ABSTRACT This paper proposes a predefined‐time command‐filtered adaptive fuzzy tracking control method applicable to stochastic nonlinear systems. Unlike existing predefined‐time stability criteria, which are exclusively formulated for deterministic systems, this paper puts forward a criterion for practical predefined‐time stability in probability, thereby extending the predefined‐time control concept to the more challenging stochastic domain. Integrating the proposed stability criterion within the backstepping framework and utilizing fuzzy logic systems (FLSs), a predefined‐time adaptive fuzzy tracking control method is developed, employing a predefined‐time command filter to resolve the classic explosion of complexity issue. This control method guarantees the convergence of the tracking error to a small neighborhood of the origin in probability within any user‐predefined time. Finally, the theoretical claims are validated through dual‐platform simulations. Comparative numerical simulations on a second‐order stochastic system demonstrate a 43.89% reduction in the overall root mean square error and significantly faster transient convergence compared to conventional fixed‐time methods, while practical feasibility is verified on a robotic manipulator.
ABSTRACT This paper studies decentralized event‐triggered control for a class of nonlinear interconnected systems with bounded time delays and actuator saturation. Saturation is handled through a polytopic representation, enabling a sampled‐data event‐triggered implementation in which each subsystem updates its control input only at triggering instants. A delay‐dependent stability analysis is developed using a Lyapunov–Krasovskii functional combined with a Bessel–Legendre integral inequality, leading to linear matrix inequality (LMI) conditions that guarantee closed‐loop stability under the proposed triggering mechanism. To connect performance and triggering design, we interpret feasible LMI solutions as computable quadratic Lyapunov–Krasovskii storage certificates (not physical energy) that can be used to compare admissible triggering parameters and reduce unnecessary transmissions. For implementation, we also discuss an optional learning‐assisted selector that chooses or selects from a precomputed library of LMI‐feasible parameters using local measurements, while preserving the same LMI‐certified stability guarantees. Simulation results on a three‐zone HVAC example, with additional validation on a multimachine power network, illustrate substantial reductions in communication events while maintaining stable closed‐loop performance.
Considering the presence of modeling errors, load variations, and time-varying parameters in collaborative robot joint modules, we propose a novel robust approximate constraint-following control algorithm based on the Udwadia-Kalaba (U-K) framework. The core innovation of this work lies in the analytical decomposition of the control law into three distinct components: A nominal term that governs the baseline dynamics, a correction term to address initial condition incompatibility, and a robust term designed to compensate for uncertainties. This structure enables constraint compliance and robustness without relying on online parameter estimation. To validate the proposed controller, we carried out both numerical simulations and real-time experiments on a joint module test platform equipped with two representative friction models, using the fast control prototype system CSPACE. Simulation results indicate that the step response stabilization time is approximately 1 s, while the sinusoidal trajectory tracking error remains below 0.01 degrees. Theoretical analysis and experimental results consistently demonstrate that the proposed method significantly improves the dynamic performance and robustness of the joint module system.
This work concentrates on the synchronization analysis of multi-layer complex dynamical networks, with particular attention on achieving inter-layer synchronization for the desired networks under asymmetric Lyapunov-Krasovskii functional (LKF) approach. Moreover, the amount of data transmission and reduced communication resources being the major concern in case of networks, a memory event-triggered control technique is developed for the considered multi-layer networks. This attempted strategy utilizes preceding information about the transmitted signal to determine whether or not to transmit the data packets, eventually reducing the amount of data transmission. With the implementation of appropriate asymmetric LKF, a set of adequate requirements are attained based on linear matrix inequalities, such that the closed-loop inter-layer synchronization error system reaches the desired asymptotic stability. Furthermore, the simulation outcomes demonstrate the practicality of the proposed findings while facilitating the discussion with the aid of two numerical examples.
In this work, we introduce a novel output-feedback approach for the formation control of multiple nonholonomic vehicles that is robust to constant input disturbances and communication delays that vary over time. The proposed controller integrates an adaptive observer capable of reconstructing both the velocities and the external disturbances, together with a proportional-plus-damping-injection term that enhances system stabilization. To address the nonholonomic constraints of the vehicles, the control law is made time-varying through the use of a vanishing time-dependent term. With sufficient damping injection, our method ensures that all vehicles achieve consensus on a desired formation pattern and that all velocities converge to zero, while providing exact estimates of the disturbances. The contribution lies not in the velocity-disturbance observer, which is adopted from our previous work, but in its closed-loop integration with a new delay-aware, smooth, time-varying formation controller, together with the corresponding Lyapunov-Krasovskii stability certificate. The effectiveness of the approach is demonstrated through realistic numerical simulations involving six Pioneer 3-DX robots in the Gazebo environment.
