Search

Diabetes Mellitus is a serious metabolic condition for global health associations. Recently, the number of adults, adolescents and children who have developed Type 1 Diabetes Mellitus (T1DM) has increased as well as the mortality statistics related to this disease. For this reason, the scientific community has directed research in developing technologies to reduce T1DM complications. This contribution is related to a feedback control strategy for blood glucose management in population samples of ten virtual adult subjects, adolescents and children. This scheme focuses on the development of an inverse optimal control (IOC) proposal which is integrated by neural identification, a multi-step prediction (MSP) strategy, and Takagi-Sugeno (T-S) fuzzy inference to shape the convenient insulin infusion in the treatment of T1DM patients. The MSP makes it possible to estimate the glucose dynamics 15 min in advance; therefore, this estimation allows the Neuro-Fuzzy-IOC (NF-IOC) controller to react in advance to prevent hypoglycemic and hyperglycemic events. The T-S fuzzy membership functions are defined in such a way that the respective inferences change basal infusion rates for each patient's condition. The results achieved for scenarios simulated in Uva/Padova virtual software illustrate that this proposal is suitable to maintain blood glucose levels within normoglycemic values (70-115 mg/dL); furthermore, this level remains less than 250 mg/dL during the postprandial event. A comparison between a simple neural IOC (NIOC) and the proposed NF-IOC is carried out using the analysis for control variability named CVGA chart included in the Uva/Padova software. This analysis highlights the improvement of the NF-IOC treatment, proposed in this article, on the NIOC approach because each subject is located inside safe zones for the entire duration of the simulation.

Motivated by the increasing complexity of systems to be controlled and different system components that cause uncertainties, threats, and disturbances, among other system stressing phenome...

Breast cancer is one of the more aggressive diseases with a high worldwide death rate in women. The early detection and timely treatment are the best methods to save life patients. Thus, the development of a new technique for medical research can represent an invaluable tool. The aim of this paper is to present an accurate and efficient numerical technique based in The Fast Slant Stack Transform as an aid to detect spiculated masses and architectural tissue distortion hidden in digital mammograms. This technique is efficient because performs this task in a very short computer processing time. The technique is validated testing 50 study cases taken from a scientific medical mammogram database. Finally, a sensitivity analysis confirms the accuracy of the here proposed technique as a promising one option particularly in developing countries where digital mammograms are highly preferred for the reduced diagnostic costs.

This paper presents an improved extension of the previous algorithm of the authors called KAdam that was proposed as a combination of a first-order gradient-based optimizer of stochastic functions, known as the Adam algorithm and the Kalman filter. In the extension presented here, it is proposed to filter each parameter of the objective function using a 1-D Kalman filter; this allows us to switch from matrix and vector calculations to scalar operations. Moreover, it is reduced the impact of the measurement noise factor from the Kalman filter by using an exponential decay in function of the number of epochs for the training. Therefore in this paper, is introduced our proposed method sKAdam, a straightforward improvement over the original algorithm. This extension of KAdam presents a reduced execution time, a reduced computational complexity, and better accuracy as well as keep the properties from Adam of being well suited for problems with large datasets and/or parameters, non-stationary objectives, noisy and/or sparse gradients.

This work focuses on the design of an intelligent controller that is a considerably large challenge for cyber-physical systems. The proposed controller can deal with unknown dynamics, actuator saturation, unknown external and internal disturbances, unknown communication delays and packet losses. Such a controller is designed using a discrete-time approach based on inverse optimal control and a recurrent high-order neural network identifier. The applicability of the proposed scheme is shown through real-time results using a tracked robot platform controlled through a wireless network under different network scenarios.

Influenza A virus infections are causes of severe illness resulting in high levels of mortality. Neuraminidase inhibitors such as zanamivir and oseltamivir are used to treat influenza; however, treatment recommendations remain debatable. In this paper, a discrete-time inverse optimal impulsive control scheme based on passivation is proposed to address the antiviral treatment scheduling problem. We adapt results regarding stability, passivity, and optimality for the impulsive action. The study is founded on mathematical models whose parameters are adjusted to data from clinical trials where participants were experimentally infected with influenza H1N1 and treated with either zanamivir or oseltamivir. Simulation results show that control-based techniques can reduce the amount of medication while simultaneously reach the efficacy levels of the treatment schedules by the Food and Drug Administration. Monte Carlo simulations disclose the robustness of the proposed control-based techniques.

