Search

This research introduces a class of non-parametric identifiers based on differential neural networks represented by Hamiltonian dynamics. The structure of the identifier corresponds to the form of a canonical Hamiltonian system that uses the evolution of generalized coordinates and momentums. The learning laws of the identifier come from applying the first integrals approach, which justifies the design of an exact identifier considering the time invariance of the Hamiltonian, with a finite number of activation functions in the identifier structure. The first integrals approach derives several learning laws for the proposed class of identifiers. The learning laws design uses the estimated derivative of the generalized momentum assessed via a super-twisting differentiator with multiple inputs and outputs. All proposed laws require the solution of differential continuous-time Riccati equations and nonlinear differential equations for the learning laws, which depend on the identification error and state constraints. The developed identifier was evaluated compared to an identifier that did not consider the Hamiltonian constraint using first integrals. This comparison included a numerical evaluation of the identifier considering its application to a classical Hamiltonian system associated with the Kepler dynamics representing satellite orbital evolution. This evaluation confirmed that the identification results were improved with the proposed learning laws regarding the class of Hamiltonian structures, and the quality indicators based on the mean square error were several times lower. [ABSTRACT FROM AUTHOR]

This study presents the design of a robust control based on the sliding-mode theory to solve both; the stabilization and the trajectory tracking problems of nonlinear systems subjected to a class of full-state restrictions. The selected nonlinear system satisfies a standard Lagrangian structure affected by nonparametric uncertainties. A barrier Lyapunov function is used to ensure the state constraints by designing a time-varying gain, which guarantees the fulfillment of the predefined state constraints even under external perturbations. The proposed design methodology for the barrier sliding-mode control (BSMC) ensures the convergence of the sliding surface in finite time to the origin. Consequently, the asymptotic convergence of the states to the corresponding equilibrium point is achieved. The finite-time stability of the origin in the closed-loop system with the proposed controller has been demonstrated using the second Lyapunov stability method. The suggested controller was evaluated on a two-link robotic manipulator. Then, the obtained results showed better stabilization and tracking performances (while the restrictions are satisfied) than the traditional first-order sliding-mode or linear state feedback controllers. [ABSTRACT FROM AUTHOR]

This study developed a non-structured mathematical model for describing the effect of carbon sources and culture pH on the mannitol production by three heterofermentative lactic acid bacteria (LAB), including Leuconostoc mesenteroides NRRL B-512F and NRRL B-523 as well as Lactobacillus fermentum NRRL 8-1840. Mannitol production was studied in batch culture using four culture media containing the same carbon sources (glucose and fructose, but different glucose concentration) and different nitrogen sources (yeast extract, peptone, and meat extract) with diverse concentrations. The results obtained in the early stages of the estimation algorithm suggested that cell growth on fructose could be neglected. Therefore, the initially proposed mathematical model was redefined and subse quently confirmed. The strains NRRL B-512 F, NRRL B-523, and NRRL B-1840 reached the highest volumetric productivities (2.26 ± 0.05, 2.30 ± 0.05, and 1.95 ± 0.05 g L-lh-1; respectively) and mannitol yields (0,95 ± 0,01,0,98 ± 0.04, and 1.00 ± 0.01 mol mannitol mol initial fructose-1; respectively) in medium A. The mannitol and biomass yields suggested a higher carbon concentration favored mannitol production and growth, but a lower nimicrobial trogen concentration directly impacted both yields. Mannitol production was associated with cell growth according to the estimated parameters. [ABSTRACT FROM AUTHOR]

This study has the aim of introducing a new type of trajectory tracking robust controllers for a class of rehabilitation robotic system considering the articulations restrictions. The robotic device consists of a suspended biped configuration. The suggested robust control considers the application of state depending gains which provide finite-time convergence for the tracking deviation. The state restrictions are fulfilled by the implementation of controller gains estimated by a class of the controlled tangent barrier Lyapunov function. Stability analysis for the tracking error yields the explicit design of the state dependent gains. The rate of convergence for the controller design is enhanced using a matrix inequality convex optimization method. Based on the forward complete characteristic of the suggested rehabilitation device, it is allowed using a finite-time convergent super-twisting-based differentiator to concrete an output feedback realization of the proposed controller. A computerized model of the tendered rehabilitation robot provides a reliable testing platform to the suggested roust controller. Numerical evaluations appear to serve as an indirect confirmation for the tracking error convergence, satisfying the articulation restrictions, and the effect of the gain optimization design. For comparison purposes, the regular state feedback control design is considered as benchmark. The faster convergence of the mean square estimation of the tracking error justifies the design of the proposed control design as well as the state feedback structure justifies the origin is a fixed-time stable equilibrium point for the space of tracking error at the same time that state space restrictions remain satisfied. The experimental evaluations of the proposed controller justifies the barrier controller which, in spite of the modeling uncertainties and the implementation issues, tracked the reference trajectories. [ABSTRACT FROM AUTHOR]

A new design of a non-parametric adaptive approximate model based on Differential Neural Networks (DNNs) applied for a class of non-negative environmental systems with an uncertain mathematical model is the primary outcome of this study. The approximate model uses an extended state formulation that gathers the dynamics of the DNN and a state projector (pDNN). Implementing a non-differentiable projection operator ensures the positiveness of the identifier states. The extended form allows producing continuous dynamics for the projected model. The design of the learning laws for the weight adjustment of the continuous projected DNN considered the application of a controlled Lyapunov-like function. The stability analysis based on the proposed Lyapunov-like function leads to the characterization of the ultimate boundedness property for the identification error. Applying the Attractive Ellipsoid Method (AEM) yields to analyze the convergence quality of the designed approximate model. The solution to the specific optimization problem using the AEM with matrix inequalities constraints allows us to find the parameters of the considered DNN that minimizes the ultimate bound. The evaluation of two numerical examples confirmed the ability of the proposed pDNN to approximate the positive model in the presence of bounded noises and perturbations in the measured data. The first example corresponds to a catalytic ozonation system that can be used to decompose toxic and recalcitrant contaminants. The second one describes the bacteria growth in aerobic batch regime biodegrading simple organic matter mixture. • A non-parametric observer is realized on Differential Neural Networks (DNNs). • The observer is applied for a class of non-negative environmental systems. • The learning laws for the weight adjustment in these DNN are obtained with the Lyapunov-like approach. • The implementation of a non-differentiable projection operator ensures the positiveness of the observer states. • Attractive Ellipsoid Method has been applied to analyze the quality of the designed observer. [ABSTRACT FROM AUTHOR]

