混合偏最小二乘回归和粒子群优化直接进给轴加速度饱和性能的优化方法

Acceleration Saturation Performance Optimization Method of Direct Feed Drive Systems Using Hybrid PLS Regression and PSO

为了提高直线电机驱动进给轴的驱动性能,对进给轴的加速度饱和性能优化方法进行了研究。分析了直接进给轴驱动过程中加速度饱和性能的影响因素,提出一种混合偏最小二乘回归(PLS)与粒子群优化(PSO)的优化方法。采用正交试验设计方法分配伺服试验参数,利用激光干涉仪测量进给轴的运动时间和速度,从而构建样本库,建立基于PLS的加速度饱和性能模型,采用PSO算法辨识加速度饱和性能优化的伺服参数。搭建了直接进给轴试验平台,进行了200组样本的随机试验。结果表明,混合优化后的伺服参数具有较好的加速度饱和性能,比样本组的平均加速度饱和性能指标高68%。

For the purpose of enhancing the driving performances of linear motor feed drive systems, an optimization method of acceleration saturation was investigated for the systems. After the analyses of influence factors of acceleration saturation performances in direct feed drive processes,an optimization method was presented based on hybrid PLS regression and PSO.Orthogonal experimental design method was used to assign experimental servo parameters.The motion time and speeds of the feed drive systems were measured by laser interferometer to establish a sample database.An acceleration saturation performance model was founded based on PLS.The servo parameters for acceleration saturation performance optimization were identified by PSO.An experimental platform was set up and the random experiments with 200 group data were run.Results show that the hybrid servo parameters have better acceleration saturation performance which is 68% higher than the average values of the sample groups.