As an important component of clean energy, wind power generation is of great significance in promoting energy structure transformation and achieving sustainable development. It is particularly crucial to develop precise and effective fault diagnosis technology to ensure the safe and stable operation of wind turbines. This paper proposes a fault diagnosis method for wind power transmission systems based on model and data fusion, called PINN_Transformer. The method integrates the nonlinear dynamics model of the planetary gearbox as physical prior knowledge into a Transformer-based diagnostic network. To achieve this fusion, a novel physics-informed loss function is constructed, which maps the system’s vibration differential equations to the time-frequency domain via wavelet transform, and simultaneously incorporates constraints based on energy conservation and time-frequency feature matching. This approach embeds the governing physical laws directly into the learning process of the deep network, addressing the interpretability gap in purely data-driven methods while enhancing feature discrimination. Experimental results demonstrate significant advantages of PINN_Transformer compared to other advanced diagnostic methods, achieving a fault diagnosis accuracy of 99.84%. Furthermore, the model maintains robust performance with an accuracy above 97.07% under additive noise conditions, confirming its good engineering application value.

Large vision-language models (LVLMs) have demonstrated remarkable capabilities in autonomous driving scene understanding tasks. However, these models occasionally generate hallucinatory texts, resulting in descriptions that seem reasonable but do not correspond to the image. To address this challenge, this paper proposes HELTA, a training-free data annotation method used in traffic scenarios, which is designed to support the offline generation of high-quality semantic datasets without hallucinations. Specifically, HELTA employs a cross-checking mechanism to filter entities and directly extracts critical objects from the given image, enriching the descriptive text. Experimental results on the POPE benchmark demonstrate that HELTA improves the F1-score of the Mini-InternVL-4B and mPLUG-Owl3 models by 12.58% and 4.28%, respectively. Additionally, qualitative results using images collected in open campus scene further highlight the practical applicability of the proposed method. Compared with the GPT-4o model, HELTA achieves comparable descriptive performance while significantly reducing costs. Finally, two high-quality semantic understanding datasets, CODA_desc and nuScenes_desc, are created for traffic scenarios to support future research. The codes and datasets are publicly available athttps://github.com/fjq-tongji/HELTA

The meteorological communication networks provide critical data support for agriculture and environmental monitoring. However, covert gradient-based attacks persistently inject subtle perturbations, threatening data integrity and increasing the operational overhead for network operators. To achieve proactive service assurance and security-aware network management, this paper proposes a data integrity monitoring mechanism as a managed network function, named EK-IGNN. Unlike traditional passive detection, EK-IGNN functions as an active security service. It first employs the Empirical Mode Decomposition Kalman Filter (EMD-KF) to extract high-fidelity attack fingerprints, which are then analyzed by an Intrinsic Graph Neural Network (IGNN). The IGNN model captures complex dependencies and adaptively amplifies weak attack features, enabling closed-loop network security management. Experimental results demonstrate that the proposed algorithm achieving an average improvement of 16.07% in accuracy and 15.27% in F1-score over state-of-the-art benchmarks.

This article revisits the federated policy gradient algorithm with environment heterogeneity for finding the optimal policy in multiagent reinforcement learning (RL). Toward this direction, personalized federated RL methods have been presented recently. However, existing personalized federated policy gradient methods may confine the personalized capacity of local policy models. In order to tackle this challenge, this article develops a provably convergent personalized federated first-order policy gradient algorithm, referred to as PfoPG, which learns a personalized policy model by adaptively mixing optimal global and local policies. Moreover, the momentum-based importance sampling is also introduced into PfoPG to improve its convergence speed. Meanwhile, this article rigorously analyzes the nonasymptotic convergence behavior of PfoPG. More specifically, PfoPG converges to a stationary policy with rateO(1/K), whereKdenotes the number of iterations. Compared to the state-of-the-art federated policy gradient methods, PfoPG can improve the convergence rate fromO(1/K2/3) toO(1/K). Finally, we verify the effectiveness of PfoPG by various experiments based on the multiagent particle environment.