Dynamic motion simulators cannot provide the same stimulation of sensory systems as real motion. Hence, only a subset of human senses should be targeted. For simulators providing vestibular stimulus, an automatic bodily function of vestibular–ocular reflex (VOR) can objectively measure how accurate motion simulation is. This requires a model of ocular response to enforced accelerations, an attempt to create which is shown in this paper. The proposed model corresponds to a single-layer spiking differential neural network with its activation functions are based on the dynamic Izhikevich model of neuron dynamics. An experiment is proposed to collect training data corresponding to controlled accelerated motions that produce VOR, measured using an eye-tracking system. The effectiveness of the proposed identification is demonstrated by comparing its performance with a traditional sigmoidal identifier. The proposed model based on dynamic representations of activation functions produces a more accurate approximation of foveal motion as the estimation of mean square error confirms. [ABSTRACT FROM AUTHOR]

Morels are gourmet wild edible mushrooms that can grow on several substrates with significant growth rate variations. Such variations have hindered the development of a standardized culture media to promote morel's sustainable production. The aim of this study is developing a novel culture media that takes advantage of coconut water as a complementary component of culture media. Coconut water has been extensively used as a growth-promoting component for plant tissue cultures; however, its application as component of fungi cultivation medium has not been fully developed. This study confirms that coconut water can be efficiently used as culture media component for morels using a kinetic characterization. Morchella sp. kinetic growth is evaluated in different cultures: agar, malt extract agar (MEA), lactose, coconut water (15%) and combinations of them. Kinetic growth parameters (lag phase, λ and maximum specific growth rate, µmax) are estimated using primary modeling methods. Among the selected models, the best fit is achieved using Baranyi's model. A significant increase from 15.8% to 43.4% of the µmax values was observed when culture media (agar, lactose, MEA) is supplemented with coconut water. The largest values of µmax are obtained in MEA-coconut cultures (21.13 ± 0.43–22.57 ± 0.35). Micro-sclerotia and late sclerotia are observed in all cultures containing coconut water justifying the development of a feasible and cost-effective way of culturing morels. The results demonstrate that coconut water can be used for formulation of standard media for morel cultivation leading to a cheap alternative to produce dense mycelium and promote sclerotia formation. [ABSTRACT FROM AUTHOR]

This paper uses a multi-head neural transformer to present the text-to-text translation/interpretation of Sign Language (SL) in the context of glosses (written SL). A Spanish to Mexican Sign Language (MSL) gloss dataset was built based on simple and compound sentences and the corresponding interpretation in MSL gloss. The interpretation process was achieved by implementing state-of-the-art tools in the natural language processing (NLP) field called neural transformers. We tried different architectures, varying the number of encoder-decoder layers and hyperparameters. The best of our models achieved 0.68 BLEU in the training phase and 0.33 in the validation phase. MSL glosses are crucial as they rule the grammatical order in which MSL has to be executed. All these quantitative and qualitative results confirm the potential applicability of neural transformers to create effective automatic translators for the Spanish language to MSL, with similar effectiveness shown by other automatic translators for other more likely languages. [ABSTRACT FROM AUTHOR]

In this study, a local approximated solution for the Hamilton–Jacobi–Bellman equation based on differential neural networks is proposed. The approximated Value function is used to obtain a feedback control law for a class of uncertain dynamical systems. The approach to deal with the uncertainties of the dynamical system is a min–max method that yields obtaining a robust-like solution for an optimal control problem. The proposed cost functional is presented in the Bolza form and the necessary and sufficient conditions for getting the min–max optimal solution with the designed robust version of the dynamic programming approach are provided. Moreover, the effect of the neural network approximation is studied. All the analysis considers a smoothness assumption regarding the Value function. The practical stability of the system with the neural network feedback optimal control is formally demonstrated by means of the Lyapunov method. Finally, the performance of the control law is compared in simulation with a classical optimal controller. The proposed controller overcomes the results produced by the classical controller, which is confirmed by the functional evaluation. [ABSTRACT FROM AUTHOR]

The aim of this work is to design a robust output‐based controller for a robotic manipulator (RM) with a master–slave configuration (MSC). The trajectories of the slave robot manipulator (SRM) must fulfil two types of constraints. The first corresponds to the boundaries of the obstacles placed in the given workspace. The second is related to the boundaries of the workspace. The controller must ensure the collision avoidance of the obstacles and the end‐effector of the SRM should remain inside of the workspace. The proposed controller satisfies a state feedback structure with time‐varying gains. The time‐variable gains include an integral compensation that guarantees the collision avoidance with the obstacle satisfying the boundary of the workspace, which is the main contribution of this study. The solution of a matrix inequality characterises the zone of convergence and the suboptimal control gains. A set of numerical simulations using a MSC based on a couple of two‐link RMs illustrates the advantages obtained with the proposed method. Also, a set of experimental results demonstrate the effectiveness of the proposed controller. [ABSTRACT FROM AUTHOR]

The design of an artificial neural network (ANN) based sub-optimal controller to solve the finite-horizon optimization problem for a class of systems with uncertainties is the main outcome of this study. The optimization problem considers a convex performance index in the Bolza form. The dynamic uncertain restriction is considered as a linear system affected by modeling uncertainties, as well as by external bounded perturbations. The proposed controller implements a min–max approach based on the dynamic neural programming approximate solution. An ANN approximates the Value function to get the estimate of the Hamilton–Jacobi–Bellman (HJB) equation solution. The explicit adaptive law for the weights in the ANN is obtained from the approximation of the HJB solution. The stability analysis based on the Lyapunov theory yields to confirm that the approximate Value function serves as a Lyapunov function candidate and to conclude the practical stability of the equilibrium point. A simulation example illustrates the characteristics of the sub-optimal controller. The comparison of the performance indexes obtained with the application of different controllers evaluates the effect of perturbations and the sub-optimal solution. [ABSTRACT FROM AUTHOR]