Infectious diseases are latent threats to humankind killing annually millions worldwide. Disease transmission modeling and control remain a central vexation for science as it involves several complex and dynamic processes. In this paper, multiscale stochastic impulsive models in combination with contact patterns are presented to describe outbreaks and epidemics more accurately. The new families of mathematical models open up a new path within the field of control theory with long-term impact and ample opportunities to control epidemics.

Wind energy is one of the most promising alternatives as energy sources, however, to obtain the best results, producers need to forecast the wind speed, generated power and energy price in order to provide the appropriate tools for optimal operation, planning, control and marketing both for isolated wind systems and for those that are interconnected to a main distribution network. For the present work, a novel methodology is proposed for the forecasting of time series in wind energy systems, it consists of a high-order neural network that is trained on-line by the extended Kalman filter algorithm. Unlike most modern artificial intelligence methods of forecasting, which are based on hybridizations, data pre-filtering or deep learning methods, the proposed method is based on the simplicity of implementation, low computational complexity and real-time operation to produce 15-step-ahead forecasting in a time series of wind speed, generated power and energy price. The proposed scheme has been evaluated using real data from open access repositories of wind farms. The results show that an on-line training of the neural network produces high precision, without the need for any other information beyond a few past observations.

Artificial intelligence techniques have been used in the industry to control complex systems; among these proposals, adaptive Proportional, Integrative, Derivative (PID) controllers are intelligent versions of the most used controller in the industry. This work presents an adaptive neuron PD controller and a multilayer neural PD controller for position tracking of a mobile manipulator. Both controllers are trained by an extended Kalman filter (EKF) algorithm. Neural networks trained with the EKF algorithm show faster learning speeds and convergence times than the training based on backpropagation. The integrative term in PID controllers eliminates the steady-state error, but it provokes oscillations and overshoot. Moreover, the cumulative error in the integral action may produce windup effects such as high settling time, poor performance, and instability. The proposed neural PD controllers adjust their gains dynamically, which eliminates the steady-state error. Then, the integrative term is not required, and oscillations and overshot are highly reduced. Removing the integral part also eliminates the need for anti-windup methodologies to deal with the windup effects. Mobile manipulators are popular due to their mobile capability combined with a dexterous manipulation capability, which gives them the potential for many industrial applications. Applicability of the proposed adaptive neural controllers is presented by simulating experimental results on a KUKA Youbot mobile manipulator, presenting different tests and comparisons with the conventional PID controller and an existing adaptive neuron PID controller.

Mobile robots commonly have to traverse rough terrains. One way to find the easiest traversable path is by determining the types of terrains in the environment. The result of this process can be used by the path planning algorithms to find the best traversable path. In this work, we present an approach for terrain classification from aerial images while using a Convolutional Neural Networks at the pixel level. The segmented images can be used in robot mapping and navigation tasks. The performance of two different Convolutional Neural Networks is analyzed in order to choose the best architecture.

In this paper, an adaptive single neuron Proportional&ndash, Integral&ndash, Derivative (PID) controller based on the extended Kalman filter (EKF) training algorithm is proposed. The use of EKF training allows online training with faster learning and convergence speeds than backpropagation training method. Moreover, the propose adaptive PID approach includes a back-calculation anti-windup scheme to deal with windup effects, which is a common problem in PID controllers. The performance of the proposed approach is shown by presenting both simulation and experimental tests, giving results that are comparable to similar and more complex implementations. Tests are performed for a four wheeled omnidirectional mobile robot. Tests show the superiority of the proposed adaptive PID controller over the conventional PID and other adaptive neural PID approaches. Experimental tests are performed on a KUKA&reg, Youbot&reg, omnidirectional platform.