Due to complex environmental factors, the position measurement data of mobile robots are susceptible to pollution from multiplicative noise and non-Gaussian characteristics, which poses a challenge for the existing adaptive filtering methods to obtain high-performance pose estimation. Based on the comprehensive consideration of non-Gaussian multiplicative noise and correlated noise, a non-Gaussian adaptive Cubature Kalman Filter design method based on Gaussian sum and maximum correntropy techniques is proposed. Firstly, an enhanced Gaussian-sum and Cubature Kalman Filter method is designed to re-derive the innovation covariance and cross-variance for the multiplicative noise system. Secondly, under the framework of the maximum correntropy filter, the design strategy of the Gaussian kernel function weight matrix is improved, and a Cubature Kalman Filter method based on the improved maximum correntropy criterion is proposed. Then, based on consideration of the correlation between multiplicative noise and measurement noise, a Gaussian sum maximum correntropy Cubature Kalman Filter method with a known correlation coefficient is established. Finally, an adaptive Gaussian sum maximum correntropy Cubature Kalman Filter method that can dynamically estimate the correlation coefficient of multiplicative noise in real time is proposed, which can perform joint high-performance estimation of mobile robot’s pose and correlation coefficient.Note to Practitioners—This paper focuses on nonlinear multiplicative noise systems which are non-Gaussian. On one hand, non-Gaussian multiplication noise is pervasive in all types of systems, seriously affecting signal transmission and data acquisition. On the other hand, the correlation between multiplicative noise and measurement noise will seriously affect filter accuracy and create new challenges. Particularly, as it is difficult to obtain the correlation coefficient in the current system, the method to estimate the correlation coefficient between non-Gaussian noise is a challenging problem. To address these issues. Firstly, the Gaussian sum technique is used to approximate the non-Gaussian noise. Secondly, the correntropy technique is used to further reduce the influence of non-Gaussian noise on the filter accuracy. Then, the problem of correlated noises is solved by improving the cost function. Finally, the problem of inaccurate noise correlation coefficient in the real system is solved by the principle of covariance correspondence. This method is suitable for systems with correlated multiplicative noise. If the accuracy of sensor measurement data is too low, it will affect estimation performance. In future research, we will address the issue of reduced filtering accuracy in various correlated noise environments.

Object detection and tracking is always a challenging issue in UAV (unmanned aerial vehicle) application. Especially, in the scene of UAV-ASV (autonomous surface vehicle) cooperative system, UAV vision based target tracking performance has been suffering from the target rotation and fast motion. Aiming at optimizing the related UAV vision tracking performance, a SDSST tracker(strong discriminative scale space tracking) with automatic initialization and self-adjusting is developed in this paper. Firstly, in the step of initialization, combining the advantages of both fast optimization of BLS (Broad Learning System) and efficient image processing of CNN (convolutional neural networks), a novel DBLS (Deep Broad Learning System) is posed for the target detection. Meanwhile, a Q-learning based DBLS architecture searching is further introduced. Then, in terms of width/height ratio self-adjusting, this article proposed a novel filter state supervisor that helps to find the abnormal estimated state caused by rotation in target scale estimation. Basically, this so called filter state supervisor could take RSV (Rolling Standard Value) as input feature and give out the filter state. Finally, the abnormal filter state would be adjusted by an appropriate alternative by searching in the proposed rotation angles memory, so that an optimized self-adjusting could be realized. Meanwhile, extensive experiments are performed on data set of USV center in Qiandao Lake, yielding a competitive result compared with five other prevalent trackers. Note to Practitioners—This paper was motivated by the problem of target lost caused by sudden change of target motion in the UAV-ASV vision tracking. Usually, the rotation motion of ASV is a major operation when ASV carries out a maritime assignment. However, the poor tracking state supervision and inflexible scale updating method as well as manual initialization in the existing approaches lead to inappropriate scale and target lost when the target undergoes rotational motion. Therefor, this paper proposes a strong vision tracker (SDSST) to enhance the original DSST in the three aspects: automatic initialization, filter state supervisor, and self-adjusting. This can allow the tracker to initialize without human interference, also bad tracker state can be informed by filter state supervisor. When bad state happens that the target scale in the tracker will be updated flexibly based on proposed rotation angles memory. Finally, The proposed method is implemented on the real filed UAV vision data collected by UAV-ASV system in the Qiandao Lake. The results show that SDSST achieves competitive result compared to 7 other prevalent trackers.

With the continuous development of Unmanned Aerial Vehicle (UAV) visual positioning technology, dynamic target detection and feature point optimization have become one of the difficult problems for UAVs to achieve high-precision visual positioning in a dynamic environment. To solve the problem of UAV target detection accuracy in a dynamic environment, this paper proposes a dynamic target detection method for airborne cameras based on background prediction and semantic compensation. Firstly, to solve the problem of high false detection rate of the traditional background difference method on the camera of moving carrier, this paper proposes a background compensation method based on region of interest prediction and uses a technique combining a scale-transformed Unscented Kalman filter (ST-UKF) and Rodrigues Formula with Perspective Transformation (RFPT) to predict the background model. Then, a moving target discrimination method based on semantic confidence is proposed to solve the problem that the traditional semantic map cannot effectively discriminate the current state of the object and leads to an excessive elimination of effective feature points; in addition, a general detection framework for airborne cameras to obtain accurate and reliable target selection boxes are proposed to improve the positioning accuracy of traditional visual positioning methods in dynamic environments, the feasibility, and innovation of the algorithm in this paper are verified through data set simulation and experimental environment.

