ISA transactions最新文献

Microgrids play an important role in stabilizing the electrical grid and they are the best route to develop green and sustainable energy. Since microgrids are expanding rapidly, it is necessary to consider the related control issues including power quality, bidirectional power flow, voltage and frequency control, and stability analysis. One of the main measurement challenges is the communication delay. It means the delay in sending data from the sensor or measuring unit to the processing unit. The communication delay gets more important when the microgrid is widespread and complex. In this paper, a novel soft switching voltage control system is proposed to solve the voltage control problem of a widespread micro-grid while there are time-varying communication delays. The novel soft switching method is based on a static output feedback controller and deep neural networks. Another novelty of this paper is considering the 33-bus microgrid as a large-scale system that helps develop local and central controllers. The simulation's results show the effectiveness of a soft switching controller in the presence of dynamic time-varying communication delays. It means that while encountering static communication delays, the static output feedback controller without a soft switching method is sufficient in a large-scale microgrid.

Approximation-free control effectively addresses uncertainty and disturbances without relying on approximation techniques such as fuzzy logic systems (FLS) and neural networks (NNs). However, singularity problems-where signals exceed preset boundaries under dynamic operating conditions-remain a challenge. This paper proposes an improved approximation-free control (I-AFC) method for the multi-agent system, which introduces a novel singularity compensator, providing a low-complexity design with exceptional adaptability while reducing the risk of singularity issues under changing working conditions (random initial values, system parameter variations, and changes in topology graph and followers' dynamics). Furthermore, theoretical analysis guides parameter selection by demonstrating the method's favorable convergence rate and appropriate control gain. Simulation results validate the approach.

In this paper, a novel recursive hierarchical parametric identification method based on initial value optimization is proposed for Wiener-Hammerstein systems subject to stochastic measurement noise. By transforming the traditional Wiener-Hammerstein system model into a generalized form, the system model parameters are uniquely expressed for estimation. To avoid cross-coupling between estimating block-oriented model parameters, a hierarchical identification algorithm is presented by dividing the parameter vector into two subvectors containing the coupled and uncoupled terms for estimation, respectively. To guarantee consistent estimation on these parameters, an auxiliary block model is designed to predict the inner unmeasurable variables of the Wiener-Hammerstein system for computational iteration. Furthermore, two adaptive forgetting factors are designed to accelerate the convergence rates on estimating both coupled and uncoupled parameters. To overcome the issue of initial value sensitivity involved with the traditional recursive least-squares based algorithms for parameter estimation, a particle swarm optimization (PSO) algorithm based on two different excitation signals is given for initial value optimization of the proposed recursive identification algorithm. Meanwhile, the convergence property of the proposed algorithm is clarified with a proof. Finally, an illustrative example and experiments on a micro-positioning stage are performed to validate the merit of the proposed method.

This paper explores a novel challenge regarding bidirectional Automated Guided Vehicles (AGVs): supervisory control amidst potential sensor faults. The proposed approach uses an event-based control architecture, guided by Supervisory Control Theory (SCT), to achieve non-blocking routing of AGVs. Unlike most routing approaches assuming full event observability, this paper investigates scenarios where events might become unobservable due to sensor faults or disturbances, which may affect the supervisor efficiency. The paper addresses two new key issues regarding AGV systems. First, it examines the diagnosis problem of automated transport systems from a discrete-event systems perspective. Secondly, it presents a control architecture enhanced with a diagnostic layer to improve fault tolerance. The theory of automata and languages is used to address control and diagnostic issues. The proposed methodology offers a systematic approach to design specification and diagnostic automata for routes shared by AGVs. The new specification automata integrate information from the diagnostic automata via synchronized transition guards, guaranteeing the synthesis of a robust supervisor that avoids deadlocks even when observability is compromised. The efficiency of the proposed architecture is examined and showcased by simulation. In addition, a modelling framework based on stochastic timed automata is introduced, applying model checking to assess system reliability which is redefined as the probability of deadlock avoidance.

An independent metering system (IMS) realizes the decoupling of the meter-in and meter-out orifices. The energy efficiency of the hydraulic system can be effectively improved by switching between different operational modes. However, the tracking accuracy of the IMS mode-switching system is difficult to ensure, which can easily lead to instability in the hydraulic system. In view of this, this paper proposes a mode switching controller based on an IMS. First, the K-filters theory is innovatively applied to the mode switching hydraulic system to estimate unmeasurable state variables of a system accurately. In addition, fuzzy logic systems (FLSs) are applied to handle the unmodeled errors and disturbances in the mechanical system dynamics model and hydraulic system. Further, aiming at the stability and trajectory tracking problems in the mode switching control (MSC) process of an IMS, the average dwell time (ADT) stability analysis method is applied to the mode switching hydraulic system to construct a set of switching rules to make the closed-loop switching system stable. Moreover, based on the prescribed performance control (PPC) theory, all state errors of a hydraulic system are guaranteed to reach the performance function constraint boundary at the specified time. Also, a dynamic surface control (DSC) technique is used to avoid the explosion of computational complexity caused by iterative differentiation inherent in the traditional backstepping method. Finally, the feasibility and effectiveness of the proposed method are verified by simulation, and experiments are carried out on mini-excavators. The results show that the designed controller can not only ensure the tracking accuracy, but also effectively suppress the instability of the hydraulic system caused by MSC.