Summary: In non‐cooperative multi‐player games, an average sub‐gradient (ASG) strategy for finding an ϵ$$ \epsilon $$‐Nash equilibrium. For convex (but not necessarily strictly convex) individual cost functions, the function convergence of the multi‐player game is demonstrated. Applying Tanaka's formula that characterizes the Nash equilibrium leads to finding the strategy for each player. A min‐max formulation defines the corresponding solution for each player. The main result is proving the reachability of the desired regime, obtaining an explicit upper bound for Tanaka's function decrement. Detailed analyses for two‐player and multi‐player games are developed. For a class of two‐player games with a convex performance function, a set of numerical studies shows the applicability of the proposed method. So, the tracking of the arm in 3‐links robot is considered as an illustrative example. [ABSTRACT FROM AUTHOR]

This research is dedicated to developing a min–max robust control strategy for a dynamic game involving pursuers, evaders, and defenders in a multiple-missile scenario. The approach employs neural dynamic programming, utilizing multiple continuous differential neural networks (DNNs). The competitive controller devised addresses the robust optimization of a joint cost function that relies on the trajectories of the pursuer–evader–defender system, accommodating an uncertain mathematical model while adhering to control restrictions. The dynamic programming min–max formulation facilitates robust control by accounting for bounded modeling uncertainties and external disturbances for each game component. The value function of the Hamilton–Jacobi–Bellman (HJB) equation is approximated by a DNN, enabling the estimation of the closed-loop formulation for the joint dynamic game with state restrictions. The controller’s design is grounded in estimating the state trajectory under the worst possible uncertainties and perturbations, providing a robustness factor through the robust neural controller. The learning law class for the time-varying weights in the DNN is generated by studying the HJB partial differential equation for the missile motion for each player in the dynamic game. The controller incorporates the solution of the obtained learning laws and a time-varying Riccati equation, offering an online solution to the control implementation. A recurrent algorithm, based on the Kiefer–Wolfowitz method, adjusts the initial conditions for the weights to satisfy the final condition of the given cost function for the dynamic game. A numerical example is presented to validate the proposed robust control methodology, confirming the optimization solution based on the DNN approximation for Bellman’s value function. [ABSTRACT FROM AUTHOR]

Convolutional neural networks (CNNs) are used to solve pattern classification problems. As this algorithm is based on establishing a relationship between an image-shaped input and its related output through the CNN structure, the training stage is a significant process in their working basis. This study develops a new-fangled and explainable algorithm to train CNNs. The input filters in the convolution layers are parameterized to keep the filter structure, implementing traditional and explainable image processing filters within the network topology. A back-propagation scheme updates the parameters in the filters and the fully connected section of the CNN. Several traditional image filters (Sobel, averaging, Gaussian, and directional, among others) are used in CNN with a learning strategy that keeps their kernel structures. The method implies that the training of these networks is applied to a single parameter instead of all coefficients in the filters, reducing the uncertainty about how each filter performs the image analysis in CNN. This approach was compared with traditional CNNs considering the analysis of the computational cost (measured in terms of time and floops required for training) and their accuracy results. Three image databases were used to evaluate the proposed algorithm. Using a cross-validation methodology, the new training algorithm based on the filter parameterization achieved higher accuracy (93.7% vs. 91.3% in average). The new algorithm got similar results regarding the computational cost compared to traditional methods. This characteristic makes the proposed training methodology an appropriate option to classify images with more explainable processing at the convolution layers. [ABSTRACT FROM AUTHOR]

• Comprehensive review of the artificial neural networks on environment engineering problems. • Detailed analysis of classical and recent contributions on this field. • Analysis of relevant open problems in the neural networks design for environmental applications. • Valuable analysis of potential novel contributions of continuous and recurrent neural network in the environmental engineering field. Artificial neural networks (ANNs) are considered efficient tools for modeling complex, non-linear processes with uncertain dynamic models. ANNs were originally applied as effective predictors of diverse processes with static dependence on the input-output information. However, when the ANN must be applied to characterize an approximate model of time-dependent input-output relationships, then it is necessary to introduce the time effect as part of the ANN, yielding to the construction of dynamic ANN or DNN. This review establishes the variants of recurrent and differential forms of DNN, their mathematically formulation as well as the methods to adjust the network weights. The characteristics of DNNs motivate their use to represent the dynamics of decontamination processes. This review details recent findings on the DNN application for the modeling and control of treatment systems based on either biological and chemical processes. The modeling application of DNN for some common methods used in the treatment of wastewater, contaminated soil and atmosphere is described. The major benefits of using the approximate DNN-based model instead of designing the complex mathematical description for each treatment are analyzed in the context of enhancing the efficiency of the decontamination treatment. This review also highlights the remarkable efficiency of DNNs as a keystone tool for modeling and control sequence of treatments. The last section in the review introduces several open researching areas for the application of DNN for decontamination systems based on biochemical and chemical treatments. [ABSTRACT FROM AUTHOR]

In the developed world, heart diseases are the major cause of death among adults. Often, the sufferers of heart disease are not aware of their condition until a catastrophic medical event occurs. Therefore, early online detection and continuous monitoring of abnormal heart rhythms shall reduce this occurrence. There are four main types of arrhythmia: ventricular arrhythmia, supraventricular arrhythmia, premature beats and asynchronous arrhythmia. In this study, an algorithm for automatic detection of atrial premature contraction, supraventricular tachyarrhythmias, fusion of ventricular and normal beat (FUSION), isolated QRS-like artifact (ARFCT), ST change, T-wave change, premature or ectopic supraventricular beat and normal beat (NORMAL) using a continuous neural network (CoNN) is presented. This kind of continuous classifier offers an online detection of classical arrhythmia observed in electrocardiographic (EKG) signals. Typically, due to its complexity and recursive nature of arrhythmia classification algorithms, they are difficult to be implemented in real time. In this work, automatic signal classification was attained by implementing a parallel CoNN algorithm using fixed point arithmetic on a field-programmable gate array (FPGA). First, the classification algorithm using a floating-point MATLAB implementation was developed and validated. This procedure served as a benchmark for the fixed point FPGA implementation on a Xilinx Zinq board. The performance of the classification algorithm was evaluated by using a fivefold cross-validation method, achieving a 93.80% accuracy and a sensitivity (TPR) average of 98% when performing the classification of the entire set of EKG signal samples. [ABSTRACT FROM AUTHOR]

