垂直气动人工肌肉系统的模型参考自适应逆补偿控制

Adaptive inverse compensation control of the vertical pneumatic artificial muscle system

中文关键词  软体机器人  垂直气动人工肌肉系统  迟滞  逆补偿控制  自适应控制系统

英文关键词  soft robotics  vertical pneumatic artificial muscle system  hysteresis  inverse compensation control  adaptive control system

中文摘要

      与传统刚性驱动系统相比, 气动人工肌肉系统具有质量轻、人机交互友好等优势, 近年来在生产生活中受到广泛关注. 然而, 该类系统的运动呈现出复杂的迟滞特性, 这给针对该类系统的跟踪控制研究带来了挑战. 本文针对垂直气动人工肌肉系统, 提出一种模型参考自适应逆补偿控制策略, 可有效克服迟滞特性以及控制过程中外界扰动和参数摄动等不确定因素对系统运动状态的影响, 实现系统高精度跟踪控制. 具体而言, 本文首先对系统的运动特性以及影响系统控制精度的不确定因素进行分析; 然后, 基于分析结果建立一个描述系统运动特性的参考模型; 进而采用逆补偿思想, 通过对所建立的参考模型求逆来构造一个逆补偿控制器, 克服迟滞特性对系统运动状态产生的影响; 随后, 基于最小均方误差算法设计自适应律, 在线辨识参考模型的权值, 同时估计逆补偿控制器的设计参数, 克服外界扰动和参数摄动等不确定因素对系统运动状态的影响; 最后, 通过实验验证所提控制策略的有效性.

英文摘要

      Compared with traditional rigid actuating systems, the pneumatic artificial muscle systems have the advantages of lightweight and friendly human-computer interaction. However, the motion of this system appears complex hysteresis characteristic, which brings large challenges in realizing tracking control of the system. This paper proposes a model reference adaptive inverse compensation control strategy for the vertical pneumatic artificial muscle system. This control strategy can effectively overcome the influence of the hysteresis characteristic, external disturbances as well as parameter perturbation on the system’s motion states, and can further realize high-precision tracking control for the vertical pneumatic artificial muscle system. Specifically, we first analyze the system’s motion characteristics and the uncertain factors that affect the control accuracy of the system. Then, based on the analysis results, we establish a reference model to describe the system’s motion characteristics. According to the idea of inverse compensation, we design an inverse compensation controller by inverting the reference model established to overcome the hysteresis characteristic’s influence on the system’s motion states. Based on the least mean square error algorithm, we design the adaptive law to identify the weights of the reference model and to estimate the design parameters of the inverse compensation controller, thereby overcoming the influence of external disturbances as well as parameter perturbation on the system’s motion states. Finally, the effectiveness of the proposed strategy is verified by some experiments.