Showing papers by "Shun-Feng Su published in 2023"

Improved Reachable Set Estimation and Aperiodic Sampled-Data for T–S Fuzzy Markovian Jump Systems

Abstract: In this article, the problem of reachable set estimation (RSE) and the design of aperiodic sampled-data controller for Takagi–Sugeno fuzzy Markovian jump systems (T–S FMJSs) with unit-energy bounded disturbance (UEBD) and unit-peak bounded disturbance (UPBD) inputs are taken into consideration. First, sufficient conditions that all states of the T–S FMJSs are encompassed by ellipsoids under zero initial conditions are acquired via constructing a mode-dependent two-sided loop-based Lyapunov function and applying a linear matrix inequality approach. Second, the RSE is taken into account in the design of the state feedback aperiodic sampled-data controller with the aim that the resulting ellipsoid encompasses the reachable set of the closed-loop system. Finally, a nonlinear mass-spring model and a tunnel diode circuit model demonstrate the efficiency of the presented approach. In addition, this method is able to obtain a larger sampled-data period than other literature, thus saving bandwidth and reducing communication resources.

Event-Triggered SMC for Networked Markov Jumping Systems With Channel Fading and Applications: Genetic Algorithm.

Abstract: The event-triggered sliding-mode control (SMC) for discrete-time networked Markov jumping systems (MJSs) with channel fading is investigated by means of a genetic algorithm. In order to reduce resource consumption in the transmission process, an event-triggered protocol is adopted for networked MJSs. A key feature is that the signal transmission is inevitably affected by fading phenomenon due to delay, random noise, and amplitude attenuation in a networked environment. With the aid of a common sliding surface, an event-triggered SMC law is designed by adjusting the system network mode. Under the framework of stochastic Lyapunov stability, sufficient conditions are constructed to ensure the mean-square stability of the closed-loop networked MJSs, and the sliding region is reached around the specified sliding surface. Moreover, based on the iteration optimizing accessibility of objective function, an effective SMC approach under genetic algorithm is proposed to minimize the convergence region around the sliding surface. Finally, the effectiveness of the proposed method is proved by the F-404 aircraft model.

Robust Kalman Filter for Position Estimation of Automated Guided Vehicles Under Cyberattacks

Abstract: Automated guided vehicles (AGVs) have become a key part of many industries, where they handle the task of managing material flows. As a result, they are very important targets for cyberattacks to cripple organizations by using methods, such as false data injection (FDI) and denial of service (DoS) on sensors and measurement data. This article introduces a new robust Kalman filter (KF) for the AGV state estimation using Huber loss function. A statistical method known as M-estimation is used to solve the regression issue robustly and establish the equivalence between the KF and a specific least squares regression problem. For adaptive estimation of the unknown a priori state and observation noise statistics concurrently with the system states, M-robust estimators are developed. The proposed method tackles the state estimation issue against different kinds of cyberattacks, such as pulse, ramp, and DoS cyberattacks. The position estimation using constant velocity tracking based on KF employing with and without the robustification methods is compared. The results confirm the effectiveness and the robustness of the proposed filter approach against the FDI due to the cyberattacks compared with traditional filters. Furthermore, the proposed method is investigated practically with Adlink-ROS AGV.

High-Intensified Resemblance and Statistic-Restructured Alignment in Few-Shot Domain Adaptation for Industrial-Specialized Employment

Abstract: The industrial status quo has been changed hastily. Facilities require new technological factors and are less datadriven to make precise inferences at most. Thus, the use of virtual knowledge can be adapted to practical applications to limit human intervention in explanation, namely, comprehensive digital transformation. Deep transfer learning has the role of enabling transferable knowledge and awareness practice of annotating endpoints. However, practical transfer tasks are burdened while considering scarce or non-annotated instances accumulated from external sampling, and these may inevitably degenerate knowledge during training and further degrade performance. In this study, a novel framework composed of the spectral correlation alignment is proposed to enable spectral patterned structure for statistical criterion to diminish the domain disparity under a fewshot alignment and multiple attention orchestration to accelerate a few-shot adaptation. This technique also increases highconcentrated recognition in cross-domain similarity verification. The demonstrations are conducted under public visual adaptation bench-marks and realistic deployment. The empirical experiment illustrates that our approach is better in efficacy and adaptability under data-limited conditions. Besides, in similar realistic applications, it is evident that the proposed scheme is deployable and can be re-practicable to less time as well as data consumption, yet is better in performance.

Practically Predefined-Time Adaptive Fuzzy Tracking Control for Nonlinear Stochastic Systems.

Abstract: This article addresses the practically predefined-time adaptive fuzzy tracking control problem of strict-feedback nonlinear stochastic systems, where the system under consideration includes stochastic disturbances and uncertain parameters. First, in this study, practically predefined-time stochastic stabilization (PPSS) in the p th moment sense is introduced, and a Lyapunov-type criterion for PPSS is proposed to assure the stabilization of the system considered. With these ideas, based on the backstepping design method, a semiglobally practically predefined-time adaptive fuzzy tracking control algorithm is proposed with a fuzzy system used to approximate the unknown part of the system. Moreover, the settling time of the system response can be arbitrarily adjusted in a mean-value sense, and such freedom can be used to improve the stochastic finite-/fixed-time control results. Finally, a practical example and a numerical example of a comparison are provided to validate the effectiveness of the proposed control strategy.

Direct Fuzzy Adaptive Regulation for High-Order Delayed Systems: A Lyapunov-Razumikhin Function Method.