This paper is concerned with the$H_{\infty }$proportional-integral-derivative (PID) control problem for Takagi-Sugeno fuzzy systems over lossy networks that are characterized by the Gilbert-Eillott model. The communication quality is reflected by the presence of two channel modes (i.e., “bad” mode and “good” mode), which switch randomly according to a Markov process. In the “bad” mode, packet dropouts are governed by a stochastic variable sequence. Considering the inaccessibility of channel modes, a mode detector is utilized to estimate the communication situation. The relationship between the actual channel mode and the estimated mode is depicted in terms of certain conditional probabilities. Moreover, a comprehensive model is constructed to represent the probability uncertainties arising from statistical errors in channel mode switching, packet dropouts, and mode detection processes. Subsequently, a robust asynchronous PID controller, based on the detected channel mode, is proposed. Sufficient conditions are then derived to ensure the mean-square stability of the closed-loop system while maintaining the desired$H_{\infty }$performance. Finally, the efficacy of the proposed design approach is demonstrated through a simulation example.

Decentralized actor-critic (AC) is one of the most dominant algorithms for dealing with multiagent reinforcement learning (MARL) problems. However, exploration-efficient, sample-efficient, and communication-efficient are difficult to achieve simultaneously by existing decentralized AC methods. For this reason, this article develops a decentralized multiagent AC algorithm by incorporating entropy regularization to improve exploration with theoretical guarantees, referred to as multi-agent AC algorithm with entropy regularization (MACE). Moreover, we rigorously prove that MACE can achieve sample complexity $\mathcal {O}(\epsilon ^{-2}\ln \epsilon ^{-1})$ and communication complexity of $\mathcal {O}(\epsilon ^{-1}\ln \epsilon ^{-1})$ , which match the best complexities at present. Finally, the performance of MACE is also evaluated on reinforcement learning (RL) tasks. The experimental results show that the proposed algorithm achieves better exploration efficiency than state-of-the-art decentralized AC-type algorithms.

At present, deep learning has demonstrated outstanding performance in the area of underwater image enhancement. However, these approaches often demand substantial computational resources and extended training time. Knowledge distillation is a widely used technique for model compression, and nowadays it has delivered outstanding results across various fields. However, it has not been utilized in the field of underwater image enhancement. To tackle the aforementioned issues, this paper introduces a knowledge distillation technique for underwater image enhancement for the first time. It is a semi-supervised self-inter feature distillation and unsupervised self-domain adversarial distillation approach. It specifically includes adaptive local self-feature distillation technique, information lossless multi-scale inter-feature distillation technique, and self-domain adversarial distillation approach in LAB-RGB space. Self-feature distillation enhances the performance of the student network by correcting other lossy feature maps with the maximum effective feature map. Inter-feature distillation enables the student network to maximize the potential information learned from the teacher network. Furthermore, an information loss-free pooling approach is suggested to achieve multi-scale loss-free information extraction. Self-domain adversarial distillation boosts the performance of student networks through unsupervised adaptive enhancement in LAB space and unsupervised domain adversarial distillation in RGB space. Finally, a self-inter alternate knowledge distillation training measure is proposed, aiming to maximize the respective benefits of self-inter knowledge distillation. Through extensive comparative experiments, it can be found that student networks with dissimilar structures trained using the knowledge distillation technique designed in this paper achieve outstanding underwater image enhancement results.

The cubature Kalman filter (CKF) overcomes the limitations of the Kalman filter in strong nonlinear systems, which has been widely used in many fields. However, in practical engineering, the abnormal measurement information obtained by the sensor causes the measurement noise covariance to change, which may deteriorate the filtering performance and even cause the filter failure. The fault-tolerant filter can deal with the state estimation problem for the systems with abnormal measurements. The key of the fault-tolerant filter is to forcefully correct filter innovation by using a fading factor. The fault-tolerant filter technology has been extensively applied in many practical systems, but it is still lack of reasonable theoretical analysis. To this end, the measurement noise model is established and the magnitude of the noise deviation is analyzed. The filtering performance under abnormal measurement is analyzed by three mean squared errors (MSEs), which are the ideal MSE, the filter calculated MSE and the true MSE. In order to solve the influence of sampling approximation deviation of CKF on fault detection, an improved fault detection algorithm is proposed. The performance of fault-tolerant CKF is analyzed from two views. The first view is about comparing the filter calculated MSEs of CKF and of fault-tolerant CKF, the second view is about comparing the relative closeness of the filter calculated MSE to the true MSE for the two algorithms. Numerical examples further verify these conclusions.

Accuracy and speed are the most important indexes for evaluating many object tracking algorithms. However, when constructing a deep fully convolutional neural network (CNN), the use of deep network feature tracking will cause tracking drift due to the effects of convolution padding, receptive field (RF), and overall network step size. The speed of the tracker will also decrease. This article proposes a fully convolutional siamese network object tracking algorithm that combines the attention mechanism with the feature pyramid network (FPN), and uses heterogeneous convolution kernels to reduce the amount of calculations (FLOPs) and parameters. The tracker first uses a new fully CNN to extract image features, and introduces a channel attention mechanism in the feature extraction process to improve the representation ability of convolutional features. Then use the FPN to fuse the convolutional features of high and low layers, learn the similarity of the fused features, and train the fully CNNs. Finally, the heterogeneous convolutional kernel is used to replace the standard convolution kernel to improve the speed of the algorithm, thereby making up for the efficiency loss caused by the feature pyramid model. In this article, the tracker is experimentally verified and analyzed on the VOT-2017, VOT-2018, OTB-2013, and OTB-2015 datasets. The results show that our tracker has achieved better results than the state-of-the-art trackers.

