Font Size: a A A

Research On PID Parameters Tuning And Load Disturbance Compensation Control For AC Servo System

Posted on:2016-04-23Degree:MasterType:Thesis
Country:ChinaCandidate:X FuFull Text:PDF
GTID:2348330470984462Subject:Mechanical engineering
Abstract/Summary:PDF Full Text Request
AC servo system has been an important part of the aerospace, NC machine too l, food package,industrial robots and other industrial fields.Now,the contro l strategy o f AC servo system most ly uses the tradit ional PID which parameter s turning related to the contro l performance.Under the different working condit ions,the AC servo system can not put into operation if the contro l parameter is not set well. The manual tuning is difficult and needs to take a long t ime to set, this article aims at researching a new method based on system ident ificat ion and genet ic algorithm o f auto-tuning PID reguktor parameters which has important significance for saving the worker's time, improving the product ion efficiency and control performance. load torque disturbance also influences the performance of servo system,in order to so lve the problem,this paper presents a methods o f load torque compensation based on extended kalman filter.At first,this paper analyzed the mathemat ical model and vector contro l o f permanent magnet synchronous motor(PMSM),and then built the three closed-loop contro l model.The transfer funct ion o f vario us co mponents of current loop,velocit y loop and posit ion loop were established from interior to exterior. According to the performance requirements o f various contro l loops,the formula e o f contro l parameters tuning were analyzed which provide theoretical basis for the contro l parameters self-tuning o f AC servo system.Secondly,according to the contro l structure of various sectors o f the servo system,a self-tuning strategy based on system ident ification and genet ic algorithm o f velocit y loop contro l parameter was proposed,first ly, the feasibilit y o f closed-loop ident ification was analyzed,then the contro lled object o f velocit y loop was identi fied with least square method. At last,the auto-tuning of the velocit y loop contro l parameters were co mpleted with genet ic algorit hm. At the same time,the posit ion loop contro l parameter was setted based on the auto-setting of velocit y loop contro l parameters,and got the proport ionalit y coefficient.Thirdly, in order to so lve the load torque disturbance problem,a method o f load torque online ident ification and co mpensat ion based on reduced-order Extended Kalman Filter(EKF) was adopted,Acordding to PMSM mecha nical motio n equat ion,the load torque ident ificat ion mathemat ical model was established,and the result o f ident ification was imported to the input o f current regulator as contro l input signal o f beforehand co mpensat ion,so that the fo llow performance was pr ovided. Under the same PID parameters,the change of posit ion error was studied with or without the load torque co mpensat ion.The simulat ion and experimental results show that the reduced-order EKF can est imate the load torque effect ively,and the co mpensat ion can reduce the posit ion err obviously and improve the system abilit y o f resist ing load disturbance.Last ly,the sett ing strategy o f velocit y loop and posit ion loop contro l parameters and the method of load torque online id ent ificat ion and compensat ion were verified by experiment,the output response performance was co mpared under the condit io n o f the default contro l parameters and self-tuning parameters.The experiment results show that the system has smaller overshoot and faster response speed,the contro l performance was improved obviously under the self-tuning parameters.At the same time, the change o f posit ion error was co mpared with or without the load torque co mpensat ion,by experimental verification,it can co me to a conclusio n that the reduced-order EKF can est imate the load torque effectively,and the co mpensat ion can reduce the posit ion error obviously and improve the system abilit y o f resist ing load disturbance.
Keywords/Search Tags:AC Servo System Ident ificat ion, Genet ic Algorithm, Self-tuning, Load Torque Ident ificat ion, Dynamic Co mpensat ion
PDF Full Text Request
Related items