Search

Wheat rust is one of the important factors leading to wheat yield decline. The traditional method of artificial identification of wheat rust remains inefficient. With the development of unmanned technology, unmanned aerial vehicles (UAVs) with advanced deep vision models can be used to monitor wheat growth, hence enabling timely disease detection and deployment of corresponding treatment measures. However, due to the limitation of embedded system hardware, it is a challenge to deploy a high‐accuracy and lightweight wheat rust detection model in an embedded system. In this paper, a training method based on transfer learning and sharpness‐aware minimization (SAM) is proposed to improve the accuracy of lightweight models. Specifically, the initial model is pretrained on an ImageNet dataset, and then the classifier of the model is fine‐tuned on the wheat rust training set; to prevent the risk of overfitting, the SAM method is applied to adjust the global parameters of the model. The experimental results on the Yellow‐Rust‐19 dataset show that the training method can effectively improve the accuracy of the wheat rust detection model, which is 3.56% higher than the existing methods. In addition, the parameter scale and computing resource requirements of four lightweight models are compared. The results indicate that MobileNetV3‐Small can achieve satisfactory accuracy with low requirements for storage and computing resource, and a detection frame rate of 101.36 frames per second in the Raspberry Pi, which is suitable for the unmanned detection of wheat rust. [ABSTRACT FROM AUTHOR]

Dust deposition on the surface of photovoltaic (PV) modules is a nonnegligible factor that reduces a PV system’s efficiency and reliability. Cleaning can remove dust, and the effect of cleaning on PV performance resembles that of maintenance. In this article, we propose a hybrid cleaning scheduling policy with periodic planning and dynamic adjustment for refining the operations and maintenance of PV systems. Specifically, the periodic planning stage aims for medium-term scheduling while the dynamic adjustment stage is tailed for short-term fine-tuning. In the former stage, we show that when the number of cleaning actions is fixed, a periodic cleaning strategy is optimal. Moreover, we derive the optimality condition under which the optimal cleaning interval can be determined. In the latter stage, based on the determined cleaning interval, we dynamically adjust the cleaning schedule with the forecast of meteorological parameters, PV power generation, and dust deposition in order to further minimize economic losses. In addition, we take the forecasting uncertainty into account and propose a new custom parameter called risk-taking tendency (RTT), which is able to quantify the risk preference of decision makers and analyze its influence on the scheduling policy. A case study is provided to illustrate the proposed strategy. [ABSTRACT FROM AUTHOR]

In this study, an MnO 2 ion sieve blend membrane with high hydrophilicity was developed to strength the lithium extraction from brines instead of the traditional adsorbent granulation procedure. The membrane-type adsorbent has the advantages of high stability and high-water permeability. That do not require high pressure because adsorption modules can be easily constructed by stacking or winding membranes. Therefore, membrane-type adsorbents are suitable for large-scale applications. It was confirmed that the adsorption rate and hydrophilicity of the membrane were negatively related to the increasing PVC content, while the tensile and compressive strengths could be maximized with a blending ratio of Polyvinyl chloride (PVC) and Polyacrylonitrile (PAN) (10:1). Besides, the optimized membrane with spongy porosities demonstrated a large flux and great adsorption performance for Li+ in the Qarham Salt Lake old brine. The lithium adsorption capacity could keep at 2800 mg m−2 and the desorption rate always exceeded 85% in the multicycle adsorption-desorption experiments，which indicates the long-term stability of this membrane-type adsorbent. In addition, this work contributes to the field by deepening the understanding of preparation conditions of blend membrane from multiple perspectives. Furthermore, a strategy is proposed for designing a blending modification membrane structure using the incorporation of two or more materials, which offers potential for future advancements in membrane technology. [Display omitted] [ABSTRACT FROM AUTHOR]