A novel culture medium to enhance the biomass and lipid production simultaneously by Chlorella vulgaris UTEX 26 was designed in three stages of optimization. Initially, a culture medium was inferred applying the response surface method to adjust six factors [NaNO 3 , NH 4 HCO 3 , MgSO 4 ·7H 2 O, KH 2 PO 4 , K 2 HPO 4 and (NH 4 ) 2 HPO 4 ], which were selected on the basement of BBM (Bold’s Basal Medium) and HAMGM (Highly Assimilable Minimal Growth Medium) culture media. Afterwards, the nitrogen source compound was optimized to reduce both, ammonium and nitrate concentrations. As result of the optimization process, the proposed culture medium improved 40% the biomass (0.73 g L −1 ) compared with the BBM medium and 85% the lipid concentration (281 mg L −1 ), with respect to HAMGM medium. Some culture media components concentrations were reduced up to 50%. Gas chromatography analysis revealed that C16:0, C18:0, C18:1, C18:2 and C18:3 were the major fatty acids produced by C. vulgaris UTEX 26. [ABSTRACT FROM AUTHOR]

This article presents a simultaneous state (via a nonlinear form of Luenberger observer) and parameter (using a proportional–integral least mean square form) estimator design method for a conventional ozonation system. The suggested state observer assumes that the only available output signal is the concentration of the ozone gas at the output of the reactor. The estimation of the reaction rate constants of ozonation in the presence of contaminants uses the suggested proportional–integral estimation method. The convergence proof of the developed state-parameter identification method was confirmed using a Lyapunov based stability analysis. This analysis characterizes the quality of estimation considering the presence of modeled uncertainties and external perturbations. The implementation of the super-twisting algorithm as a robust and exact differentiator allowed to perform the estimation of the reaction rate constants of the ozonation, the temporal evolution of the dissolved ozone and the evolution of contaminants concentrations. The simultaneous state and parameter estimator design method was implemented in real-time using phenol as a model contaminant. The numerically simulated and real-time implementations showed that the method provides accurate estimates of the contaminant concentration and the reaction rate coefficient in all the evaluated cases. • Simultaneous state and parameter estimation for conventional ozonation system is presented. • Identification of reaction rates of ozone-contaminants uses a proportional–integral algorithm. • The parameter identification is developed based on a Lyapunov-like function. • Numerical simulations demonstrate the successful of proposed algorithm. • State observer assumes that the available output signal is the ozone concentration at the reactor output. [ABSTRACT FROM AUTHOR]

The human gut is home to a complex array of microorganisms interacting with the host and each other, forming a community known as the microbiome. This community has been linked to human health and disease, but understanding the underlying interactions is still challenging for researchers. Standard studies typically use high-throughput sequencing to analyze microbiome distribution in patient samples. Recent advancements in meta-omic data analysis have enabled computational modeling strategies to integrate this information into an in silico model. However, there is a need for improved parameter fitting and data integration features in microbial community modeling. This study proposes a novel alternative strategy utilizing state-of-the-art dynamic flux balance analysis (dFBA) to provide a simple protocol enabling accurate replication of abundance data composition through dynamic parameter estimation and integration of metagenomic data. We used a recurrent optimization algorithm to replicate community distributions from three different sources: mock, in vitro, and clinical microbiome. Our results show an accuracy of 98% and 96% when using in vitro and clinical bacterial abundance distributions, respectively. The proposed modeling scheme allowed us to observe the evolution of metabolites. It could provide a deeper understanding of metabolic interactions while taking advantage of the high contextualization features of GEM schemes to fit the study case. The proposed modeling scheme could improve the approach in cases where external factors determine specific bacterial distributions, such as drug intake. [ABSTRACT FROM AUTHOR]

The properties of robustness and finite‐time convergence provided by sliding mode (SM) theory have motivated several researches to deal with the problems of control and state estimation. In the SM theory, the super‐twisting algorithm (STA), a second‐order SM scheme, has demonstrated remarkable characteristics when it is implemented as a controller, observer or robust signal differentiator although the presence of noise and parametric uncertainties. However, the design of this algorithm was originally developed for continuous‐time systems. The growth of microcomputers technology has attracted the attention of researchers inside the SM discrete‐time domain. Recently, discretisations schemes for the STA were studied using majorant curves. In this study, the stability analysis in terms of Lyapunov theory is proposed to study a discrete‐time super twisting‐like algorithm (DSTA) for non‐linear discrete‐time systems. The objective is to preserve the STA characteristics of robustness in a quasi‐sliding mode regime that was proved in terms of practical Lyapunov stability. An adequate combination of gains obtained by the same Lyapunov analysis forces the convergence for the DSTA. The problem of state estimation is also analysed for second‐order mechanical systems of n degrees of freedom. Simulation results regarding the design of a second‐order observer using the DSTA for a simple pendulum and a biped model of seven degrees of freedom are presented. [ABSTRACT FROM AUTHOR]

In this study, a neuro-controller with adaptive deadzone compensation for a class of unknown SISO non-linear systems in a Brunovsky form with uncertain deadzone input is presented. Based on a proper smooth parameterisation of the deadzone, the unknown dynamics is identified by using a continuous time recurrent neural network whose weights are adjusted on-line by stable differential learning laws. On the basis of this neural model so obtained, a feedback linearisation controller is developed in order to follow a bounded reference trajectory specified. By means of Lyapunov analysis, the boundedness of all the closed-loop signals as well as the weights and deadzone parameter estimations is rigorously proven. Besides, the exponential convergence of the actual tracking error to a bounded zone is guaranteed. The effectiveness of this scheme is illustrated by a numerical simulation. [ABSTRACT FROM AUTHOR]