Emerging technologies seek to provide effective solutions to the most severe health problems such as type 1 diabetes mellitus (T1DM). In fact, the number of diabetics around the world has increased as well as the mortality rate associated with this condition. T1DM is caused by an autoimmune failure which disables the pancreas to produce insulin; therefore, glucose is not correctly metabolized to be used as efficient energy. Consequently, the most important fact is to keep the patient's blood glucose level within normal ranges in order to avoid long-term complications. Recently, engineering innovative approaches based on intelligent systems such as artificial neural networks have been proposed for control in biomedical systems. In this work, a novel neuro-fuzzy control scheme for blood glucose regulation in virtual T1DM patients is proposed. The glucose-insulin dynamics is modeled by a recurrent high-order neural network and then a neural multistep predictor is incorporated in order to know the glucose behavior within a 15-min horizon; thereby, allowing the knowledge of future values to determine the convenient basal infusion insulin rate as defined by the fuzzy membership functions. Test using the well-known Uva/Padova simulator illustrated that the proposed neuro-fuzzy controller maintains normoglycemia in virtual populations of adults, adolescents, and children digressing from two other neuro control approaches. Thus, intelligent systems based on neural networks offer enormous potential for health improvement of T1DM patients. The present contribution illustrates very encouraging results to closed-loop glucose level regulation regarding the autonomous artificial pancreas.

The study presents the hybridization of global sensitivity analysis with data-driven techniques to evaluate the Mexican electricity market interaction and assess the impact of individual parameters concerning locational marginal prices. The study case pertains to Yucatan, Mexico's electricity grid and market characteristics. A comparison of three artificial intelligence techniques in the electricity market is presented to forecast electricity prices in real-time market conditions. The study contemplates exogenous input parameters classified as regional, operational, meteorological, and economic indicators. A sensitivity analysis was carried out to the model with the best performance of the Artificial Intelligence techniques. The results showed that the impact of the variables fluctuates according to market and consumption conditions. In this study, the most relevant variables were electricity generation (17.06%), fossil fuel costs (natural gas 12.54% and diesel 8.63%), load zone (11.17%), and the day of the year (8.51%). From the qualitative point of view, the complex behavior of the parameters was analyzed; moreover, the quantitative results weighted the relevance of the variables in the Locational Marginal Prices. The meteorological and economic parameters allow assessing the environment where it interacts and serves as an instrument for decision-making in the planning of the energy sector. The presented methodology can be implemented as an alternative tool for market participants to analyze electricity prices.

Impulsive systems describe processes with at least one state variable is impulsively changeable. The design of optimal control policies in impulsive systems is a complex task. In order to relax the solution for the Hamilton-Jacobi-Bellman equation, a meaningful cost functional can be proposed a posteriori in the inverse optimal problem. The main contribution of this paper is a neural inverse optimal control for discrete-time impulsive systems. Control policies for discrete-time impulsive systems are derived by combining inverse optimal control into a recurrent high order neural network (RHONN) trained with the Extended Kalman filter (EKF). The neural network avoids the development of a mathematical model to represent the studied system. For illustration, we apply the proposed neurocontrol to personalized drug treatment in influenza infection disease, whose nonlinear model is included and described for completeness. The robustness of the proposed framework is tested through Monte Carlo simulations.

The solution of the inverse kinematics problem is an essential capability for robotic manipulators. This capability is used to solve tasks such as path planning, control of manipulators, object grasping, etc. In this paper, we present an approach for solving the inverse kinematics of robot arm manipulators using a soft computing approach. Given a desired end effector pose, the proposed approach is able to solve both the position and orientation for the inverse kinematic problem. In addition, the proposed approach avoids singularities configurations, since, it is based on the forward kinematics equations. We present simulations and experiments, where a comparative study among some selected soft computing algorithms is realized. The simulations and experiments illustrate the effectiveness of the proposed approach.

