Design Of Discrete-time Auto Disturbance Rejection Control For The Linear Servo System Based On Swarm Intelligence Parameter Optimization

The subject comes from the National Natural Science Fund project, the research aboutPermanent Magnet Linear Synchronous Motor (PMLSM) control problems in modernhigh-end CNC machine tool feeding system. For the problem of internal and externaldisturbances exist in linear servo system directly influence the performance issues,launched a study of the servo system controller. Mainly for the study of the ActiveDisturbance Rejection Control (ADRC) control technology, fuzzy control theory andswarm intelligence optimization algorithm study. The main research work is as follows.First, this paper discussed the purpose and significance of the research, and introducedthe PMLSM development situation, current research ADRC technology. Gives the basicstructure and operating principle of PMLSM, linear motor mathematical model is analyzed,and discussed the multiple disturbance factors that affect the operation of the linear motor.Second, for the issue of linear servo system has multiple disturbances. Using ADRCcontrol technology designed a first-order discrete ADRC speed and current double closedloop controller for linear servo system. Because of the first order ADRC compensationcapacity is restrained, we further designed the second-order discrete-time ADRC speedcontroller. Results from the simulation can be seen: the second-order discrete ADRC speedcontroller performance is better. But the parameters of ADRC are more difficult to control.Again, in order to overcome the difficulty of multi parameters not in tuning, bacterialforaging optimization (BFO) algorithm was introduced. Considering the slow convergenceof BFO algorithm and the good convergence of particle swarm optimization (PSO)algorithm, based the parameter tuning of ADRC a novel method named BFO-PSOalgorithm was proposed. The simulation is verified that through the BFO-PSO tuningADRC has good control effect.Finally, for further to improve the dynamic performance of the linear servo systemdesigned discrete speed fuzzy ADRC. To obtain fuzzy rule in fuzzy ADRC with expertknowledge and empirical knowledge needs numerous adjustments, and sometimes the fuzzy rule is not always an optimal one. A method based on PSO without expertknowledge or empirical knowledge is presented to generate fuzzy rule automatically.Results of the simulation experiments comparesion shows that the dynamic quality of thesystem based on the proposed method is more effective than fuzzy ADRC, and the trackingperformance is also improved.