Research on the Effect of Parameter Self-Tuning on the Accuracy of PCR Temperature Control Based on Improved BP Neural Network PID Controller
In the PCR process,temperature control largely determines the final experimental accuracy and results.As the most important PID temperature control in process control,it adopts the control strategy of proportion,integral and differen-tial links.PID controller is widely used in industrial control field because of its simple algorithm,good reliability and high control robustness.The proportional link provides a fast response by reflecting the size of the error;The integral link ensures the steady-state accuracy of the system by eliminating the static difference.The differential component can suppress overharmonization and improve the dynamic performance of the system by predicting the variation trend of error.PID controller plays an irreplaceable role in the field of temperature control,so that the deviation between the temperature set value and the actual temperature value can be effectively controlled in an acceptable range,thus ensuring the requirements of temperature stability and accuracy in industrial production.In PID control,how to control the three parameters efficiently is very important for the accuracy of PID control.The traditional PID control algorithm can achieve good steady-state perfor-mance,but poor dynamic performance.Fuzzy control has excellent dynamic performance,but poor steady-state performance.In order to solve this problem,the three parameters of traditional temperature control PID algorithm are optimized and adjusted by neural network.In order to realize the application of low cost miniaturization platform,we successfully deployed the optimized BP neural network algorithm on STM32F407IGT6 single chip microcomputer,and used to self-tune the PID algorithm parameters.Through the temperature control detection on the built experimental platform,we have reached the national design standards and achieved a temperature control accuracy of 0.2℃.
万方数据
维普
