基于运动学正解和自适应逆控制的并联机构波形复现研究

项目名称： 基于运动学正解和自适应逆控制的并联机构波形复现研究

项目编号： No.51305066

项目类型： 青年科学基金项目

立项/批准年度： 2014

项目学科： 机械、仪表工业

项目作者： 代小林

作者单位： 电子科技大学

项目金额： 24万元

中文摘要： 液压驱动并联机构可用作重型车辆的道路模拟试验设备，基于运动学反解算法的铰点空间控制策略比较简单但控制精度不高、基于动力学模型的控制策略对模型参数比较敏感易出现不稳定、基于迭代补偿的波形复现算法只能做离线和开环试验故准备时间长且不能响应系统状态的变化。本项目以并联机构铰点空间控制策略为基础，采用高精度、高效运动学正解算法以实时获取试验台的姿态反馈信号，研究用非线性滤波器和变步长LMS算法实现并联机构模型的高精度辨识，在逆模型中引入偏值权以克服系统的漂移，并研究用变步长X-LMS算法进行自适应逆控制，实现一种在线和闭环的并联机构波形复现算法，具有重要的理论意义和实际应用价值。

中文关键词： 并联机构；波形复现；运动学正解；模型辨识；自适应逆控制

英文摘要： Hydraulically driven parallel mechanism can used as a kind of heavy vehicles road simulation equipment.Joint space control strategy based on inverse kinematcis is simple ,but has pool performace and accuracy. Dynamic model based control strategy is sensitive to the parameters and its stability is not good. Waveform replication based interave compensation can only do offline and open loop test, this method need long time to prepared and cannt deal with the state changes of the system. This project act joint control of the parallel platform as inner loop, and ues high precision and high efficiency forward kinematics to get the displacement and attitude of the simulator in real time, then a nonlinear filter and variable step size LMS algorith is disscussed to get the accurate model of the parallel test table, finally a bias coefficient is employed in the inverse model of the platform and variable step size X-LMS algorithm is uesed to achieve online and closed-loop waveform replication of the parallel mechanism. This project has important theoretical and practical value.

英文关键词： parallel mechanism；waveform replication；forward；model identification；adaptive inverse control