Predicting the remaining useful life for operational devices plays a critical role in prognostics and health management. As the models based on the stochastic processes are widely used for characterizing the degradation trajectory, an adaptive skew-Wiener model, which is much more flexible than traditional stochastic process models, is proposed to model the degradation drift of industrial devices. To make full use of the prior knowledge and the historical information, an on-line filtering algorithm is proposed for state estimation, a two-stage algorithm is adopted to estimate unknown parameters as well. For remaining useful life prediction, a novel result is presented with an explicit form based on the closed skew normal distribution. Finally, sufficient Monte Carlo simulations and an application for ball bearings in rotating electrical machines are used to validate our approach. [ABSTRACT FROM AUTHOR]

This paper focuses on the safety of railway vehicles. A new semi-active control strategy is proposed based on the skyhook control theory. In view of the main railway vehicle safety performance indicators, the new control strategy aims at reducing the derailment coefficient of railway vehicles by restraining the lateral vibrations of the bogie and the wheelset. Furthermore, to evaluate the improvement of the safety performance brought about by the new control strategy, a complete railway vehicle model is established using the ADAMS/Rail software package. In further co-simulations, five conventional control methods are compared with the proposed approach under the same conditions. Co-simulation results indicate that the new control strategy is effective in improving the safety performance of railway vehicles. [ABSTRACT FROM AUTHOR]

Nonlinear degradation trajectories are encountered frequently, and not all of them evolve homogeneously in practical systems. To take nonlinearity, heterogeneity, and the entire historical degradation data into account, we propose a nonlinear heterogeneous Wiener process model with an adaptive drift to characterize degradation trajectories. A state-space based method is employed to delineate our model. Due to the introduction of the adaptive drift, it is difficult to directly apply Kalman filter methods to update the distribution of the estimated degradation drift. To address this issue, we develop an online filtering algorithm based on Bayes’ theorem. The expectation-maximization (EM) algorithm, as well as a novel Bayes’-theorem-based smoother, are adopted to estimate the unknown parameters in our model. Moreover, the distribution of the predicted remaining useful life (RUL) incorporating the complete distribution of the estimated degradation drift is achieved analytically. Finally, a simulation, and a case study are provided to validate the proposed approach. [ABSTRACT FROM AUTHOR]

A dynamic preventive maintenance policy for system with continuously degrading components is investigated in this paper. Different from traditional cost-centric preventive maintenance policy, our maintenance strategy is formulated from the value perspective. Component value is modelled as a function of component reliability distribution. Maintenance action is triggered whenever the system reliability drops below a certain threshold. Our policy mainly consists of two steps: (i) determine which component to maintain; (ii) determine to what degree the component should be maintained. In Step 1, we introduce the yield-cost importance to select the most important component. In Step 2, the optimal maintenance level is obtained by maximizing the net value of the maintenance action. Finally, numerical examples are given to illustrate the proposed policy. [ABSTRACT FROM AUTHOR]

Background and objectives: The most widely used method to compare dissolution profiles is the similarity factor f 2. When this method is not applicable, the confidence interval of f 2 using bootstrap methodology has been recommended instead. As neither details of the estimator nor the types of confidence intervals are described in the guidelines, the suitability of five estimators and fourteen types of confidence intervals were investigated in this study by simulation. Methods: One million individual dissolution profiles were simulated for the reference and test populations with predefined target population f 2 values, where random samples of different sizes were drawn without replacement. From each pair of random samples, five f 2 estimators were calculated, and fourteen types of confidence intervals were obtained using 5000 bootstrap samples. The whole process was repeated 10000 times and the percentage of the similarity conclusions was measured. In addition, the uncertainty associated with the current practice of using f ^ 2 point estimate alone for the statistical inference was evaluated. Results: When combined with different types of confidence intervals, the estimated f 2 (f ^ 2), the bias-corrected f 2 (f ^ 2 , bc ), and the variance- and bias-corrected f 2 (f ^ 2 , vcbc ) are not suitable estimators due to higher-than-acceptable type I errors. The estimator f ^ 2 , exp , calculated based on the mathematical expectation of f ^ 2 , and f ^ 2 , vcexp , the variance-corrected f ^ 2 , exp , showed acceptable type I errors when combined with any of the ten percentile intervals. However, they have the drawback of low power, which might be addressed by increasing the sample size. To properly control the type I error, samples with at least 12 units should be used. Conclusion: The best combinations of estimator and type of confidence interval are f ^ 2 , exp and f ^ 2 , vcexp combined with any of the ten types of percentile intervals. When the sample f 2 value is close to 50, the use of the confidence interval of f 2 is recommended even when the variability of the dissolution profiles is low and the prerequisites defined in the regulatory guidelines for using the conventional f 2 method are fulfilled in order to control the type I error rate. [ABSTRACT FROM AUTHOR]