Wavelets are designed to have compact support in both time and frequency, giving them the ability to represent a signal in the two-dimensional time–frequency plane. The Gaussian, the Mexican hat, and the Morlet wavelets are crude wavelets that can be used only in continuous decomposition. The Morlet wavelet is complex-valued and suitable for feature extraction using continuous wavelet transform. Continuous wavelets are favoured when a high temporal resolution is required at all scales. In this paper, considering the properties from the Morlet wavelet and based on the structure of a recurrent high-order neural network model, a novel wavelet neural network structure, here called recurrent wavelet first-order neural network, is proposed in order to achieve a better identification of the behavior of dynamic systems. The effectiveness of our proposal is explored through the design of a centralized neural integrator backstepping control scheme for a two degree-of-freedom robot manipulator evolving in the vertical plane. The performance of the overall neural identification and control scheme is verified through numerical simulation using the mathematical model for a benchmark prototype. Moreover, real-time results validate the effectiveness of our proposal when using a robotic arm, of our own design, powered by industrial servomotors.

Structure from Motion algorithms offer good advantages, such as extract 3D information in monocular systems and structures estimation as shown in Hartley & Zisserman for numerous applications, for instance; augmented reality, autonomous navigation, motion capture, remote sensing and object recognition among others. Nevertheless, this algorithm suffers some weaknesses in precision. In the present work, we extent the proposal in Arana-Daniel, Villasenor, Lopez-Franco, & Alanis that presents a new strategy using bio-inspired intelligence algorithm and Conformal Geometric Algebra, based in the object mapping paradigm, to overcome the accuracy problem in two-view Structure form motion algorithms. For this instance, we include two new experiments and the inclusion of the circle entity; the circle carries stronger information about its motion than other geometric entities, as we will show.

Quorum Sensing (QS) is a complex process of cell to cell communication that allows bacteria to share information and regulate gene expression. This bacterial process is difficult to model because of their complexity, nonlinearity, and stochastic nature. While nonlinear systems can be derived using mathematical modeling, neural networks are an excellent tool to infer their dynamics. In this paper, we present a Recurrent High Order Neural Network (RHONN) to identify the communication system used by Escherichia coli (E. coli) to regulate its genetic expression. Simulation results show the applicability of the identifier.

In this work, a germinal center optimization (GCO) algorithm which implements temporal leadership through modeling a non-uniform competitive-based distribution for particle selection is used to find an optimal set of parameters for a recurrent high order neural network observer (RHONNO). The RHONNO is trained with an extended Kalman filter algorithm and it is capable of giving a model of the system besides of just giving state estimation. Furthermore, the RHONNO does not need previous knowledge of the system model, nor measurements, estimation or bounds of delays and disturbances. Applicability of the proposed methodology is presented using simulation results.

This paper presents the results of the use of training algorithms for recurrent neural networks based on the extended Kalman filter and its use in electric energy price prediction, for both cases: one-step ahead and n-step ahead. In addition, it is included the stability proof using the well-known Lyapunov methodology, for the proposed artificial neural network trained with an algorithm based on the extended Kalman filter. Finally, the applicability of the proposed prediction scheme is shown by mean of the one-step ahead and n-step ahead prediction using data from the European power system.

A nested high order sliding mode controller is proposed for speed control of a brushless direct current motor using an extended park transformation. Conventional park transformation applied to non sinusoidal back electromotive force motors results in non constant values in the (d, q) reference frame, a modified park transformation is used to obtain constant variables after transformation. Sliding mode control is robust in the presence of matched parametric variations and uncertain disturbances but unmatched variations have a negative effect in the system’s performance. A nested high order sliding mode scheme is presented to improve controller robustness to unmatched perturbations. Simulations are used to show the performance of the proposed controller with load torque and resistance variations.

This work presents a neural identifier-control scheme for uncertain nonlinear discrete-time systems with unknown time-delays. This scheme is based on a neural identifier to get a model of the system and a discrete-time block control technique based on sliding modes to generate the control law. The neural identifier is based on a Recurrent High Order Neural Network (RHONN) trained with an Extended Kalman Filter (EKF) based algorithm. Applicability is shown using real-time test results for linear induction motors. Also, a Lyapunov analysis is added in order to prove the semi-globally uniformly ultimately boundedness (SGUUB) of the proposed neural identifier-control scheme.

