Research And Implement On PID Controller Parameters Self-tuning Methods

PID(Proportional,Integral and Differential) controller is used widely in kinds of industry circumstance of its simple structure,easy implementation and strong robustness.And as an abroad control law,during a long period,PID control has not been knocked out from the emergence of different advanced algorithms,and on the contrary,it plays an major role in sundries of control technologies.So far,a variety of advanced PID controller parameter tuning methods has appeared,but in practical applications these advanced tuning methods did not produce perfect control as expected.This is mainly because the simplicity of its decision on the structure of the PID controller has limitations on quality control,and that makes a simple PID controller,with large time-delay or unstable objects,for the object of the control performances are not good;At the same time,PID controller can not meet different performance requirements on the track and curb interference.This article first describes the algorithm characteristics of PID control,the influences of control parameters for performance,and analyses PID controller's basic self-tuning algorithm; especially focuses on the basic principles for the parameters self-tuning algorithm of PID controller based on the relay feedback.According to the method put forward by the Sweden scholars K.J.Astrom,which observes the limit cycle oscillations of the controlled process under the relay feedback with Z-N formula,derive the parameters tuning formula of PID controller in detail.The module converts temperature into digital value,the host reads the converted value through serial communication,then processes the data in fuzzy method,gets the control result by querying fuzzy control table,and outputs the control result to DUT6000, directly controls the digital output,adjusts heated power through PWM method that means changing the width of output pulse in order to control temperature.The algorithm doesn't need to understand the object model,and it is easy to achieve in the industrial control system with a good control result.At the same time,for SISO model,the article proposes PID parameter tuning method based on intelligent methods.And it has done detailed descriptions for the methods of on-line modifying parameters based on neural networks and fuzzy algorithms,also the simulation results are given to understand its certain practical value clearly. Finally in the summarization and prospect part,the research and development problems are summarized as while as the future possible work for the improvements are looking forward to.