Neural Network Prescribed Performance Sliding Mode Control for Pump Control Electrohydraulic Servo System

针对泵控电液伺服系统中常见的参数不确定性和未知干扰问题，文章设计了一种结合RBF神经网络的预设性能滑模控制方法(RBFPPCBSMC)。首先，基于模型设计了一种干扰观测器(DOB)对未知扰动进行估计，并采用反步法设计改进趋近律的滑模控制律，通过双曲正切函数(tanh)构造滑模面切换函数，有效抑制滑模控制中的高频抖振现象。其次，设计径向基函数(RBF)神经网络对系统未建模动态进行自适应补偿。然后，引入规定性能约束(PPC)，确保瞬态和稳态位置响应在要求的有界范围内，进一步降低系统的跟踪误差。通过Lyapunov稳定性理论，证明了采用该控制方法的闭环系统的稳定性。为了验证RBFPPCBSMC的有效性，文章进行了详细的仿真对比。仿真结果表明，该控制器能够实现泵控电液伺服系统的精准控制，并有效应对模型参数不确定性和外部扰动带来的挑战。To address the common issues of parameter uncertainty and unknown disturbances in pump control electrohydraulic servo systems, this paper proposes a novel control strategy combining RBF neural networks with preset performance sliding mode control (RBFPPCBSMC). Initially, a disturbance observer (DOB) is designed based on the system model to estimate unknown disturbances. The sliding mode control law is then formulated using a backstepping approach. The sliding surface switching function is constructed via the hyperbolic tangent function (tanh), effectively mitigating the high-frequency chattering phenomenon inherent in sliding mode control. In parallel, a radial basis function (RBF) neural network is synergistically designed to achieve adaptive compensation for the system’s unmodeled dynamics. To ensure that both transient and steady-state position responses remain within desired bounds and to minimize tracking errors, a prescribed performance constraint (PPC) is incorporated. The stability of the closed-loop system under this control method is rigorously proven using the Lyapunov stability theory. Detailed simulation comparisons are conducted to validate the effectiveness of the RBFPPCBSMC. Simulation results demonstrate that the proposed controller achieves precise control of the electro-hydraulic servo system and effectively handles challenges posed by model uncertainties and external disturbances.