This paper investigates robust exponential stability and control for a class of discrete-time switched stochastic Hopfield neural networks with impulsive effects. First, a unified analysis framework is established for discrete-time switched stochastic impulsive systems. By combining Lyapunov theory with the average dwell-time approach, sufficient conditions in terms of linear matrix inequalities (LMIs) are derived to guarantee robust exponential stability and prescribed disturbance attenuation. Then, these results are extended to switched stochastic Hopfield neural networks with impulsive effects by incorporating sector conditions on the activation functions, leading to tractable state-feedback controller design criteria. The developed method explicitly captures the coupled influence of switching, stochastic disturbances, and impulsive behavior. Finally, the effectiveness of the proposed method is verified through numerical examples of photovoltaic energy storage charging stations and networked unmanned vehicle systems.

This paper investigates finite-time stochastic stability for discrete-time switched stochastic systems with delayed impulses and asynchronous switching, typical in cyber-physical systems like microgrids and vehicle platoons. Using average dwell-time method, stability criteria are established for both stabilizing and destabilizing impulses, revealing a bidirectional regulation mechanism between impulse frequency and system stability. To address controller switching lags, finite-time stochastic stability conditions are derived via piecewise Lyapunov functions, and LMI-based controller synthesis is provided. The proposed framework offers a unified solution for security control under impulsive disturbances and asynchronous switching. Theoretical results are validated through microgrid frequency regulation and vehicle platoon anti-attack control, confirming effective suppression of random disturbances from impulses and switching.

In this paper, a cross-sensor generative self-supervised learning network is proposed for fault detection of multi-sensor. By modeling the sensor signals in multiple dimensions to achieve correlation information mining between channels to deal with the pretext task, the shared features between multi-sensor data can be captured, and the gap between channel data features will be reduced. Meanwhile, in order to model fault features in the downstream task, the salience module is developed to optimize cross-sensor data features based on a small amount of labeled data to make warning feature information prominent for improving the separator accuracy. Finally, experimental results on the public datasets FEMTO-ST dataset and the private datasets SMT shock absorber dataset (SMT-SA dataset) show that the proposed method performs favorably against other STATE-of-the-art methods.

Path planning in complex, dynamic environments presents a significant challenge. Deep Reinforcement Learning (DRL) offers an end-to-end solution but suffers from critical sample inefficiency and a “cold-start” problem. Imitation Learning (IL) accelerates training but is constrained by a performance ceiling and poor generalization. To address these limitations, we propose a novel Imitation–Reinforcement Hybrid Machine Learning Algorithm (Hybrid IL-RL). This framework balances exploration and performance via a two-stage process: First, an offline pre-training phase uses Behavioral Cloning (BC) with “non-expert” A* data from static environments for a “warm start”. Second, an online fine-tuning phase uses a DRL algorithm (SAC) to adapt this policy in complex, dynamic environments, allowing the agent to surpass the teacher’s limitations. Simulation experiments validate the approach. The framework demonstrates significantly faster convergence than DRL algorithms trained from scratch. Most critically, in the dynamic environment, our Hybrid IL-RL algorithm achieved the highest success rate (82.4%), while pure IL methods (BC, GAIL) failed due to poor generalization (e.g., 82.1% collision rate) and pure DRL methods struggled (approx. 51–56% success rate). Our results confirm the hybrid framework effectively solves the cold-start problem while using DRL to break the IL performance ceiling.

This paper investigates the pth moment globally asymptotically stable (p-MGAS) of discrete-time Markovian jump impulsive stochastic systems, which are crucial for modeling complex dynamics influenced by impulsive effect and Markovian jump. We propose a new framework using Lyapunov functions and stochastic analysis to derive stability conditions, incorporating both impulsive and non-impulsive time-varying parameters. The average impulsive interval method is used to handle impulsive effects more flexibly. The results are applied to neural networks with Markovian jumps and impulses via linear matrix inequality (LMI), demonstrating practical relevance. Two numerical examples verify the effectiveness of our methods. This study offers a robust approach to stability analysis for such complex systems.