In this paper the authors present an approach to track a ground vehicle with a quadrotor equipped with a monocular vision system. The proposed approach is based on a visual servoing technique, to estimate the quadrotor desired velocities. The paper includes simulation and experimental results to show the effectiveness of the algorithm

Pose estimation is one of the most important tasks in mobile robotics. The problem consist in estimate the position of the mobile robot using a sensor on the robot. In this work we assume that a vision sensor is mounted on the robot, with the visual information provided by this sensor we estimate the motion of the mobile platform. In the 3D space the problem can be seen as the estimation of the transformation between two consecutive frames. In this paper the authors propose a pose estimation scheme which is based on particle swarm optimization (PSO). The proposed algorithm is designed to estimate the pose in the 3D space, and to be robust to outliers.

Glucose–insulin models, testing glucose sensors and support systems for health care decisions play an important role in synthesis of glucose control algorithms. In this work we propose an online glucose–insulin identification using the Recurrent High Order Neural Network (RHONN). Then, the model obtained is used to predict n -steps forward of glucose levels, also by RHONN. The used data for identification is from a Type 1 Diabetes Mellitus (T1DM) patient, it was collected from the Continuous Monitoring Glucose System (CMGS) by MiniMed Inc ® and an insulin pump by Paradigm Real-time Insulin Pump ®. RHONN is trained online by Extended Kalman Filter (EKF). The results suggest that it is possible to make a prediction of up to 35 min in the future, which it would help to prevent risky events (hypoglycemia and hyperglycemia). Also shows that, it could be directly connected to a CGMS to help the patient improve the glucose control and even an automatic glucose control algorithm. The proposed Neural Network shows good performance compared to baseline methods in terms of evaluation criteria.

In this paper, we present an impulsive pinning control algorithm for discrete-time complex networks with different node dynamics, using a linear algebra approach and a neural network as an identifier, to synthesize a learning control law. The model of the complex network used in the analysis has unknown node self-dynamics, linear connections between nodes, where the impulsive dynamics add feedback control input only to the pinned nodes. The proposed controller consists of the linearization for the node dynamics and a reorder of the resulting quadratic Lyapunov function using the Rayleigh quotient. The learning part of the control is done with a discrete-time recurrent high order neural network used for identification of the pinned nodes, which is trained using an extended Kalman filter algorithm. A numerical simulation is included in order to illustrate the behavior of the system under the developed controller. For this simulation, a 20-node complex network with 5 different node dynamics is used. The node dynamics consists of discretized versions of well-known continuous chaotic attractors.

Seasonal and pandemic influenza A virus (IAV) infections are a cause of severe morbidity and mortality worldwide. In this work, we study the problem of influenza treatment from a control theory perspective. Combined techniques of impulsive control and inverse optimal control are applied to an IAV model. Numerical results show that control-based strategies could improve virological efficacy and at the same time may reduce the drug amount in comparison to current clinical recommendations of pandemic regimens. This work discusses the advantages of theoretical approaches to tackle influenza infections.

This work proposes a discrete-time non-linear neural observer based on a recurrent high order neural network in parallel model trained with an algorithm based on the extended Kalman filter for discrete-time multiple input multiple output non-linear systems with unknown dynamics and unknown time-delay. To prove the semi-globally uniformly ultimately boundedness of the proposed neural observer the stability analysis based on the Lyapunov approach is included. Applicability of the proposed observer is shown via simulation and experimental results.

Visualizing physical phenomena is a central tool for nowadays research. In particular, volumetric representations are a critical factor in the diagnosis of diseases and surgery planning. In the last years, rendering techniques have been essential for medical practice, but these approaches are suitable for representing non-rigid motion in tissue and internal organs. In the present chapter, we introduce a mapping algorithm capable of track non-rigid deformations on free-form objects. The proposed method uses k-means for partition algorithm and covariance ellipsoid, afterward the Germinal Center Optimization is used to adapt the ellipsoid parameters. We offer experimental results over the Stanford Repository and tumors.

