液压伺服关节自适应模糊神经网络控制补偿方法  

Control compensation methods for hydraulic servo joint with adaptive-network-based fuzzy inference system

作　　者：朱兴龙[1] 周骥平[1] 

机构地区：[1]扬州大学机械工程学院,江苏扬州225009

出　　处：《控制理论与应用》2005年第5期694-698,共5页Control Theory & Applications

基　　金：江苏省高校自然科学基金资助项目(03KJB460161)

摘　　要：三自由度液压伺服关节在实现位置跟踪时存在跟踪误差,原因在于液压伺服关节是一个具有饱和、结构死区和强耦合的动力学系统.为了解决这一问题,建立了该关节的动力学模型.通过比较几种控制方法在该关节位置跟踪问题上存在的不足,提出了一种自适应模糊神经网络控制补偿方法.该方法采用样本训练自学习,自适应调整变增益系数.该方法不但消除了饱和、结构死区和强耦合引起的位置跟踪误差,而且解决了控制向量在大范围内变化实现准确位置跟踪.最后,通过仿真试验验证了该动力学系统是稳定的,提出的方法是可行的.Tracking errors occur during the position tracking because the three degrees of freedom （d.o.f.） hydraulic servo joint is a nonlinear dynamic system with saturation, dead band and stroke coupling. In order to solve this problem, the dynamic model of the joint is established. A control compensation method with adaptive-network-based fuzzy inference system （ANFIS） is presented by comparing the advantages and disadvantages of several control methods in position tracing problem. The method enables to adjust the alterable gain coefficients by the sample data sets training and self-learning. The position tracking errors, caused by saturation, dead band and stroke coupling, are eliminated. Simultaneously the accurate position tracking is implemented by the method even if the control vectors are fluctuating in a large scale. Eventually, the simulation results show that the dynamic system is steady and the method is feasible.

关 键 词：控制补偿方法 自适应模糊神经网络 液压伺服关节 机器人 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]