Abstract: A slow time-delay assumption restricts the application of control approaches for numerous systems which are constantly affected by multiple uncertainties, including parameters, control coefficients, and the asymmetric dead-zone input. This work presents a new adaptive method for a class of high-order nonlinear delayed systems by removing the so-called slow time-delay assumption and multiple uncertainties. Remarkably, with a novel Lyapunov-Razumikhin (L-R) function and a direct fuzzy adaptive regulation scheme, a memoryless adaptive feedback controller is skillfully constructed to guarantee that the output tracks the given reference signal while keeping the boundedness of all closed-system signals. Finally, the presented scheme is applied to control a single-link robot system.

Fuzzy Observer-Based Output Feedback Control of Continuous-Time Nonlinear Two-Dimensional Systems

Abstract: The fuzzy observer-based output feedback control of continuous-time nonlinear two-dimensional (2-D) systems is studied by Takagi–Sugeno (T–S) fuzzy models in this work. The plant information of the 2-D systems evolves along two independent directions dynamically. The nonlinear 2-D systems are firstly expressed by T–S fuzzy models with parameter uncertainties. By a Lyapunov method together with some convexification techniques, two methods are developed for fuzzy observer-based output feedback controller synthesis of the underlying fuzzy 2-D systems, and novel output feedback controller synthesis results are proposed within a convex optimization setup. Simulation studies are provided to illustrate the validity of the proposed methods.

Distributed Localization Estimation for Dynamic Multiagent Systems

Abstract: This article investigates the real-time localization problem of dynamic multiagent systems with repetitive operation characteristics under directed graph. A distributed localization estimation algorithm based on iterative learning is proposed. The barycentric coordinates calculated based on the relative distance are used to estimate the real coordinates of the agent. Different from the traditional estimation methods along the time axis, the proposed method utilizes the information of iteration axis simultaneously. In this method, the current estimation coordinates are updated by using the estimation coordinates of the same sampling time in previous iteration, the estimation accuracy is improved, and the velocity constraint is removed. Additionally, the real-time localization problem of dynamic multiagent systems under arbitrary deployment is concerned. An improved distributed localization estimation algorithm with signed coefficients based on iterative learning is proposed. Meanwhile, the results are also extended to the localization estimation of multiagent systems with arbitrary deployment in 3-D space. By introducing Richardson iteration and infinite norm, the global asymptotic convergence of the proposed methods is guaranteed. Finally, numerical simulations and the Qbot-2e robot experiment are provided to show the effectiveness and validity of the obtained results.

Adaptive Asymptotic Tracking Control for Flexible-Joint Robots With Prescribed Performance: Design and Experiments

Abstract: This study reports the adaptive asymptotic tracking control problem for flexible-joint (FJ) robot systems, the output tracking error can be kept within the prescribed range in the initial stage of system operation, as time approaches infinity, the asymptotic tracking result can be obtained. The prescribed performance function and the positive integrable time-varying function are introduced simultaneously in the control design of FJ robot systems for the first time. The control scheme is designed under the frame of the adaptive backstepping method and command filtered technique, which successfully avoids the problem of complexity explosion. The radial basis function neural networks are used to deal with unknown uncertainties and the adaptive laws are designed to approximate the norms of weight vectors and approximation errors. Finally, the feasibility of the proposed scheme is proved by the simulation and the experiment of the 2-link FJ robot on the Quanser platform.

Novel Adaptive Control for Flexible-Joint Robots With Unknown Measurement Sensitivity

Abstract: For the existing tracking control schemes of flexible-joint robots, precise sensor measurement is an implicit premise. However, idealized sensors are difficult to achieve due to manufacturing technology or other external factors. To this end, this paper further investigates the tracking control problem for flexible-joint robots with unknown measurement sensitivity. Specifically, for such multi-input multi-output Euler-Lagrange systems with completely unknown system dynamics, a novel measurement values-based adaptive control method is proposed by fusing sensitivity information and system variables into Lyapunov function candidates, where the restriction on system states in other the approximation lemma-based results is removed, since unknown nonlinearities are scaled by the structural characteristics of system variables. Above all, even if there are measurement errors, satisfactory tracking performance can be obtained by adjusting the design parameters, which is proved by rigorous theoretical analysis. Finally, hardware experiments further verify the effectiveness of the proposed method. Note to Practitioners —This work is motivated by the trajectory tracking control problem for flexible-joint robots under imprecise sensor measurements. Due to manufacturing technology limitations and component aging, there is inevitably a deviation between the measured values of sensors and real values, and this problem may become more prominent as the working environment of flexible-joint robots tends to become more complex. To our knowledge, most of the existing solutions for flexible-joint robots are developed based on precise sensor measurements, and they may fail to achieve satisfactory performance when real state information is not available. Moreover, the prior knowledge about model parameters and measurement sensitivity is difficult or impossible to exactly obtain in practice, which seriously hinders the further application of control methods that are dependent on system dynamics. To this end, this paper proposes a novel tracking control scheme based on measurement information for flexible-joint robots with unknown measurement sensitivity, where the dependence on model information is eliminated with the elaborately constructed Lyapunov function candidates, and the real tracking error is still adjusted to an acceptable range even if there are measurement errors. Preliminary experiments on a flexible-joint robot developed by Quanser company demonstrate the feasibility and effectiveness of the proposed method. In future studies, designing an effective scheme to achieve direct preset tracking control is the focus of the work.