Abstract: Over the past few years, fault detection for computer networks has attracted extensive attentions for its importance in network management. Most existing fault detection methods are based on active probing techniques which can detect the occurrence of faults fast and precisely. But these methods suffer from the limitation of traffic overhead, especially in large scale networks. To relieve traffic overhead induced by active probing based methods, a new fault detection method, whose key is to divide the detection process into multiple stages, is proposed in this paper. During each stage, only a small region of the network is detected by using a small set of probes. Meanwhile, it also ensures that the entire network can be covered after multiple detection stages. This method can guarantee that the traffic used by probes during each detection stage is small sufficiently so that the network can operate without severe disturbance from probes. Several simulation results verify the effectiveness of the proposed method. [Copyright &y& Elsevier]

Abstract: A scheme for implementing secure communication based on chaotic maps and strong tracking filter (STF) is presented, and a modified STF algorithm with message estimation is developed for the special requirement of chaotic secure communication. At the emitter, the message symbol is modulated by chaotic mapping and is output through a nonlinear function. At the receiver, the driving signal is received and the message symbol is recovered dynamically by the STF with estimation of message symbol. Simulation results of Holmes map demonstrate that when message symbols are binary codes, STF can effectively recover the codes of the message from the noisy chaotic signals. Compared with the extended Kalman filter (EKF), STF has a lower bit error rate. [Copyright &y& Elsevier]

Intelligent rice disease recognition methods based on deep neural networks can predict the degree of disease on the basis of, for example, the number of disease spots on an image, so that preventive measures can be taken. Currently, intelligent recognition methods for rice diseases suffer from the disadvantages of poor versatility and low accuracy. This paper uses eight common image classification networks to classify and identify four rice diseases. ResNet50 was selected as the feature extraction network and an enhanced feature fusion and target adaptive network (EFFTAN), referred to as EFFTAN, is proposed. The EFFTAN was used to detect four rice spot diseases in the rice leaf disease image samples dataset; the mean average precision of the final detection was 95.3%, and effective detection was also achieved for the dense spot features. [ABSTRACT FROM AUTHOR]

Inter- and intra-batch variability of the quality attributes contribute to the uncertainty for demonstrating equivalent microstructure of post-approval changes and generic/hybrids of semisolid topical products. Selecting a representative sample size to describe accurately the in vitro properties of semisolids and to reach enough statistical power to demonstrate similarity between two semisolid topical products is currently challenging. The objective of this work is to establish the number of batches and units per batch to be compared based on different inter-batch and intra-batch variability to demonstrate equivalence in the physical characteristics of the products that ensure a similar microstructure of the semisolid. This investigation shows that the minimum number of batches to be compared of each product is 3 and the minimum number of units per batch could be 6 in the case of low intra- and inter-batch variability. If the products are not identical, i.e., 2.5–5% differences that are expected due to differences in the manufacturing process or the suppliers of excipients, 12 units and 6 batches are needed. If intra- or inter-batch variability is larger than 10%, the number of batches and/or the number of units needs to be increased. As the interplay between inter- and intra-batch variability is complex, the sample size required for each combination of inter- and intra-batch variability and expected difference between products can be obtained in the attached tables. [ABSTRACT FROM AUTHOR]

