【摘要】 两相混合式步进电机是一个由脉冲驱动的多变量、强耦合的非线性系统，针对步进电机运行易受内外界扰动的影响从而产生失步、抖振等问题，提出了一种基于扩张状态观测器(ESO)的反步控制系统，在传统ESO的基础上，引入了径向基函数神经网络(RBF),对ESO的参数进行实时调整，同时引入粒子群算法(PSO),对反步控制器参数进行优化。研究结果表明，与传统反步控制相比，采用改进ESO的反步控制系统能在空载及负载状态下有效的提高系统的跟踪精度，并提高系统的响应速度。更多还原

【Abstract】 The two-phase hybrid stepper motor is a multi-variable, strongly coupled non-linear system driven by pulses, so the operation of the stepper motor is susceptible to internal and external disturbances, resulting in problems such as lost steps and jitter.To address this problem, this paper proposes a backstepping control system based on an expansive state observer(ESO),which introduces a radial basis function neural network(RBF) to adjust the parameters of the ESO in real time and a particle swarm algorithm(PSO) to optimise the backstepping controller parameters on the basis of the conventional ESO.The results show that the backstepping control system with the improved ESO can effectively improve the tracking accuracy and increase the response speed of the system under no-load and load conditions compared with the traditional backstepping control.更多还原