In recent years, exoskeleton robots have attracted great interest from researchers in the area of robotics due to their ability to assist human functionality improvement. A wearable lower limb exoskeleton is aimed at supporting the limb functionality rehabilitation process and to assist physical therapists. Development of a stable and robust control system for multi-joint rehabilitation robots is a challenging task due to their non-linear dynamic systems. This paper presents the development of a Swarm-Initialized Adaptive (SIA) based controller, which is a combination of a swarm-based intelligence, named Swarm Beetle Antenna Searching (SBAS), and an adaptive Lyapunov-based controller. The SBAS initializes the parameters of SIA to efficiently improve the performance of the control system and then these controller parameters are updated by an adaptive controller. The control system is validated in a lower limb exoskeleton prototype with four degrees of freedom, using a healthy human subject for sit-to-stand and walking motions. The experimental results show the applicability of the proposed method and demonstrate that our approach obtained efficient control performance in terms of steady-state error and robustness and can be used for a lower limb exoskeleton to improve human mobility.

This paper addresses the critical challenge of interpretability in machine learning methods for machine fault diagnosis by introducing a novel ad hoc interpretable neural network structure called Sparse Temporal Logic Network (STLN). STLN conceptualizes network neurons as logical propositions and constructs formal connections between them using specified logical operators, which can be articulated and understood as a formal language called Weighted Signal Temporal Logic. The network includes a basic word network using wavelet kernels to extract intelligible features, a transformer encoder with sparse and structured neural attention to locate informative signal segments relevant to decision-making, and a logic network to synthesize a coherent language for fault explanation. STLN retains the advantageous properties of traditional neural networks while facilitating formal interpretation through temporal logic descriptions. Empirical validation on experimental datasets shows that STLN not only performs robustly in fault diagnosis tasks, but also provides interpretable explanations of the decision-making process, thus enabling interpretable fault diagnosis.

Relying on composite nonlinear feedback, an output-feedback controller is robustly addressed in the singular models with uncertainties, disturbances and time-delays. For this purpose, an observer-based compensator is utilized to realize the purpose. In the presence of disturbance and uncertainty, it is demonstrated that the tracking error and the states of the overall system are ultimately bounded. Moreover, the asymptotic stability would be specifically established without the external disturbance and uncertain terms. Employing the linear matrix inequality, the control design is translated into an optimization problem. Hence, in solving such an optimization issue, the coefficients of the estimator and the control law are determined simultaneously. Some simulations are provided to show the advantages of the planned strategy compared to a similar one.

Existing cross-domain mechanical fault diagnosis methods primarily achieve feature alignment by directly optimizing interdomain and category distances. However, this approach can be computationally expensive in multi-target scenarios or fail due to conflicting objectives, leading to decreased diagnostic performance. To avoid these issues, this paper introduces a novel method called domain feature disentanglement. The key to the proposed method lies in computing domain features and embedding domain similarity into neural networks to assist in extracting cross-domain invariant features. Specifically, the neural network architecture designed based on information theory can disentangle key features from multiple entangled latent variables. It employs the concept of contrastive learning to extract domain-relevant information from each data point and uses the Wasserstein distance to determine the similarity relationships across all domains. By informing the neural network of domain similarity relationships, it learns how to extract cross-domain invariant features through adversarial learning Eight multi-target domain adaptation tasks were set up on two public datasets, and the proposed method achieved an average diagnostic accuracy of 96.82%, surpassing six other advanced domain adaptation methods, demonstrating its superiority.

This paper investigates the self-triggered control for stabilizing an n-dimensional linear time-invariant system under communication constraints, including finite bit rates and transmission delay. The concerned system is further perturbed by bounded process noise. To resolve these issues, a self-triggering strategy is proposed. Specifically the proposed self-triggering strategy selects the next sampling time from a set of pre-designed time instants based on the sampled system states. By fully exploiting the encoded information of receive time instants of feedback packets, we can achieve the desired input-to-state stability (ISS) at a lower bit rate than that of periodic sampling. Moreover, the proposed self-triggering strategy is free of the burdens of continuously monitoring the system state compared with event-triggered sampling strategies. The efficiency of the proposed self-triggering strategy is further confirmed by simulations.