This article concentrates on solving the k -winners-take-all (k WTA) problem with large-scale inputs in a distributed setting. We propose a multiagent system with a relatively simple structure, in which each agent is equipped with a 1-D system and interacts with others via binary consensus protocols. That is, only the signs of the relative state information between neighbors are required. By virtue of differential inclusion theory, we prove that the system converges from arbitrary initial states. In addition, we derive the convergence rate as O(1/t) . Furthermore, in comparison to the existing models, we introduce a novel comparison filter to eliminate the resolution ratio requirement on the input signal, that is, the difference between the k th and (k+1) th largest inputs must be larger than a positive threshold. As a result, the proposed distributed k WTA model is capable of solving the k WTA problem, even when more than two elements of the input signal share the same value. Finally, we validate the effectiveness of the theoretical results through two simulation examples.

This article addresses the problem of observer-based fuzzy proportional-integral-derivative (PID) control for a class of nonlinear systems subject to degraded measurements and randomly perturbed sampling periods (RPSPs). In the existing results, the degraded measurements and RPSPs are handled separately, where the sampling of different sensors is usually assumed to be synchronous. In our work, a comprehensive model is built to reflect the joint effects of degraded measurements and RPSPs by using a series of stochastic variable sequences and a set of Markov processes. In this model, the sampling periods of each sensor are allowed to be diverse, time-varying, and randomly perturbed, thereby fully capturing the environmental effects and device constraints. Different from the existing literature that uses proportional type controllers, an observer-based fuzzy PID controller with a modified structure is proposed, which fully utilizes the system information. To overcome the difficulties of the incomplete measurement information, some auxiliary variables related to the sampling periods are introduced under which the measurement output is transformed into a form delayed with stochastic delays. Subsequently, by using the special variable separation and inequality technique, sufficient conditions are derived to ensure the exponentially ultimate boundedness of the closed-loop system in the mean-square sense. The desired gains for the observer and PID controller are obtained through the solution of an optimization problem. Last, the effectiveness of the developed approach is demonstrated through simulation examples.

This article is concerned with the secure distributed economic dispatch (DED) problem of microgrids. A quantized distributed optimization algorithm using the Paillier encryption–decryption scheme is developed. This algorithm is designed to optimally coordinate the power outputs of a collection of distributed generators (DGs) in order to meet the total load demand at the lowest generation cost under the DG capacity limits while ensuring communication efficiency and security. First, to facilitate data encryption and reduce data release, a novel dynamic quantization scheme is integrated into the DED algorithm, through which the effects of quantization errors can be eliminated. Next, utilizing matrix norm analysis and mathematical induction, a sufficient condition is provided to demonstrate that the developed DED algorithm converges precisely to the optimal solution under finite quantization levels (and even the three-level quantization usingsigntransmissions). Moreover, an encryption–decryption scheme is developed based on quantized outputs, which ensures confidential communication by leveraging the homomorphic property of the Paillier cryptosystem. Finally, the effectiveness and superiority of the implemented secure distributed algorithm are confirmed through a simulated example.

A credibility-based and improved Gaussian sum square root cubature Kalman filter (IGS-SRCKF) estimation method for nonlinear non-Gaussian systems is proposed to improve the accuracy of unmanned surface vehicles (USVs) motion state estimation in complex environments. This method elevates the linear Gaussian credibility theory to the nonlinear non-Gaussian credibility theory by utilizing the improved Gaussian sum algorithm and constructing the square root cubature Klaman filter pseudomatrix, solving the problem of defining trust factors in nonlinear non-Gaussian systems, and measuring the credibility of the filter. At the same time, the conditions for designing high-performance filters are obtained using this credibility theory. Then, a new USV motion state estimation method based on credibility and the IGS-SRCKF framework is proposed by combining the hunter–prey optimizer and dingo optimization algorithm based on the obtained conditions for designing high-performance filters. Finally, the progressiveness and superiority of this algorithm are verified by simulation and actual data.

Real-time and high-performance semantic segmentation is a crucial task in the scene understanding of autonomous vehicles. This paper focuses on this issue and proposes a transformer and convolutional neural networks (CNN) hybrid encoder-decoder structure SegTransConv. Firstly, we present a four-stage hierarchical encoder, and the feature extractor in each stage is composed of two transformer layers and CNN modules in series. In this way, the encoder better exploits the global contexts of the input and expands the receptive fields. In the U-shape decoder, the feature maps are upsampled through the proposed feature enhancement upsampling module (FE_Up). Then the knowledge distillation strategy is leveraged to improve the model performance under the guidance of the teacher network STDCNet. Finally, a novel evaluation metric is designed to comprehensively assess the accuracy, speed, floating-point operations (FLOPs), and parameters of real-time segmentation methods. Extensive experiments on two public datasets and self-collected images have evaluated the effectiveness of our method. SegTransConv-A and SegTransConv-B obtain 72.8% and 73.0% mIoU, respectively, at the inference speed of 68.0 FPS with an input resolution of$1024\times 512$.

