PID Self-tuning Method Based On Predictive Control And Information Entropy

Due to its advantages of simple algorithm and strong robustness,PID control is still the most widely used control algorithm in industrial processes.When the PID controller is running in the industrial field,due to changes in external conditions or changes in the controlled object itself,the control function will change,resulting in PID control parameters no longer applicable to the current controlled object,the control effect becomes worse,and need to be re Set PID control parameters.Therefore,this paper studies the online self-tuning method of PID control parameters,so that the PID controller has self-tuning function.When the dynamic characteristics of the controlled object change,it can automatically adjust the PID parameters online to adapt to the new working conditions.For single variable system,the closed-loop online recursive identification is implemented for the dynamic model of controlled process.When the perturbation signal is added on the output or setpoint of PID controller,or the controller parameters change,the identifiability proof of closed-loop identification is given.The model mismatch can be identified from the input and output data and the time to restart online identification can be found.The PID parameter self-tuning is proposed based on predictive control or information entropy respectively.Based on the state space model of the closed-loop PID control system,the weighted square sum of predicted deviation and control variation is taken as the objective function to optimize the PID control parameters online.Moreover,the information entropy of controlled variable and manipulated variable is selected as the objective function,the online optimization of PID control parameters is carried out to achieve the maximum order.Both of them adopt closed-loop online identification method to conduct feedback correction of process model parameters.For multivariable system,the tuning scheme of decentralized PID control loops for large-scale system is presented based on the closed-loop coupling degree.First,the concepts of closed-loop relative gain and coupling degree are put forward.It can be found that,the closed-loop relative gain under ideal control is equivalent to the open-loop relative gain;the feedback of closed control loop aggravates the coupling degree of the whole system.Second,the sequential tuning rules of PID control loops are proposed based on closed-loop coupling degree for large-scale system.According to the open-loop relative gain array of controlled process,the variable pairing and tuning sequence of PID control loops is determined.Then self-tuning of each PID control loop is implemented in sequence and the coupling degree is calculated for each PID control loop.When the coupling degree exceeds the threshold value,the self-tunings for all PID control loops will be finished.