Abstract: This paper addresses the design of a discontinuous finite time convergent learning law for neural networks with continuous dynamics. The neural network was used here to obtain a non-parametric model for uncertain systems described by a set of ordinary differential equations. The source of uncertainties was the presence of some external perturbations and poor knowledge of the nonlinear function describing the system dynamics. A new adaptive algorithm based on discontinuous algorithms was used to adjust the weights of the neural network. The adaptive algorithm was derived by means of a non-standard Lyapunov function that is lower semi-continuous and differentiable in almost the whole space. A compensator term was included in the identifier to reject some specific perturbations using a nonlinear robust algorithm. Two numerical examples demonstrated the improvements achieved by the learning algorithm introduced in this paper compared to classical schemes with continuous learning methods. The first one dealt with a benchmark problem used in the paper to explain how the discontinuous learning law works. The second one used the methane production model to show the benefits in engineering applications of the learning law proposed in this paper. [Copyright &y& Elsevier]

In this study, a neuro‐controller with adaptive deadzone compensation for a class of unknown SISO non‐linear systems in a Brunovsky form with uncertain deadzone input is presented. Based on a proper smooth parameterisation of the deadzone, the unknown dynamics is identified by using a continuous time recurrent neural network whose weights are adjusted on‐line by stable differential learning laws. On the basis of this neural model so obtained, a feedback linearisation controller is developed in order to follow a bounded reference trajectory specified. By means of Lyapunov analysis, the boundedness of all the closed‐loop signals as well as the weights and deadzone parameter estimations is rigorously proven. Besides, the exponential convergence of the actual tracking error to a bounded zone is guaranteed. The effectiveness of this scheme is illustrated by a numerical simulation. [ABSTRACT FROM AUTHOR]

State estimation for uncertain systems affected by external noises is an important problem in control theory. This paper deals with a state observation problem when the dynamic model a plant contains uncertainties or it is completely unknown. Differential neural network (NN) approach is applied in this uninformative situation but with activation functions described by wavelets. new learning law, containing an adaptive adjustment rate, is suggested to imply the stability condition for the free parameters the observer. Nominal weights are adjusted during the preliminary training process using the least mean square (LMS) method. Lyapunov theory is used to obtain the upper bounds for the weights dynamics as well as for the mean squared estimation error. Two numeric examples illustrate this approach: first, a nonlinear electric system, governed by the Chua's equation and second the Lorentz oscillator. Both systems are assumed to be affected by external perturbations and their parameters are unknown. [ABSTRACT FROM AUTHOR]

This study presents an innovative modeling approach for systems with uncertain dynamics, utilizing a novel long-short-term memory (LSTM) architecture featuring continuous temporal evolution. The proposed approximate model is resolved by introducing a non-parametric identifier, the trajectories converge to the states of systems exhibiting uncertain dynamics in an ultimately bounded manner. Given the multi-layered nature of continuous dynamics, we introduce two identifiers, considering the possibility of accessing or not accessing the LSTM's internal state. Consequently, two identifiers have been devised, considering the influence of the internal state of the LSTM. Applying a control Lyapunov function to each identifier enables the derivation of weight evolution in the LSTM's two layers: the output and the hidden layers. These weights are linked to input states' short- and long-term effects on the LSTM's dynamics. To validate the proposed identifiers' effectiveness compared to traditional Hopfield-like differential neural networks, we provide a numerical example and conduct complementary experimental validations. These results confirm the theoretical findings regarding the impact of short and long-term temporal information on the LSTM's current state and demonstrate superior identification quality compared to traditional neural networks exhibiting continuous dynamics adhering to the Hopfield form. Collectively, these findings substantiate the advancements presented in this study. [ABSTRACT FROM AUTHOR]

In recent years, hundreds of technical papers have been published which describe the use of sliding mode control (SMC) techniques for power electronic equipment and electrical drives. SMC with discontinuous control actions has the potential to circumvent parameter variation effects with low implementation complexity. The problem of controlling time-varying DC loads has been studied in literature if three-phase input voltage sources are available. The conventional approach implies the design of a three-phase AC/DC converter with a constant output voltage. Then, an additional DC/DC converter is utilized as an additional stage in the output of the converter to generate the required voltage for the load. A controllable AC/DC converter is always used to have a high quality of the consumed power. The aim of this study is to design a controlled continuous signal generator based on the sliding mode control of a three-phase AC-DC power converter, which yields the production of continuous variations of the output DC voltage. A sliding mode current tracking system is designed with reference phase currents proportional to the source voltage. The proportionality time-varying gain is selected such that the output voltage is equal to the desired time function. The proposed new topology also offers the capability to get rid of the additional DC/DC power converter and produces the desired time-varying control function in the output of AC/DC power converter. The effectiveness of the proposed control design is demonstrated through a wide range of MATLAB/Simulink simulations. [ABSTRACT FROM AUTHOR]

The aim of this study is to design a non-parametric identifier based on Differential Neural Networks (DNNs) for a class of positive systems described by uncertain mathematical models. The inclusion of state constraints and the existence of equilibrium points outside the origin are considered in the design of the non-parametric identifier with the implementation of asymmetric barrier Lyapunov functions. The application of a stability analysis yields the design of learning laws for the weights adjustment. A class of hybrid learning laws depending on the relative difference of each state with respect to its corresponding component of the equilibrium point provides the ability of handling the positiveness of all the states, which is ensured considering he implementation of non-linear state dependent gains. A numerical example confirms the efficiency of the proposed state non-parametric identifier in the presence of bounded noises and perturbations affecting the dynamics of the evaluated positive systems. The example corresponds to a pharmaceutical compartmental system which reproduces the immunotherapy dynamics for the cancer treatment. The comparison of the proposed DDN approximated model with the classical non-barrier identifier confirms the ability of reproducing positive systems trajectories satisfying state constraints. [ABSTRACT FROM AUTHOR]

Many studies relating the gut microbiota composition and autism spectrum disorder focus on finding the statistical differences in microbiome composition between neurotypical and autistic subjects. Since microbiota composition involves high-dimensional variables, establishing inferential or causal relationships using only statistical information is complex, hindering advances toward early functional treatment. Complementary machine learning strategies related to the study of autism spectrum disorder are focused on early diagnosis, substituting the expensive screening tests without providing a possible guide to future alternatives to reduce autism spectrum disorder symptoms. Such techniques may offer better outcomes as a direct approach complemented with statistical analysis to optimize patient healthcare based on an early and simplified detection process. This work evaluates several classic machine learning models, including random forests, support vector machines, k-nearest neighbors, Naïve Bayes, and artificial neural network models. The developed models were assessed to identify less-known patterns and their underlying structures to prior published research on the relationship between gut microbiome composition and an autism spectrum disorder. The differences and similarities between the discovered patterns and existing research are discussed to detect a minimal set of strains that may define the presence of autism spectrum disorder. The best-evaluated models were an artificial neural network and a k-nearest neighbor model, reaching an accuracy of 94.7% in the testing partition with only two missed classifications from 38 previously unseen testing samples. These outcomes support the potential of machine learning strategies to construct a useful pre-diagnostic tool for autism spectrum disorder based on relative gut microbiome distribution. [ABSTRACT FROM AUTHOR]