In this article, a finite‐horizon optimal trajectory control strategy is developed for near space hypersonic vehicle (NSHV) longitudinal model with multi‐constraints including external disturbance, system modeling error, and input saturation. The whole control process has two parts: inner‐loop attitude control and outer‐loop trajectory control. First, the feedback linearization method is applied to design a tracking controller for outer‐loop system, and reference signals for inner‐loop attitude control can be obtained using Newton iteration method. Second, for the inner‐loop attitude system with multi‐constraints, a finite‐horizon optimal tracking control scheme consists of feedforward control input and adaptive dynamic programming based optimal feedback controller is designed. In this way, not only the adverse effects of above multi‐constraints are eliminated, but also the optimally tracking performances are guaranteed. Finally, the Lyapunov analysis method is utilized to ensure the stability of the entire closed‐loop control system, and simulation tests with respect to NSHV longitudinal trajectory tracking are supplied to verify the availability of the proposed strategy.

In this article, an adaptive event-triggered constrained control strategy is proposed for uncertain nonlinear systems with input constraints by using reinforcement learning (RL) technology and disturbance observer (DO). By constructing an actor-critic neural network (NN) framework, the unknown uncertainties can be tackled by online learning and more accurate compensation. The actor-NN is adopted for generating actions (regarded as compensation signals), and the critic-NN is employed to evaluate the performed actions (regarded as to monitor and assess the actor-NN performance, including the control performance). Moreover, a self-learning DO with learning ability is designed to estimate the external disturbance. On the basis of the backstepping control technology, the event-triggered control (ETC) method and the smooth approximation of input saturation nonlinearity, an improved event-triggered constrained control strategy is presented using RL technique, and the rigorous theoretical proofs of the closed-loop system stability and the avoidance of Zeno behavior are presented. The application for the quadrotor unmanned aerial vehicle (UAV) validates the effectiveness of the developed ETC approach.

The work presented in this article focuses on designing a disturbance-observer-based fixed-time controller for the quadrotor unmanned aerial vehicle (QUAV) system, which aims to research the attitude control issues of QUAV with state constraints and external disturbances. Firstly, a novel controller depend on adding one power integrator and fixed-time control technique is introduced to implement finite time convergence regardless of original states. Then, the external disturbances of QUAV system are estimated via the disturbance observer. Futhermore, the proposed method is demonstrated to be fixed-time stable through barrier Lyapunov analysis. In the end, theoretical results are also illustrated and supported by simulation studies.

This paper proposes a robust trajectory tracking control strategy based on the high-order fully actuated system approach and disturbance observer. In spite of the strong coupling characteristic and high nonlinearity of the quadrotor unmanned aerial vehicle (QUAV) system, the linear time-invariant system with desired eigenstructures can be obtained by designing a greatly simple robust controller via the direct parametric method. Meanwhile, a disturbance observer (DO) is designed to estimate the external disturbances, and the output of the DO can be adapted for the controller design to compensate for the affect of external disturbances on control performance. Eventually, the numerical simulation results of the QUAV trajectory tracking demonstrate the effectiveness of the developed robust control scheme.

In this paper, an asynchronous advantage actor-critic (A3C) based intelligent temperature control method is proposed for Continuous Stirred Tank Reactor (CSTR) system. Firstly, the overall framework of CSTR system is composed of global network, thread network, environment model and interaction mechanism is designed. Secondly, the process of interaction is designed: the thread network is trained to interact with environmental models by using an asynchronous multi-threaded approach, the trained parameters are updated and synchronized to the global network. Finally, the effectiveness and stability of A3C algorithm is verified by the simulation experiments. The simulation experiments show that the trained A3C network can control the output temperature of the CSTR system to reach the preset temperature and keep it the preset temperature, so the temperature control of the CSTR system is achieved in the A3C algorithm.