In multiagent reinforcement learning, policy evaluation is a central problem. To solve this problem, decentralized temporal-difference (TD) learning is one of the most popular methods, which has been investigated in recent years. However, existing decentralized variants of TD learning often suffer from slow convergence due to the sensitive selection of learning rates. Inspired by the great success of adaptive gradient methods in the training of deep neural networks, this article proposes a decentralized adaptive TD$(\lambda )$learning algorithm for general$\lambda $with linear function approximation, referred to asD-AMSTD$(\boldsymbol {\lambda })$, which can mitigate the selective sensitivity of learning rates. Furthermore, we establish the finite-time performance bounds ofD-AMSTD$(\boldsymbol {\lambda })$under the Markovian observation model. The theoretical results show thatD-AMSTD$(\boldsymbol {\lambda })$can linearly converge to an arbitrarily small size of neighborhood of the optimal weight. Finally, we verify the efficacy ofD-AMSTD$(\boldsymbol {\lambda })$through a variety of experiments. The results show thatD-AMSTD$(\boldsymbol {\lambda })$outperforms existing decentralized TD learning methods.

Overcharging is an important safety issue in the charging process of electric vehicle power batteries, and can easily lead to accelerated battery aging and serious safety accidents. It is necessary to accurately predict the vehicle’s charging time to effectively prevent the battery from overcharging. Due to the complex structure of the battery pack and various charging modes, the traditional charging time prediction method often encounters modeling difficulties and low accuracy. In response to the above problems, data drivers and machine learning theories are applied. On the basis of fully considering the different electric vehicle battery management system (BMS) charging modes, a charging time prediction method with charging mode recognition is proposed. First, an intelligent algorithm based on dynamic weighted density peak clustering (DWDPC) and random forest fusion is proposed to classify vehicle charging modes. Then, on the basis of an improved simplified particle swarm optimization (ISPSO) algorithm, a high-performance charging time prediction method is constructed by fully integrating long short-term memory (LSTM) and a strong tracking filter. Finally, the data run by the actual engineering system are verified for the proposed charging time prediction algorithm. Experimental results show that the new method can effectively distinguish the charging modes of different vehicles, identify the charging characteristics of different electric vehicles, and achieve high prediction accuracy.

Light propagates in water with certain attenuation, which causes quality problems, such as color cast, low contrast, and low illumination, in underwater images. Moreover, it is generally difficult to obtain a large number of real-world underwater images and the corresponding ground truth in convoluted underwater environments. To address these two problems, this article recommends an underwater image enhancement scheme based on few-shot learning and multi-color space, called UIE-FSMC. Specifically, for the first time, we propose a lightweight underwater image enhancement network based on few-shot learning. We design a new strategy to train the network. Specific training steps include the following: First, synthetic underwater images are used for large-scale pre-training, which can obtain the initial weight of the network. Then, according to the characteristics of the data, meta-learning based on supervised and unsupervised loss is suggested. It further enhances the feature expression aptitude of the network by learning the external and internal characteristics of the data. Finally, fine-tuning based on supervised and unsupervised loss is designed. It further improves the accuracy, robustness, and generalization of the network through an average strategy. In addition, we design a post-processing method founded on the RGB and LAB color spaces. In the RGB color space, we suggest a local region-based adaptive color correction method for underwater images. In the LAB color space, we design a multi-scale local adaptive contrast enhancement method for the L channel and a local region-based color balance strategy for the AB channels. Quantitative and qualitative experiments on five different underwater image datasets show that the outcomes of UIE-FSMC are superior to those of other techniques. Furthermore, application research further validates the excellent performance of UIE-FSMC.

Localization in unknown environments is an essential requirement for vision navigation of robotic vehicles in intelligent transportation systems. However, moving objects in dynamic scenarios usually bring about great difficulty for robot localization, because motion estimation of robotic vehicles is disturbed by increasing feature outliers caused by moving objects. In order to improve the accuracy and robustness of robot localization, a novel saliency-induced moving object detection (SMOD) approach is proposed to filter out feature outliers for RGB-D-based simultaneous localization and mapping (SLAM) in complex dynamic workspaces. Firstly, three complementary motion saliency potentials, including motion energy (ME), spatiotemporal objectness (STO), and dynamic superpixels (DS), are modeled by fully analyzing spatial, temporal, appearance and depth cues in RGB-D inputs. They can be used to identify the dynamic objects effectively from diversely changing backgrounds. Then, a superpixel-level graph-based motion saliency (MS) measure is proposed to generate the MS map for reliable localization of the moving objects. The edge weights and background nodes on the graph are determined reasonably by fusing ME, STO, and DS, which is not vulnerable to background interferences. Furthermore, the SMOD approach is embedded into the front-end of ORB-SLAM3 as a pre-processing stage, in order to filter out feature outliers associated with the moving objects. Finally, the extensive experiments are performed to verify the accuracy and robustness of the proposed approach on the public dynamic datasets. The experimental results show that the SMOD method can detect the moving objects effectively in a variety of challenging dynamic environments, and separate the dynamic regions reliably from the irrelevant background. The data comparison demonstrates that the SMOD-SLAM navigation system can outperform other state-of-the-art dynamic visual SLAM (vSLAM) systems.