Bearing is a key component in rotary machines. Their failures may cause the abrupt shutdown of these machines, which would result in substantial economic losses. Therefore, the prediction of the remaining useful life (RUL) of bearings is regarded as one of the critical approaches to avoid failure of bearings and their systems. In this article, an ensemble data-driven approach is proposed to predict the RUL of bearings. It uses feature extraction, an attention mechanism, and uncertainty analysis. First, the features embedded in the bearings’ vibration signals are extracted. Second, a stacked gated recurrent unit (GRU) is constructed to predict the bearing RUL. A novel attention mechanism based on dynamic time warping (DTW) is developed to improve the performance of information extraction, and a Bayesian approach is employed to analyze the prediction uncertainty. Finally, the proposed approach is validated using two benchmark-bearing data sets. The results show that the proposed approach can predict the bearing RUL effectively, and the prediction uncertainty can also be evaluated. [ABSTRACT FROM AUTHOR]

Highlights • Multi-protocol label switching networks is optimized based on reliability and costs. • Decoupled the problem into two sub-problems: the cost and the reliability problems. • Working and backup paths are the first K+1 paths regardless of rank of back path. • An algorithm is proposed to reduce the searching space for working and backup paths. Abstract The multi-protocol label switching (MPLS) networks have been widely applied in the backbone network in recent years with offering better Quality-of -Service (QoS). The reliability of MPLS networks is of great importance. In this paper, we consider optimal design for MPLS networks, where working and backup paths are optimally selected to minimize networks' constructing costs under reliability constraints. As both the networks' constructing costs and the network reliability are related to the lengths of the selected paths, the optimal network design problem is decoupled into two sub-problems: the cost related and the reliability related problems, where the cost related problem has been solved by Sohn et al.. Considering the reliability related problem, we proposed a reliability related algorithm to optimally search for the working and backup paths by greatly reducing the searching space. Both the cost related and the reliability related problems can be solved using the same searching criteria, so the optimal set of the working and backup paths of the proposed problem can be derived by comparing the results obtained from the cost related and reliability related problems. Numerical examples are also provided to verify the reliability related algorithm and the whole multi-path transmission design approach. [ABSTRACT FROM AUTHOR]

Deep learning (DL)-based methods for remaining useful life (RUL) prediction have received increasing research attention due to excellent feature extraction abilities. Most DL methods rely on abundant labeled samples for supervised training. However, because of the adoption of the over-maintenance strategy of equipment, the monitored data for the degradation of equipment usually consists of few labeled samples and a large amount of unlabeled samples, which limits the performance of DL methods. To take advantage of the value of unlabeled samples, this paper proposed a contrastive learning framework for RUL prediction. First, an unlabeled sample augmentation is developed firstly to extend the sample set. Then, an unlabeled sample learning (USL) architecture is proposed to learn the information of degradation from unlabeled samples to promote general DL models' performance on RUL prediction. Based on the proposed framework, USL-convolutional neural network and USL-long short-term memory network are used to validate its performance based on datasets of turbofan engine and bearing. Results show that the performance of RUL prediction based on the proposed framework can be enhanced by unlabeled samples and verify the good scalability and generalization ability of the proposed framework. • Proposed framework boosts RUL prediction with unlabeled samples despite few labeled. • Proposed framework is designed for general deep learning methods like CNN and RNN. • Two case analyses validated the effectiveness and generalization ability. [ABSTRACT FROM AUTHOR]

• Introduce a novel formulation that integrates technical and organizational resilience. • Propose a comprehensive resilience analysis framework incorporating various metrics. • Utilize a new multi-objective linear programming model for optimizing resource allocation. • Present a detailed flowchart for validating the resilience analysis framework in a power system. In urban infrastructure systems, resilience is crucial for maintaining functionality, minimizing losses, and expediting recovery during disruptive incidents. Effective allocating resources across various phases of emerging disturbances can generally enhance the system's capacity to cope with disasters. However, it is imperative to recognize that distinct resource allocation strategies may lead to divergent outcomes in terms of resilience performance. Therefore, this study develops a framework for optimizing resource allocation based on multiple resilience objectives by understanding the interplay between resilience performance and dynamic decision-making. The resilience processes are first formalized into distinct stages, considering the technical and organizational resilience of the infrastructure system in the event of disruption. Building upon this foundation, five decision scenarios are proposed, contingent on the allocation or non-allocation of resources to each resilience stage. A multi-resilience-objective mixed-integer linear programming (MROMILP) model is formulated to optimize the resource allocation scheme for each resilience stage within the constraints of internal resources. Finally, the model and framework are tested using a power system as a tangible example. The integrated multi-stage quantitative resilience assessment and optimization method proposed in this study can assist decision-makers in making dynamic and continuous trade-offs between resources and resilience targets. [ABSTRACT FROM AUTHOR]