In this paper, an adaptive fuzzy decentralized output-feedback control is developed for the uncertain switched large-scale interconnected nonlinear system, which contains fast time-varying delays and un-known nonconstant control coefficients. By introducing and proving a new lemma, an improved analysis method is given based on the Lyapunov-Krasovskii (LK) functional, and it gets rid of the dependence of LK functional design on the bound of the derivative of the delay. Immeasurable states are estimated by co-designing the decentralized state observer and the innovative compensation system, which eliminates a widely adopted but strict assumption that the unknown control coefficient must be constant. By comprehensively designing the state observer, the compensation system and LK functionals, a novel adaptive fuzzy decentralized output-feedback control law is proposed, and stability of the closed-loop system is guaranteed theoretically based on an improved average dwell time. Finally, the effectiveness is further verified by complete simulation experiments.(c) 2023 The Franklin Institute. Published by Elsevier Inc. All rights reserved.

During the production of the shell of laptop and during the installation of the laptop its surface may be damaged by external factors. Therefore, its surface quality inspection is an essential and important part of the entire production process. At this stage, the detection of laptop appearance defects within the industry mainly relies on manual inspection, but manual inspection methods are inefficient and costly. In order to reduce the cost of manual labor, realize the intelligence of industrial production as well as improve the efficiency of inspection, in this paper, the YOLOv5 algorithm was used to create a deep learning model to investigate an effective method for detecting scratches defects on the appearance of laptops. In order to speed up the operation of the algorithm and improve the accuracy of the defect detection, the C3 module is used, and the activation function of the Conv module was modified, and the SiLU activation function was used instead of the Hardswish activation function; the experimental results show that the deep learning model trained with the improved YOLOv5 algorithm has a better performance for detecting the scratch defects on the appearance of laptops, not only accelerates the training speed of the model but also achieves an accuracy of 95.0% and a recall of 88%.

The trajectory planning and tracking control problem is studied in this paper based on prescribed performance method (PPM) for the small-scale unmanned autonomous helicopter (UAH) with wind-gust disturbances (WGDs) and unmeasurable states. For the purpose, the nonlinear model with flapping dynamics is established, and the transformation performance function is used to ensure that the errors of trajectory tracking satisfy the corresponding performance. A fractional-order observer is designed for unmeasurable states to estimate the flapping angles in actual flight, and the fractional-order extended state observer (ESO) is constructed to estimate the WGDs, respectively. On the basis of PPM and the designed observers, the fractional-order theory-based backstepping trajectory tracking control scheme is developed for the UAH system, and the three-dimensional trajectory is planned by the improved wolf pack algorithm. Then the stability of the entire system is proven through strict theoretical analysis. Finally, the simulation analysis on the UAH is presented to demonstrate the efficiency of the designed method.

针对具有随机噪声输入干扰、泊松随机波动干扰和控制输入饱和的近空间飞行器(near space vehicle,NSV)纵向轨迹控制模型,提出一种自适应鲁棒轨迹跟踪随机控制方案,实现对高度和速度参考信号的有效跟踪.对于外环轨迹控制,分别针对高度子系统和速度子系统,设计鲁棒随机控制器,通过数值计算方法对等效控制输入转化,得到内环姿态控制所需的姿态角参考信号.针对内环姿态控制问题,基于反步控制(backstepping control,BC)方法、动态面控制(dynamic surface control,DSC)技术和随机鲁棒控制方法,设计一种基于辅助系统的自适应鲁棒随机控制方案,实现随机干扰下的NSV姿态精确跟踪,衰减多重随机噪声干扰和参数不确定的影响,获得满意的鲁棒H¥控制性能.数值仿真进一步验证了所提方案的有效性.

针对一类具有未知外部干扰和系统不确定性的两级化学反应器系统,提出了一种基于干扰观测器的回馈递推跟踪控制方案.首先根据两级化学反应器的实际反应过程,建立严格反馈非线性系统模型,将系统反应过程中的未知外部干扰和系统不确定视为复合干扰,设计一种非线性干扰观测器估计复合干扰,并将干扰观测器输出用于控制器设计;进而利用回馈递推控制技术,设计两级化学反应器系统鲁棒控制方案,使系统实际输出跟踪上期望的参考信号,并通过Lyapunov稳定性理论分析闭环系统信号的有界稳定性;最后通过对两级化学反应器系统进行数值仿真实验,验证所提出跟踪控制方案的有效性.