Influenced by many complex factors, it is very difficult to obtain high-performance industrial power load forecasting. The industrial power load forecasting is deeply studied by fusing some machine-learning methods for industrial enterprise power consumers. As a result, a novel power load forecasting method is proposed by taking into account the variation of load characteristics in different regions, industries, and production patterns. First, through the improved K -means clustering analysis, the historical load data are classified as the production patterns to which they belong. Then, the prediction algorithm combining reinforcement learning with particle swarm optimization and the least-squares support vector machine is proposed. Finally, the improved algorithm in this article is used for short-term load forecasting separately by the load data in different patterns after the above processing. The forecasting method in this article is based on data driven with real datasets. The results of the simulation experiment show that the improved prediction algorithm can distinguish the changes in different production patterns and identify the load characteristics of different regions and industries with high prediction accuracy, which has practical application value.

The rapid popularization of mobile devices makes it more convenient and cost-efficient to collect synchronized WiFi received signal strength (RSS) and inertial measurement unit sequences by crowdsensing. The transition model has proven to be a promising unsupervised localization approach that captures the transition relationship between the change of RSS signal space and the change of physical space, alleviating the need of extra knowledge for creating radio map. However, it faces two essential challenges in real-world deployments. First, model coverage affects its locating performance, because a specific transition model only represents its local space. Second, the instability of RSS leads to a conflicting relationship between changes of two spaces because of the complex environment and the heterogeneous type of devices. To address these challenges, we propose Lightgbm-CTMM, a novel unsupervised localization framework. First, a clustering method is adopted to capture the expected relationship to ensure robust coverage. Second, direction filter is employed to guarantee that the change in signal space corresponds to the change in physical space. The feasibility and effectiveness of Lightgbm-CTMM are evaluated by extensive experiments, and the locating performance of Lightgbm-CTMM is better than that of conventional approaches. Moreover, Lightgbm-CTMM reduces the work on quality assessment of trajectories.

Adaptive gradient algorithms have recently become extremely popular because they have been applied successfully in training deep neural networks, such as Adam, AMSGrad, and AdaBound. Despite their success, however, the distributed variant of the adaptive method, which is expected to possess a rapid training speed at the beginning and a good generalization capacity at the end, is rarely studied. To fill the gap, a distributed adaptive subgradient algorithm is presented, called D-AdaBound, where the learning rates are dynamically bounded by clipping the learning rates. Moreover, we obtain the regret bound of D-AdaBound, in which the objective functions are convex. Finally, we confirm the effectiveness of D-AdaBound by simulation experiments on different datasets. The results show the performance improvement of D-AdaBound relative to existing distributed online learning algorithms.

We consider decentralized large-scale continuous submodular constrained optimization problems over networks, where the goal is to maximize a sum of nonconvex functions with diminishing returns property. However, the computations of the projection step and the whole gradient can become prohibitive in high-dimensional constrained optimization problems. For this reason, a decentralized randomized block-coordinate Frank-Wolfe algorithm is proposed for submoduar maximization over networks by local communication and computation, which adopts the randomized block-coordinate descent and the Frank-Wolfe technique. We also show that the proposed algorithm converges to an approximation fact$(1-e^{-p_{\max }/p_{\min }})$of the global maximal points at a rate of$\mathcal {O}(1/T)$by choosing a suitable stepsize, where$T$is the number of iterations. In addition, we confirm the theoretical results by experiments.

As the power system develops from informatization to intelligence. Research on data services based on the Internet of Things (IoT) focuses more on application functions, but the research on the data quality of the IoT itself is insufficient. Long-term continuous operation of the big data IoT system has the risk of performance degradation or even partial fault, which leads to a decrease in the availability of collected data for intelligent analysis. In this article, based on the power IoT message data, the characteristics are established through a variety of improved detection methods, and then the abnormal data type is obtained through Q learning and fusion of the random forest (RF) identification features. Finally, the topology of the specific power user IoT system is combined with kernel principal component analysis (KPCA) + improved RF algorithm getting the abnormal location of the IoT. The results show that the research method has a significantly higher positioning accuracy (from 61% to 97%) than the traditional RF method, and the combination method has more advantages in parameter adjustment and classification accuracy than directly using a multilayer perceptron (MLP).