Data-driven methods routinely achieve promising results on remaining useful life prediction, but under a window-manner end-to-end paradigm, they suffer from unsatisfying generalization ability and low interpretability, as the consequence of neglecting diverse modes among the entire degradation processes of different entities. This article proposes a novel Transformer-based network, to tackle the problem by integration of global and local information. During offline training, the paired inputs containing full life and piece data are constructed, and then using cross-attention between the encoder and the decoder, the consistent position of the piece data in the full life is derived, which is directly associated with the degradation state. The designed paired inputs and model architecture ensures the strong generalization because the prediction result considering global information is adaptive to diverse degradation modes. Further, the designed cross-attention discrepancy utilizes prior knowledge of the consistent position such that similar degradation states are aligned more properly. Such a consistent position, visualized by the cross-attention distribution, is supposed to represent the intuitive relationship between degradation level and monitoring data, thus provides inherent interpretability about the prediction process. Finally, predictions of the online monitoring piece data with respect to all historical full lives with different degradation modes are aggregated to the final prediction. Extensive experiments on two datasets of turbofan and bearing show that our model provides competitive performance, especially under complicated working conditions and fault modes, achieving averagely 5.9% score reduction compared with the state-of-the-art method. • The union consideration of global and local information ensures adaptation to diverse entities. • The paired inputs of full life and piece data derive interpretable consistent positions. • Two case studies on typical industrial applications validate the effectiveness. [ABSTRACT FROM AUTHOR]

Background: Doxylamine succinate, an ethanolamine-based antihistamine, is used in the short-term management of insomnia because of its sedative effects. No data on the dose proportionality of the pharmacokinetics of doxylamine are available, although this drug has been marketed in European countries for more than 50 years. Objective: The objective of this study was to evaluate and compare the dose proportionality between two marketed strengths (12.5 mg and 25 mg) of doxylamine hydrogen succinate after a single oral dose administration under fasting conditions in healthy human subjects. Study Design: This was a single-center, randomized, single dose, laboratory-blinded, two-period, two-sequence, crossover study. Setting: The study was conducted in a phase I clinical unit. Subjects and Methods: A single oral dose of doxylamine hydrogen succinate of 12.5 mg (equivalent to 8.7 mg of doxylamine base) or 25 mg (equivalent to 17.4 mg of doxylamine base) was administered to healthy volunteers under fasting conditions in each study period. The drug administrations were separated by a wash-out period of 7 calendar days. Blood samples were collected for up to 60 h post-dose, and plasma doxylamine levels were determined by an ultra high-performance liquid chromatography method with tandem mass spectrometry detection. Pharmacokinetic parameters were calculated using non-compartmental analysis. Dose proportionality was assessed based on the parameter area under the concentration-time curve (AUC). Safety was evaluated through assessment of adverse events, standard laboratory evaluations, vital signs and 12-lead electrocardiogram (ECG). Results: In total, 12 healthy volunteers (3 male; 9 female) were included in the study. Mean maximum observed plasma concentration ( C) and area under the concentration-time curve from time zero to time t (AUC) of doxylamine hydrogen succinate 12.5 mg and 25 mg tablets increased linearly and dose-dependently [12.5 mg: mean C 61.94 ng/mL, coefficient of variation (CV) 23.2 %; mean AUC 817.33 ng·h/mL, CV 27.4 %; and 25 mg: mean C 124.91 ng/mL, CV 18.7 %; mean AUC 1630.85 ng·h/mL, CV 22.8 %]. Mean AUC was 815.43 ng·h/mL, CV 22.8 % for 25 mg. The dose-normalized geometric mean ratio (%, 12.5 mg/25 mg) of AUC was 98.92 (90 % CI: 92.46, 105.83). The most common adverse event was somnolence. Conclusions: Exposure to doxylamine was proportional over the therapeutic dose range of 12.5-25 mg in healthy volunteers. Based on the results, a predictable and linear increase in systemic exposure can be expected. Doxylamine hydrogen succinate was safe and well tolerated. [ABSTRACT FROM AUTHOR]