Timely detection of notebook appearance defects is an important means to prevent products from being delivered to customers before leaving the factory.In industrial production, more emphasis is placed on fast and accurate detection methods, but the existing difficulties: 1. Defect samples are rare and difficult to obtain; 2. In high-resolution images, there are slight differences between abnormal samples and normal samples; 3. Slowly detection and insufficient accuracy.The existing methods mainly use a large amount of abnormal samples, so it is difficult to extend to the field of notebook appearance anomaly detection.To solve this problem, we designed a method that firstly uses unsupervised PatchCore which the algorithm was trained on normal samples and Defect GAN is used in test phase. To create a large number of verisimilitude abnormal samples and test these samples with PatchCore. On TKP-Surface datasets, the AUROC score of image-level anomaly detection achieves 96.1%, which meets the requirements of industrial applications.

In this paper, considering the simultaneous influence of multi-source disturbances, system modeling uncertainties and input–output constraints, an adaptive robust attitude tracking control scheme is proposed for near space vehicles (NSVs) which is expressed as a stochastic nonlinear system. A multi-dimensional Taylor polynomial network (MTPN) is utilized to handle the system uncertainties, and the nonlinear disturbance observer (NDO) based on MTPN is designed to estimate the external disturbances. Furthermore, by constructing the auxiliary system to tackle the input saturation and introducing the Tan-type barrier Lyapunov function (TBLF) to solve the output constraint, the constrained control strategy can be obtained. Combining with backstepping control (BC) technique and stochastic control method, an adaptive robust stochastic control scheme is developed based on NDO, MTPN, and auxiliary system, and the closed-loop system stability in the sense of probability is analyzed based on stochastic Lyapunov stability theory. Finally, numerical simulations further demonstrate the feasibility of the proposed tracking control scheme.

In present paper the finite-time bounded tracking control of discrete-time systems with external disturbance is considered. We constructed a new system so that the problem can be transformed into finite-time boundedness(FTB) problem of the output vector error system. Based on this, a tracking controller for the original system is designed to ensure that the system is finite-time bounded tracking of the reference signal. To illustrate the feasibility of the tracking controller, a numerical simulation result is also provided.

This article attempts to study the high angle of attack maneuver from the perspective of switched system control. In view of the complex aerodynamic characteristics, an improved longitudinal attitude motion model is presented, which is a switched stochastic nonstrict feedback nonlinear system with distributed delays. The significant design difficulty is the completely unknown diffusion and drift terms and distributed delays with all state variables. Based on a technical lemma and neural networks, an improved smooth state feedback control law for nonstrict feedback systems is proposed without any growth assumptions. To eliminate the influence of distributed delays, an improved Lyapunov–Krasovskii function is constructed, which skillfully removes the constraint of the upper bound of the delay change rate. Then, by combining the average dwell‐time scheme and stochastic backstepping technique, an adaptive neural network tracking control law is designed, which extends a newly proposed switched system stability condition to the stochastic switched system. Theoretical analysis and flight control simulation experiments are provided to illustrate the effectiveness of the proposed control method.

This article proposes a high-order disturbance observer (HODO) and dynamic surface control (DSC) technique-based adaptive fuzzy control scheme for nonlinear systems subjected to input saturation and external time-varying disturbances. First, based on a Sigmoid function, the saturation input is tackled by utilizing a well-defined nonlinear smooth function. Furthermore, HODO and fuzzy logic systems are used to estimate the external disturbances and to handle the lumped unknown functions, respectively. Then, by using the backstepping method and DSC technique, a HODO-based adaptive fuzzy tracking control scheme is proposed for nonlinear systems with saturation nonlinearity, uncertainties, and external disturbances. The Lyapunov analysis method is used to prove that all signals in the entire system are semiglobally uniformly ultimately bounded (SGUUB). In addition, the tracking error converges to a compact set with a tunable error bound determined by some design parameters. Finally, a numerical simulation of two-stage chemical reactor shows the effectiveness of the developed tracking control strategy.