With the popularity of mobile devices and the emergence of various audio-editing tools, it becomes easier to produce and forge audio files. Many criminals will forge false audio information as evidence. Therefore, audio forensics technology becomes particularly important. Audio recording device identification technology, which can verify the authenticity and uniqueness of the evidence obtained, is one of the promising branches of audio forensics technology. In this article, a novel neural-network-based framework using the device noise feature is proposed to identify the source of recording according to the device traces generated by the device during the recording. We also propose a new neural network model RARS (Recognition of Audio Recording Source based on residual neural network). The proposed framework achieves state-of-the-art performance on MOBIPHONE, the only publicly available dataset in this field. Moreover, we build a new dataset based on the latest mobile phones and tablet devices. Our method achieves good performance on both the two datasets, which proves that our model has a certain degree of reusability and robustness.

This paper is concerned with the finite-time synchronization (FTS) of memristor-based recurrent neural networks (MRNNs) combined with inertial items and mixed delays, where both the discrete delays and bounded distributed delays are included. First, MRNNs with inertial items are of second-order state derivatives, thereby differing from the classical first-order MRNNs and bringing difficulties to study the dynamics of such systems. By using the order-reduction method, such kind of second-order MRNNs is transferred into conventional first-order differential systems. Then, under two kinds of designed feedback controllers, several sufficient conditions are derived ensuring the FTS of MRNNs with inertial items and mixed delays. Finally, numerical simulations are provided to show the effectiveness of the results and one application is also presented in pseudorandom number generation.

In this paper, adaptive quantized state estimation fusion is deeply studied. To approach the model mismatching problem induced by random quantization, some quantized Kalman filters have been presented in the previous work, such as the quantized Kalman filter with strong tracking filtering (QKF-STF), the variational Bayesian adaptive quantized Kalman filter (VB-AQKF), and a centralized fusion frame-based complex quantized filter called variational Bayesian adaptive QKF-STF (VB-AQKF-STF). Based on the previous work for the single sensor system, a distributed complex quantized filter is designed in this paper. A novel quantized Kalman filter based on multiple-method fusion scheme (QKF-MMF) is proposed. Similar to the VB-AQKF-STF, the QKF-MMF can also realize joint estimation on the state and the quantization error covariance under the distributed fusion frame. Furthermore, it extends the single sensor results to multisensor tracking systems by using centralized and distributed fusion frames. Two multisensor quantized fusion estimators are proposed for a parallel structure with main-secondary processors in the fusion center. The weighted fusion and embedded integration ways are deeply applied to design the multisensor quantized fusion methods. The proposed work can perfect the quantized estimation algorithms and provide different choices for practical engineering applications.

As the main performance self-evaluation index of the Kalman filter, the estimation error covariance (EEC) has been used to design the allocation cost function of task and resources for sensor tracking networks. For nonlinear systems, the sensor allocation method based on the EEC needs to adjust the allocation plans after obtaining the filtering results. Meanwhile, recent investigations have indicated that the self-evaluation function EEC of the Kalman filtering is universally inapplicable in practical applications, for which the estimation models are generally mismatched due to difficulty in accurately training parameters and approximation of nonlinear systems. Thereby, the sensors cannot be properly allocated by using the EEC as a preliminary criterion. Alternatively, observable degree (OD) is a naturally quantitative measure on observability and can be utilized to effectively measure the estimation performance. In this paper, the OD analysis with scale transform invariance for nonlinear systems is studied by using the unscented Kalman filter, the pseudostate transition matrix, and the pseudo observation matrix on the basis of the results of linear systems. Afterward, the OD of nonlinear fusion systems, the sensor utilization efficiency, the priority of tasks, and the sensor performance and sensitivity are jointly considered to formulate the optimization problem for sensor allocation. The genetic algorithm with intelligent learning function is employed to solve the optimization problem. Moreover, extensive simulation demonstrates the feasibility of the proposed approach.

In this paper, we address the problem of sound source localization in reverberant environments. Time-delay estimation (TDE) methods are widely employed to locate sound sources based on the time differences of arrival (TDOAs) of signals received at different microphone pairs. In strong reverberations, the highest peak of the localization function is not necessarily from the true source resulting from the multi-path effect. Our previously proposed method based on the optimal peak association (OPA) aims to extract multiple peaks from the localization function for each microphone pair and find out the optimal association of TDOAs corresponding to the same sound source. However, due to the limitation of geometric configuration of microphones and possible missed detections, some microphone pairs fail to provide high-quality TDOA measurements. An improved OPA method is developed in this work based on the robust least squares which can determine the weights adaptively in terms of their respective observation accuracy. Experimental results demonstrate the superiority of the proposed method compared with the original OPA method in reverberant environments.

This paper mainly regulates a flexible wing of a micro aerial vehicle to track two spatiotemporally varying trajectories. By utilizing Lyapunov's direct method, two boundary control laws are designed to guarantee uniform boundedness of the closed-loop target system along the time axis. Based on Schur complement lemma, nonlinear inequalities derived from the theoretical deduction are rewritten as matrixes, which are solved through the LMI toolbox in MATLAB. In addition, the tracking control problem is formulated as an optimization problem. The simulation examples are conducted to prove the effectiveness of the proposed boundary control laws.

