双电机驱动伺服系统径向基函数神经网络反推自适应控制

双电机驱动伺服系统中存在齿隙非线性环节,为了削弱齿隙非线性对系统的动态和稳态性能产生的不利影响,本文提出了一种新的自适应控制方法.首先给出了系统的状态空间模型并分析了双电机同步联动控制的原理,然后应用改进的反推方法,在考虑系统所有的状态变量都能收敛的基础上,引入虚拟控制量,通过逐步递推选择Lyapunov函数,利用径向基函数(radial basis function, RBF)神经网络在线逼近系统中的不确定函数,设计了基于状态反馈的RBF神经网络反推自适应控制器,并进行了稳定性分析.将单纯的反推控制和RBF神经网络反推自适应控制的仿真结果对比,发现后者的优越性高于前者.最后在实际系统中进行试验,验证了所提控制策略的可行性.

Radial-basis-function neural network backstepping adaptive control of dual-motor driving servo system

Backlash nonlinearity exists in dual-motor driving servo systems. To weaken the adverse effect of backlash nonlinearity on system dynamic and steady performance, a new adaptive control strategy was proposed. The state-space model of the system was first established and then the principle of dual-motor synchronous control was analyzed. By introducing the virtual control quantity on the basis of considering that all of the state variables of the system can converge, using an improved backstepping approach and recursively selecting the Lyapunov function, and adopting a radial-basis-function (RBF) neural network to approximate the uncertain function in the system, a state feedback-based RBF neural network backstepping adaptive controller was developed and its stability was analyzed. By comparing the simulation results of the mere backstepping control and the RBF neural network backstepping adaptive control, it is obvious that the superiority of the latter is higher than that of the former. Finally, experiments were carried out in the actual system to verify the feasibility of the proposed control strategy.