基于电流分配系数自适应修正的平面电机解耦控制

发布时间：2019-01-26 08:59

【摘要】：对于模型参数具有不确定性的平面电机运动系统,电流分配策略是引起其多自由度耦合行为的本质。但现有解耦控制的研究主要集中于各运动自由度控制算法的设计,没有从电流分配的物理层面解决耦合问题。论文针对平面电机三自由度解耦控制,建立了各线圈单元通电电流与三自由度位移输出之间的ARMA(auto regressive and moving average)模型,以实测输出与估计输出之间的残差平方最小为优化目标,提出了一种基于电流分配系数自适应修正的解耦控制算法,并给出了可用于实际控制的递推离散化方法。通过在线调节,使电流分配系数估计值逼近实际模型参数,从物理本质减小了多自由度耦合作用。采用轨迹跟踪实验,并与基于数据解耦控制方法进行对比,结果表明该方法提高了平面电机的定位与运动精度,验证了其对一类MIMO(multiple input multiple output)运动系统多自由度解耦的有效性。
[Abstract]:For planar motor motion system with uncertain model parameters, current allocation strategy is the essence of coupling behavior of multi-degree of freedom. However, the current research on decoupling control is mainly focused on the design of control algorithms for each motion degree of freedom, and the coupling problem is not solved from the physical level of current distribution. Aiming at the decoupling control of planar motor with three degrees of freedom, the ARMA (auto regressive and moving average) model between the electrified current of each coil unit and the output of the displacement of three degrees of freedom is established. The minimum residual square between the measured output and the estimated output is taken as the optimization objective. A decoupling control algorithm based on adaptive modification of current distribution coefficient is proposed, and a recursive discretization method for practical control is presented. Through on-line adjusting, the estimation of current distribution coefficient is close to the actual model parameters, and the coupling effect of multi-degree of freedom is reduced from the physical essence. The trajectory tracking experiment is used and compared with the data based decoupling control method. The results show that this method improves the positioning and motion accuracy of the planar motor. The effectiveness of the decoupling method for a class of MIMO (multiple input multiple output) motion systems with multiple degrees of freedom is verified.
【作者单位】：清华大学机械工程系摩擦学国家重点实验室;电子科技大学机械电子工程学院;
【基金】：国家科技重大专项(2012ZX02702006) 国家自然科学基金(51175296)资助项目
【分类号】：TM301.2

【参考文献】