基于BP神经网络PID的黑体温控系统  

BLACKBODY TEMPERATURE CONTROL SYSTEM BASED ON BP NEURAL NETWORK PID

作　　者：王志刚[1,2] 韩愈 王景辉[3] Wang Zhigang;Han Yu;Wang Jinghui(Tianjin Key Laboratory for Advanced Mechatronic System Design and Intelligent Control,Tianjin University of Technology,Tianjin 300384,China;National Demonstration Center for Experimental Mechanical and Electrical Engineering Education,Tianjin University of Technology,Tianjin 300384,China;Division of Thermophysics and Process Measurements,National Institute of Metrology of China,Beijing 100013,China)

机构地区：[1]天津理工大学天津市先进机电系统设计与智能控制重点实验室,天津300384 [2]天津理工大学机电工程国家级实验教学示范中心,天津300384 [3]中国计量科学研究院热工所,北京100013

出　　处：《计算机应用与软件》2022年第11期73-77,159,共6页Computer Applications and Software

基　　金：天津市优秀科技特派员项目(19JCTPJC59300)。

摘　　要：在辐射测温研究中黑体辐射源作为其中的重要组成部分,其温度控制过程伴随着温变性、非线性、时滞等特性。为了提高控温品质,设计一种基于BP神经网络PID的镓点面源黑体温控系统。对黑体辐射源进行建模,分别将BP神经网络PID算法和传统PID算法通过MATLAB/Simulink软件进行比较分析。实验仿真可知BP神经网络PID算法具有精度高、振荡少等优势。通过LabVIEW软件编写BP神经网络PID算法控制程序;利用Compact RIO温度控制器控制并采集镓点面源黑体的温度。实验结果分析,与常规PID控制相比,基于BP神经网络PID算法的控制性能得到明显改善,能够提高黑体辐射源的控温品质。In the study of radiation temperature measurement,the blackbody radiation source is an important part,and its temperature control process is accompanied by temperature change,nonlinearity and time delay.In order to improve the precision of temperature control,a gallium point surface source blackbody temperature control system based on BP neural network PID is designed.The mathematical model of blackbody radiation source was established,the BP neural network PID algorithm and traditional PID algorithm were compared and analyzed respectively through MATLAB/Simulink software.Experimental simulation shows that the BP neural network PID algorithm has the advantages of high precision and less oscillation.The control program of BP neural network PID algorithm was realized by LabVIEW software.The Compact RIO temperature controller was used to control and collect the temperature of gallium point plane blackbody.The experimental results show that compared with conventional PID control,the control performance of PID algorithm based on BP neural network is obviously improved,which can improve the temperature control quality of blackbody radiation source.

关 键 词：神经网络 PID算法 面源黑体 温控系统 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术] TP273