The current study presents the treatment of water contaminated with the dye Direct Red 28 in the presence of Na2SO4 at three different concentrations (10, 40 and 80 g/L) by simple ozonation. This wastewater treatment was considered in a wastewater recycling process. At the end of each ozonation cycle, ozonation dynamics, ozone consumption, pH, electrical conductivity and UV/Vis spectra variation were analyzed and correlated with the stoichiometric analysis balance. The efficiency of the wastewater treatment in each reuse cycle was evaluated by a novel image processing method. The image results show that the presence of sulfate improves more than 45% the dye fixation on a cotton sample. However, at high additive concentrations, the color quality decreases 25%, compared with the system with low additive concentration, due to the presence of peroxysulfate ions and the compounds accumulated through the ozonation process. [ABSTRACT FROM AUTHOR]

• A feasible model describes the lactic acid production by Lactobacillus casei using residual agro-industrial wastes. • The model includes a delay characterizing the lagged biomass growth by the acclimation forced by the complex substrate. • A robust differentiator and a time-delay least mean square identifier yields the parameters estimation. • The model reproduced the experimental data with correlation factor of 0.98 in average for all the evaluated conditions. The aim of this study was to obtain a feasible mathematical model that may describe the lactic acid production by Lactobacillus casei using residual agro-industrial wastes. The model includes a delay term which characterizes the lagged biomass growth by the acclimation period induced by the considered complex substrate. A novel parametric identification algorithm leads to the characterization of the time delay as well as the kinetic parameters. The application of the estimated parameters in the proposed model defined the behavior of biomass growth, substrate consumption and lactic acid accumulation. A combination of robust exact differentiators (based on the sliding mode algorithm named super twisting) and a time-delay least mean square identification algorithm provides the on-line calculus of the required parameters. The identification algorithm was tested with the experimental data obtained from a bacterial culture system containing all the required nutritional sources to complete the target carbon to nitrogen ratio of 4.13. This value has been characterized as an adequate ratio to promote the efficient production of lactic acid. The proposed model reproduced the experimental data with correlation factor of 0.98 in average for the evaluated experimental conditions. The identified parameters agreed with the reported results in similar studies showing the effectiveness of the identification method as well as the possibility of using non-conventional substrates (milk whey and vegetable residues) to produce lactic acid. [ABSTRACT FROM AUTHOR]

This study focuses in the output feedback stabilisation of constrained linear systems affected by uncertainties and noisy output measurements. The system states are restricted inside a given polytope and a classical Luenberger observer is used to reconstruct the unmeasurable states from output observations. Based on the observed states, a state feedback is proposed as the control input. The stability analysis and the control design are done using an extended version of the attractive ellipsoid method (AEM) approach. To avoid the violation of state constraints, this work proposes a barrier Lyapunov function (BLF) based analysis. The control parameters are obtained throughout the solution of some optimisation problems such that the BLF ensures an approximation of the constraints by a maximal ellipsoidal set and the AEM provides the characterisation of a minimal ultimately bounded set for the closed-loop system solutions. Numerical simulations show the advantages using the BFL-AEM methodology against classical sub-optimal controllers in academic second order and third order examples. Then, the proposed control strategy is applied over a Buck DC-DC converter. In all the cases, the method proposed here prevails over the other controllers. [ABSTRACT FROM AUTHOR]

This work presents the development of a state-input affine differential neural network that approximates a class of nonlinear systems with uncertain dynamics and perturbations. The feasibility of using differential neural networks is based on the approximation capabilities of artificial neural networks, assuring that the approximation error of the right-hand side of the system with uncertain dynamics is bounded. The dynamics of the free parameters of the differential neural network is obtained using the Lyapunov's second stability method, which provides the convergence of the identification error to an invariant set around the origin. The approximation capabilities of the proposed identifier are compared with a classical differential neural network, aiming to identify the dynamics of a Stewart platform, which is used as testing example. The results show that the cumulative norm of identification's error is smaller using the state-input affine differential neural network, which demonstrates better identification outcomes with the proposed modified identifier. An additional advantage of the state-input affine differential neural network is that it has a quasi-linear structure that could help formulate a controller for several systems, which is a further research interest. • Learning laws for State-input affine differential neural network (SIADNN) identifier. • Design based on the second stability method of Lyapunov. • The novel DNN identifier has more applications than the traditional DNN identifier. • Evaluation of performing the non-parametric identification of a Stewart platform. [ABSTRACT FROM AUTHOR]

This work reports the study of the decomposition of three PAHs (anthracene, phenanthrene, and fluorene) in aqueous solution at different pHs (2, 7, and 11) and in the presence of methanol as a co-solvent by two different methods: conventional ozonation (O3) and O3/activated carbon (AC). The results showed that without AC, the decomposition of anthracene and fluorene proceeded mostly by molecular O3 at pHs 2 and 7. The indirect ozonation mechanism expected at basic pH was inhibited by the presence of high concentrations of methanol, which acted as a radical scavenger. The ozone consumed by the radical formation under pH >7 induced a larger decomposition reaction rate (five times) for all PAHs at acidic and neutral conditions compared with the basic ones. The presence of AC modulated the decomposition rates for all PAHs at the pHs evaluated. The latter was confirmed by the similarity among the decomposition dynamics. [ABSTRACT FROM AUTHOR]

Mathematical modelling applied to biological systems allows for the inferring of changes in the dynamic behaviour of organisms associated with variations in the environment. Models based on ordinary differential equations are most commonly used because of their ability to describe the mechanisms of biological systems such as transcription. The disadvantage of using this approach is that there is a large number of parameters involved and that it is difficult to obtain them experimentally. This study presents an algorithm to obtain a finite-time parameter characterization of the model used to describe changes in the metabolic behaviour of Escherichia coli associated with environmental changes. In this scheme, super-twisting algorithm was proposed to recover the derivative of all the proteins and mRNA of E. coli associated to changes in the concentration of oxygen available in the growth media. The 75 identified parameters in this study maintain the biological coherence of the system and they were estimated with no more than 20 % error with respect to the real ones included in the proposed model. [ABSTRACT FROM AUTHOR]

