基于遗传神经网络的直线伺服系统定位误差补偿  

Positioning Error Compensation of Linear Servo System Based on Genetic Algorithm and Neural Network

作　　者：林健[1] 汪木兰[1] 李宏胜[1] 

机构地区：[1]南京工程学院自动化学院先进数控技术江苏省高校重点建设实验室,南京211167

出　　处：《组合机床与自动化加工技术》2011年第2期86-88,92,共4页Modular Machine Tool & Automatic Manufacturing Technique

基　　金：江苏省高校自然科学基础研究项目(08KJB460003);江苏省高校自然科学基础研究(10KJD460001);江苏省"六大人才高峰"高层次人才资助项目(2008163)

摘　　要：针对数控直线伺服系统的定位误差补偿,采用激光干涉仪测量工作台的定位误差,建立基于RBF算法的神经网络误差模型。提出遗传算法的训练方案优化RBF的网络参数,为了评价优化后RBF网络预测的精度,运用部分误差样本进行训练和仿真。构建了以DSP为核心的直线电机定位误差实验平台,根据误差校正值进行误差实时补偿实验。仿真和实验结果表明:经过遗传算法训练的神经网络模型对工作台的误差具有良好的学习能力和泛化能力,工作台定位精度显著提高。In view of the positioning error compensation of NC linear servo system,the positioning errors were measured by the laser interferometer,the neural network error model is set up by Radial Basis Function（RBF） algorithm.The training method of the genetic algorithm is also proposed to optimize the network parameters,in order to evaluate the accuracy of RBF network prediction method,part of the error samples are used to train and simulate.A DSP-core linear motor positioning error experimental platform was built,the error compensation experiments are conducted.The simulation and experimental results indicate that the RBF neural network error model trained by the genetic algorithm has a good learning ability and a generalization ability,the positioning accuracy is improved significantly.

关 键 词：直线伺服系统 定位误差 径向基函数神经网络 遗传算法 

分 类 号：TM303.1[电气工程—电机]