Background: Doxylamine succinate, an ethanolamine-based antihistamine, is used in the short-term management of insomnia because of its sedative effects. The data available on the pharmacokinetic profile of doxylamine in humans are limited, notwithstanding that this drug has been marketed in European countries for more than 50 years. In fact, no data on the effect of food on the pharmacokinetic parameters of doxylamine are available. Objective: The objective of this study was to evaluate the pharmacokinetic parameters of doxylamine following a single oral dose of doxylamine hydrogen succinate 25 mg in healthy human subjects under fed and fasting conditions. Study Design: This was a single-center, randomized, single-dose, laboratory-blinded, two-period, two-sequence, crossover study. Setting: The study was conducted in a phase I clinical unit. Subjects and Methods: A single oral dose of doxylamine hydrogen succinate 25 mg (equivalent to 17.4 mg of doxylamine base) was administered to healthy volunteers under either fed conditions (high-fat, high-calorie food intake) or fasting conditions in each study period. The drug administrations were separated by a wash-out period of seven calendar days. Plasma samples were collected for up to 60 hours postdose, and plasma doxylamine concentrations were determined by a high-performance liquid chromatography method with tandem mass spectrometry detection. Pharmacokinetic parameters were calculated using noncompartmental analysis. Safety was evaluated through assessment of adverse events, standard laboratory evaluations, vital signs, and 12-lead electrocardiography. Results: In total, 24 healthy subjects (12 male and 12 female) were included in the study. Doxylamine succinate 25 mg tablets exhibited similar oral bioavailability of doxylamine in the fasting state (mean maximum plasma drug concentration [C] 118.21 ng/mL, coefficient of variation [CV] 19.2%; mean area under the plasma concentration time curve from time zero to time t [AUC] 1746.97 ng · h/mL, CV 31.6%) and in the fed state (mean C 120.99 ng/mL, CV 15.0%; mean AUC 1712.20 ng · h/mL, CV 26.7%). No statistically significant between-treatment differences were observed for any of the pharmacokinetic parameters under study. The fed: fasting ratios of the geometric least squares means with corresponding 90% confidence intervals for C and AUC were within the range of 80-125%. Conclusion: High-fat, high-calorie food intake does not affect the kinetics of doxylamine in healthy subjects. The drug was safe and well tolerated by the subjects in this study. [ABSTRACT FROM AUTHOR]

The coal pulverizing system is an important auxiliary system in thermal power generation systems. The working condition of a coal pulverizing system may directly affect the safety and economy of power generation. Prognostics and health management is an effective approach to ensure the reliability of coal pulverizing systems. As the coal pulverizing system is a typical dynamic and nonlinear high-dimensional system, it is difficult to construct accurate mathematical models used for anomaly detection. In this paper, a novel data-driven integrated framework for anomaly detection of the coal pulverizing system is proposed. A neural network model based on gated recurrent unit (GRU) networks, a type of recurrent neural network (RNN), is constructed to describe the temporal characteristics of high-dimensional data and predict the system condition value. Then, aiming at the prediction error, a novel unsupervised clustering algorithm for anomaly detection is proposed. The proposed framework is validated by a real case study from an industrial coal pulverizing system. The results show that the proposed framework can detect the anomaly successfully. [ABSTRACT FROM AUTHOR]