The aim of this study was to characterize the in vivo volumetric distribution of three folate-based biosensors by different imaging modalities (X-ray, fluorescence, Cerenkov luminescence, and radioisotopic imaging) through the development of a tridimensional image reconstruction algorithm. The preclinical and multimodal Xtreme imaging system, with a Multimodal Animal Rotation System (MARS), was used to acquire bidimensional images, which were processed to obtain the tridimensional reconstruction. Images of mice at different times (biosensor distribution) were simultaneously obtained from the four imaging modalities. The filtered back projection and inverse Radon transformation were used as main image-processing techniques. The algorithm developed in Matlab was able to calculate the volumetric profiles of 99mTc-Folate-Bombesin (radioisotopic image), 177Lu-Folate-Bombesin (Cerenkov image), and FolateRSense™ 680 (fluorescence image) in tumors and kidneys of mice, and no significant differences were detected in the volumetric quantifications among measurement techniques. The imaging tridimensional reconstruction algorithm can be easily extrapolated to different 2D acquisition-type images. This characteristic flexibility of the algorithm developed in this study is a remarkable advantage in comparison to similar reconstruction methods. [ABSTRACT FROM AUTHOR]

This study designed a suboptimal output control strategy to characterise an attractive and invariant set for the state trajectories of perturbed linear systems with noisy measurements and state constraints. The state constraints were defined by a given polytope formed of by n-dimensional vectors. An adaptive linear controller enforced the existence of an attractive and invariant set (centred at the origin) for the trajectories of the perturbed system. A barrier Lyapunov function (BLF) and the attractive ellipsoid method (AEM) derived the adjustment law of the adaptive gain. The controller design used the linear matrix inequality technique to solve two optimisation problems. The first solution provided a maximal set where the initial conditions must belong without violating the state constraints. The second optimisation solution characterised the invariant minimal attractive set for the system trajectories. An academic example verified how the proposed adaptive control generated the system trajectories that converged to the minimal attractive ellipsoid while keeping them inside the polytope defining the state constraints. The simulation script showed the advantages of the adaptive BLF controller (ABLC) against classical AEM controller. A second numerical example considered a direct current motor showing the advantages of the ABLC against the sliding mode technique. [ABSTRACT FROM AUTHOR]

This study designed a suboptimal output control strategy to characterise an attractive and invariant set for the state trajectories of perturbed linear systems with noisy measurements and state constraints. The state constraints were defined by a given polytope formed of by n ‐dimensional vectors. An adaptive linear controller enforced the existence of an attractive and invariant set (centred at the origin) for the trajectories of the perturbed system. A barrier Lyapunov function (BLF) and the attractive ellipsoid method (AEM) derived the adjustment law of the adaptive gain. The controller design used the linear matrix inequality technique to solve two optimisation problems. The first solution provided a maximal set where the initial conditions must belong without violating the state constraints. The second optimisation solution characterised the invariant minimal attractive set for the system trajectories. An academic example verified how the proposed adaptive control generated the system trajectories that converged to the minimal attractive ellipsoid while keeping them inside the polytope defining the state constraints. The simulation script showed the advantages of the adaptive BLF controller (ABLC) against classical AEM controller. A second numerical example considered a direct current motor showing the advantages of the ABLC against the sliding mode technique. [ABSTRACT FROM AUTHOR]

The aim of this work was to design a non-parametric model of interconnected systems represented by uncertain second-order systems with incomplete information (only the generalized position vector is measurable). Artificial neural networks appeared to be a plausible alternative to get a non-parametric representation of the aforementioned interconnected systems. The modelling strategy used a set of spatial distributed second-order continuous neural networks (CNN). Each node in the interconnected system was represented as a second-order continuous neural network added by the super-twisting discontinuous sliding mode algorithm. The non-parametric modelling problem was reduced to design a feasible expression for the CNN weights in order to reproduce the states (including the generalized derivative of position vector) of all the nodes dynamics together and simultaneously. The adaptive laws for the CNN weights ensured the convergence of the CNN trajectories to the states of the uncertain interconnected system. To investigate the qualitative behaviour of the suggested methodology, two numerical examples were proposed. The first one represents the interconnection of three mass–spring–damper mechanical systems. The second example considers the problem of the non-parametric modelling problem for a wave partial differential equation. A set of three-dimensional graphic representations were used to demonstrate the identification abilities achieved by the CNN designed in this study for the second case. [ABSTRACT FROM AUTHOR]

This research focuses on designing a min–max robust control based on a neural dynamic programming approach using a class of continuous differential neural networks (DNNs). The proposed controller solves the robust optimization of a proposed cost function that depends on the trajectories of a system with an uncertain mathematical model satisfying a class of non-linear perturbed systems. The dynamic programming min–max formulation enables robust control concerning bounded modelling uncertainties and disturbances. The Hamilton–Jacobi–Bellman (HJB) equation's value function, approximated by a DNN, permits to estimate the closed-loop formulation of the controller. The controller design is based on an estimated state trajectory with the worst possible uncertainties/perturbations that provide the degree of robustness using the proposed controller. The class of learning laws for the time-varying weights in the DNN is produced by studying the HJB partial differential equation. The controller uses the solution of the obtained learning laws and a time-varying Riccati equation. A recurrent algorithm based on the Kiefer–Wolfowitz method leads to adjusting the initial conditions for the weights to satisfy the final condition of the given cost function. The robust control suggested in this work is evaluated using a numerical example confirming the optimizing solution based on the DNN approximate for Bellman's value function. [ABSTRACT FROM AUTHOR]