Long Short-Term Memory (LSTM) is a type of recurrent neural network that has become important in machine learning research thanks to its high precision to solve problems such as speech recognition, handwriting recognition, natural text compression, sequential data processing among others. Although classic LSTM are powerful tools to solve such problems, their adaptation is far from showing a smooth behavior which represents a drawback to LSTM be used in applications such as real-time control of physical systems in which to fulfill restrictions of ranges of values of the control variables is important in order to preserve the physical integrity of the systems. In this paper we present a design of architecture of LSTM that overcomes the non-smooth adaptation problem by using a single forget gate for all the LSTM units and furthermore improves the accuracy of classic LSTMs in problems such as rebber grammar learning, time series forecasting and control of physical systems as it is shown in the experimental and comparison results.

p53 regulates the cellular response to genotoxic damage and prevents carcinogenic events. Theoretical and experimental studies state that the p53-Mdm2 network constitutes the core module of regulatory interactions activated by cellular stress induced by a variety of signaling pathways. In this paper, a strategy to control the p53-Mdm2 network regulated by p14ARF is developed, based on the pinning control technique, which consists into applying local feedback controllers to a small number of nodes (pinned ones) in the network. Pinned nodes are selected on the basis of their importance level in a topological hierarchy, their degree of connectivity within the network, and the biological role they perform. In this paper, two cases are considered. For the first case, the oscillatory pattern under gamma-radiation is recovered; afterward, as the second case, increased expression of p53 level is taken into account. For both cases, the control law is applied to p14ARF (pinned node based on a virtual leader methodology), and overexpressed Mdm2-mediated p53 degradation condition is considered as carcinogenic initial behavior. The approach in this paper uses a computational algorithm, which opens an alternative path to understand the cellular responses to stress, doing it possible to model and control the gene regulatory network dynamics in two different biological contexts. As the main result of the proposed control technique, the two mentioned desired behaviors are obtained.

State estimation of biological systems is a difficult task due to their complexity and stochasticity. In particular, bilinear and Michaelis-Menten terms are the base for many biological models such as in infectious diseases, cancer, diabetes, and many others. In this paper, mentioned non-linear terms are formulated into a polynomial form with state-dependent matrices driven by additive white Gaussian noises over linear observations. To show the effectiveness of the approach, two different models widely used for modeling viral infectious diseases are considered and compared with the extended Kalman filter (EKF) algorithm. Numerical results show the applicability of the polynomial approach.

The effect of meal on blood glucose concentration is a key issue in diabetes mellitus because its estimation could be very useful in therapy decisions. In the case of type 1 diabetes mellitus (T1DM), the therapy based on automatic insulin delivery requires a closed-loop control system to maintain euglycaemia even in the postprandial state. Thus, the mathematical modelling of glucose metabolism is relevant to predict the metabolic state of a patient. Moreover, the eating habits are characteristic of each person, so it is of interest that the mathematical models of meal intake allow to personalise the glycaemic state of the patient using therapy historical data, that is, daily measurements of glucose and records of carbohydrate intake and insulin supply. Thus, here, a model of glucose metabolism that includes the effects of meal is analysed in order to establish criteria for data-based personalisation. The analysis includes the sensitivity and identifiability of the parameters, and the parameter estimation problem was resolved via two algorithms: particle swarm optimisation and evonorm. The results show that the mathematical model can be a useful tool to estimate the glycaemic status of a patient and personalise it according to her/his historical data.

Cooperative manipulation in dual-arm robotic systems is a fundamental capability to perform many problems such as human-like tasks. In this article, we present an approach to solve dual-arm cooperative manipulation tasks using the differential evolution algorithm. In this work, manipulator kinematics are represented using the Denavit–Hartenberg model. The proposed method is able to avoid singularities because it does not require the inversion of any Jacobian matrix. In addition, the proposed approach handles joint limit constraints based on penalty functions. As a final remark, this approach is suitable for robotic systems of redundant or non-redundant serial manipulators composed of revolute or prismatic joints. Simulation experiments illustrate the effectiveness of the proposed approach under different dual-arm configurations. Furthermore, real experiments were performed using a dual-arm KUKA Youbot system to show the applicability of the proposed approach.