电液伺服系统多模型鲁棒自适应控制

Multiple Model Robust Adaptive Control of Electro-Hydraulic Servo Systems

针对电液伺服系统中存在不确定非线性和强参数不确定性的问题,提出一种多模型鲁棒自适应控制算法。根据系统参数不确定性范围建立了多个辨识模型,在辨识模型中设计非线性鲁棒项,以抑制干扰、未建模动态等不确定非线性的影响,提高系统的鲁棒性。基于辨识模型设计相应的控制器,采用基于辨识误差的性能指标函数作为切换依据,选取最佳控制器作为当前控制器,解决了传统自适应控制对参数自适应初值敏感的问题。该方法能够克服不确定非线性和强参数不确定性的影响,使系统得到渐进跟踪的性能,提高系统的瞬态响应性能。实验结果表明,该算法能抑制建模不确定性的影响,系统期望跟踪指令幅值为10mm时,跟踪误差大约为0.038 6mm,相对跟踪误差约为0.386%,系统跟踪精度得到了提高。

An multiple model robust adaptive control algorithm is proposed for motion control of electro hydraulic servo systems subjected to both uncertain nonlinearities and strong parametric uncertainties. According to the range of parametric uncertainties, multiple identification models are designed and nonlinear robust items are added to the models to suppress the influence of uncertain nonlinearities and to improve the robustness of the system. The controllers are designed based on the identification models and then the performance index function designed based on the identification error is considered to select the optimal controller as the current controller. This method solves the problem that the traditional adaptive control is sensitive to the initial value. The proposed control algorithm can overcome the influence of uncertain nonlinearities and strong parametric uncertainties, guarantee an asymptotic output tracking performance, and improve the transient response performance of the system. Experimental results showed that the proposed algorithm can suppress the influence of modeling uncertainties. Its tracking error is 0. 038 6 mm when the amplitude of the desired tracking trajectory is 10 mm and the relative tracking error reaches 0. 386%, thus the system tracking accuracy is obviously improved.