This study describes the development (design, construction, instrumentation, and control) of a nursing mobile robotic device to monitor vital signals in home-cared patients. The proposed device measures electrocardiography potentials, oxygen saturation, skin temperature, and non-invasive arterial pressure of the patient. Additionally, the nursing robot can supply assistance in the gait cycle for people who require it. The robotic device's structural and mechanical components were built using 3D-printed techniques. The instrumentation includes electronic embedded devices and sensors to know the robot's relative position with respect to the patient. With this information together with the available physiological measurements, the robot can work in three different scenarios: (a) in the first one, a robust control strategy regulates the mobile robot operation, including the tracking of the patient under uncertain working scenarios leading to the selection of an appropriate sequence of movements; (b) the second one helps the patients, if they need it, to perform a controlled gait-cycle during outdoors and indoors excursions; and (c) the third one verifies the state of health of the users measuring their vital signs. A graphical user interface (GUI) collects, processes, and displays the information acquired by the bioelectrical amplifiers and signal processing systems. Moreover, it allows easy interaction between the nursing robot, the patients, and the physician. The proposed design has been tested with five volunteers showing efficient assistance for primary health care. [ABSTRACT FROM AUTHOR]

This study presents a novel robust controller strategy for submersible autonomous robotized vehicles (SARV). This controller applies the averaged sub-gradient (ASG) descendant method to optimize the tracking of well-posed reference trajectories. The motion control form of the SARV is done in three stages (cascade-like control design). The phases consist of using the translation velocities as pseudo controllers to adjust the SARV position, regulating the angular velocities to attain the requested translation velocities, and using the proposed ASG to control the thrusters to get the needed angular velocities. ASG implementation optimizes a convex cost function depending on the integral of the tracking error using the ASG method. The application of Barbalat's lemma justifies the tracking error is converging the origin asymptotically. A class of integral sliding mode controllers with a variable state-dependent gain solves the tracking of the designed reference trajectories in each of the three stages. The sliding surface depends on the tracking error for each pseudo-controller, its integral, and the cost function average. The optimization of the cost function can be done without complete knowledge of the SARV dynamics. A numerical example is presented in this study to confirm the suggested control design's effectiveness based on the cost function's time evolution analysis. The forced motion by the proposed controller is compared with the movement obtained by a proportional–integral–derivative (PID) controller. The proposed controller exhibits a better tracking of the reference trajectory than the PID version. • This study presents a robust robust controller for underwater robotic autonomous devices. • An averaged sub-gradient (ASG) descendant control minimizes the trajectory tracking error. • Sub-gradient realization solves an extremum seeking control to stabilize the tracking error. • The controller is of sliding mode type with variable gain based on a state dependent sliding surface. • The dynamic nature of the MRAD regulated by a back-stepping like controller design. [ABSTRACT FROM AUTHOR]

Ozonated oils have demonstrated promising results for clinical applications. The reaction of ozone with the unsaturated compounds of oils produces by-products such as ozonides and poly peroxides. A deeper knowledge of thedynamics of by-product formation is helpful in determining the required ozonation degree to obtain therapeutic effects. Theaimof this paper is to showthe relationship between ozonation degree and structuralandviscosity changes during the ozonation of grape seed(GS) and sunflower (SF) oils. Structural characterization was done by Fourier transform infrared (FT-IR) and hydrogen-1 nuclear magnetic resonance (¹H NMR) spectroscopy, with iso-ozonides being identified. Viscosity showed a significant increase during ozonation, a fact associated with poly peroxide formation. We have made use of the total unsaturation(TU) method tomeasure the ozonation degree. TheTUof non-ozonatedGSoilwas found to be higher than for SF oil (5.94 and 4.49mmol per g of oil, respectively), and their by-product distributions were also found to differ. InGSoil, three reaction stepswere observed for double-bond conversion into isoozonides and poly peroxides, while the ozonides and poly peroxides were formed in parallel in SF oil. The studies we implemented characterized the differences in the reactivities of these oils with ozone. [ABSTRACT FROM AUTHOR]

Physiological solutions ozonated are widely used in the medical field (dentistry and surgery) as an effective bactericide. In this investigation, the inactivation of Escherichia coli and Pseudomonas aeruginosa by ozone dissolved in physiological solution was studied. There is a poor knowledge of the inactivation efficiency of this solution for different bacteria. The efficiency of the microorganisms’ inactivation was evaluated by the total unsaturation of lipids measured by the so-called Double Bond Index (DB-index). This is a sensitive analysis to evaluate the quantity of carbon-carbon double bonds (>C=C<) available in organic and biological samples, with high efficiency and in a short time. DB-index results were compared with the quantity of colony forming units (CFU) available in the culture. Three experimental systems were evaluated to determine the relationship between the DB-index variable and the microorganism’s inactivation: 1) using BHI agar as a culture medium to evaluate the dynamic growing curve when the ozone dissolved was dosed over the strain’s surface; 2) using a glucose solution (5%) as culture media and keeping the ozone concentration constant, which was dissolved in different physiological solutions, to observe the effects of solvent type over the bacteria growth; and, 3) using a glucose solution (5%) as culture media, and physiological solution of NaCl (0.9%) as a dissolved media for ozone at different concentrations. From the experimental data, a model of the ozone inactivation of each pathogen was built to obtain the inactivation kinetics. The model obtained showed a correlation between the CFU behavior and DB-index to each bacteria, since Pseudomona aeruginosa was more resistant to being oxidized than Escherichia coli. [ABSTRACT FROM PUBLISHER]

The interaction between magnetic objects has been widely explored due to their range of applications, which usually involves complex calculations regarding magnetic forces. This paper presents approximation functions to describe the force generated between cylindrical electromagnets and spherical permanent magnets. These functions consider the electrical current that passes through the electromagnets, and the geometry of the devices to associate their relationship through parametric adjustments. A nonlinear least mean square parameter estimation method yields electromagnetic functions in axial and lateral directions in the relative configuration of the magnetic devices. An own developed measurement device is used to get the experimental data that leads to the functions characterization. The functions are developed considering a piecewise approximation depending on the relative distance between the electrical and permanent magnets. The modeling strategy leads to an accurate and precise representation of the force between the magnetic devices, as the correlation index and relative errors indicate. • The proposal of a novel model to approximate the magnetic force between electromagnets and spherical permanent magnets. • The development of an experimental testing device to characterize the magnetic interaction between magnetic elements. • The evaluation of the proposed model using quality indicators such as relative errors and correlation indexes. [ABSTRACT FROM AUTHOR]