Performance of the Kalman filter (KF) is degraded when dealing with nonlinear dynamic systems. For a kind of nonlinear biomechatronics system, a fifth-degree ensemble iterated cubature square-root information filter (EsFICIF), which can effectively improve estimation performance, is proposed by combing many estimation schemes. Moreover, the associated multisensor fusion is deeply studied based on this proposed nonlinear filter in this paper. That is, four classic nonlinear fusion methods, which include augmented measurements fusion, weighted measurements fusion, sequential filtering fusion, and distributed filtering fusion, are compared on estimation performance. The motivation of this paper is to extend the work on estimation performance comparison of nonlinear fusion methods based on the conventional extended KF and to validate some basic conclusions existed in the traditional linear data fusion theory based on the proposed EsFICIF. The estimation accuracies of the four nonlinear fusion methods are compared and the exchanging property of measurements update order is also discussed. It is observed that, when the measurement properties are identical, the estimation accuracies of augmented measurements fusion, weighted measurements fusion, and distributed feedback fusion are equivalent, while the sequential filtering fusion does not hold. Furthermore, the exchanging property of the measurements update order of the sequential filtering fusion can no longer be guaranteed. These results further show some basic conclusions existed in linear fusion theory are no longer valid for nonlinear systems and the conclusions based on the EKF are still available for more complex nonlinear filters. Finally, numerical examples are provided to validate the results given in this paper.

This paper investigates dynamic balance optimization and control of quadruped robots with compliant/flexible joints under perturbing external forces. First, we formulate a constrained dynamic model of compliant/flexible joints for quadruped robots and a reduced-order dynamic model is developed considering the robot interaction with the environment through multiple contacts. A dynamic force distribution approach based on quadratic objective function is proposed for evaluating the optimal contact forces to cope with the external wrench, and fuzzy-based adaptive control of compliant/flexible joints for quadruped robots is proposed to suppress uncertainties in the dynamics of the robot and actuators. The dynamic surface control approaches and fuzzy learning algorithms are combined in the proposed framework. All the signals of the closed-loop system have proven to be uniformly ultimately bounded through Lyapunov synthesis. Simulation experiments were performed for a quadruped robot with compliant/flexible joints. The benefits of its tracking accuracy and robustness indicate that the proposed framework is promising for the robots with payload uncertainties and external disturbances.

Existing work on flooding in wireless sensor networks (WSNs) mainly focuses on single-packet problem, while the work on sequential multipacket problem is surprisingly little. This paper proposes OppCode, a new opportunistic network-coding-based flooding architecture for multipacket dissemination in WSNs, where both unreliable and correlated links commonly exist. Instead of flooding a single packet each time, each node encodes multiple native packets chosen from a specific fixed-size page to an encoded packet and then rebroadcasts it further. The key idea consists of two parts. One is opportunistically coding decision, in which each node grasps every possible coding opportunity greedily to maximize its aggregate coding gain of all neighbors based on the probabilistic estimation of packets each neighbor already has. The other is paged collective acknowledgements (ACKs), in which one rebroadcast acts as not only an implicit ACK of successful disseminations of all packets in the entire page for the sender, but also probabilistic ACK to update page-scale per-packet coverage estimations for its neighbors in a batch. Experiments based on extensive simulations and 21-node testbed show that OppCode significantly increases performance of multipacket flooding in terms of reliability, transmission overhead, delay, and load balance.

The identification of vessel collision risk for a Maritime Intelligent Transport System (MITS) is crucial for maritime safety and management. This paper considers the identification of the Systematic Collision Risk (SCR) for an MITS based on AIS data, which is obtained by wireless communication among vessels and between vessels and shore-based stations. SCR is modeled as a function of the collision risk of each vessel. A computing method for the SCR of a two-vessel case is proposed. Meanwhile, a hierarchical clustering based simplification algorithm is provided and applied to transform the topology of an MITS, thus simplifying the computing of the SCR. Based on the two-vessel case and transformation, a bottom-to-top weighted fusion method is employed to calculate the SCR for an MITS. Extensive numerical examples of simulative and real AIS data verify the effectiveness of our modeling and computing.

Square-root cubature Kalman filter (SCKF) is more effective for nonlinear state estimation than an unscented Kalman filter. In this paper, we study the design of nonlinear filters based on SCKF for the system with one step noise correlation and abrupt state change. First, we give the SCKF that deals with the one step correlation between process and measurement noises, SCKF-CN in short. Second, we introduce the idea of a strong tracking filter to construct the adaptive square-root factor of the prediction error covariance with a fading factor, which makes SCKF-CN obtain outstanding tracking performance to the system with target maneuver or abrupt state change. Accordingly, the tracking performance of SCKF is greatly improved. A universal nonlinear estimator is proposed, which can not only deal with the conventional nonlinear filter problem with high dimensionality and correlated noises, but also achieve an excellent strong tracking performance towards the abrupt change of target state. Three simulation examples with a bearings-only tracking system are illustrated to verify the efficiency of the proposed algorithms.