Design And Development Of The Teaching Experiment Platform For Smith Predictive Control Based On Air Heating Mixing Process

Smith predictive control is a kind of process control technology widely used in industry for large time-delay controlled objects.It is one of the key contents in the process control teaching of automation major,whose content involves many teaching knowledge points such as model,time-delay,parameter estimation,prediction,controller design and self-tuning.Existing teaching methods still depend on classroom teaching and simulation experiment based on MATLAB,emphasizing formula deduction and reasoning demonstration.In this paper,a new Smith predictive control teaching experiment platform based on air heating mixing process is designed and developed with the support of the National Natural Science Foundation of China.The platform is highly consistent with the actual industrial environment from the controlled object,detection instrument and actuator to the control system.At the same time,as a teaching experimental platform,the experimenter can arbitrarily configure the time delay of the controlled object,carry out off-line identification,on-line identification,Smith predictive control,self-tuning Smith predictive control and other physical experiments in a miniature industrial environment,and can also carry out comparative experiments with other conventional control methods.These contents can not only help students understand Smith's predictive control knowledge system in an all-round way,but also strengthen engineering training,shape engineering literacy and further improve classroom teaching effect.The main work of this paper is as follows:(1)On the basis of a detailed review of the development of the existing experimental teaching platform for automation,the problem of time delay in control and Smith's predictive control,the experimental functional requirements,platform functional requirements and human-computer interaction functional requirements of the experimental platform for Smith predictive control are analyzed in detail from the perspective of the objectives of the assistant teaching experiment.(2)The system architecture of Smith predictive control teaching experiment platform is designed,which consists of air heating mixing process,testing instrument and actuator,flexible switching intelligent channel system,multiple PLC/DCS control systems and model computer.The design of flexible switching intelligent channel system realizes that the process delay of any size can be configured and modified on the generalized controlled object sub-platform,around Smith predictive control has designed a relatively complete teaching experiment related function module,which can realize the functions of the controlled object loop matching,on-line time delay configuration and modification,free switching between different PLC/DCS control systems and I/O channel configuration.It can carry out off-line identification and online identification,Smith predictive control,self-tuning Smith predictive control and other teaching experiments.(3)Based on Siemens S7-300 PLC control system,STEP7,WinCC and MATLAB GUI software,Smith predictive control teaching experiment platform software based on air heating mixing process is developed.The software of flexible switching intelligent channel system and the sodtware of process control are developed by Siemens industrial control software,and the software of model computer is developed by using MATLAB GUI to realize the complex calculation functions of model identification,model conversion and controller design,such as polynomial operation and matrix calculation.The function module of MATLAB GUI in model computer realizes data communication with PLC control system through OPC protocol.(4)On this teaching experiment platform,a series of teaching experiments related to Smith predictive control are established.The results show that the platform has the advantages of highly simulating the industrial field environment,configurable time delay of controlled objects,intuitive,easy to understand,safe,reliable,convenient and flexible.At the same time,it helps students to understand the relevant knowledge system of Smith predictive control in an all-round way,and improves the teaching effect.