Temperature Control System Design And Optimization For A 4L Jacketed Reactor

Jacketed crystallization reactor is a commonly used reacting device in chemical crystallization process,for which temperature control is the main factor to influence the producing process.Because there is a long-time delay in the process,the traditional control methods hardly could achieve good control result,usually companied with temperature overshoot,delay in disturbance rejection and system instability.This thesis designs a temperature control system which can tackle above problems,proposes a crystallization reactor temperature response transfer function identification method based on step-response identification experiment,gives a control scheme based on predictor and combines predictor with iterative learning method,which utilizing batches operation to modified the set-point reference cure and obtain better tracking effect.This thesis designs a temperature control system which regard LabVIEW software as upper machine to monitor the process,PLC as lower machine to control executors,respectively.For heating process,PLC gives pulse-width modulation signal to solid-state relay,through the on/off state of the solid-state relay to control the power of immersion heater indirectly;for refrigerating process,PLC gives the frequency value to the frequency invertor,through V/F control method to adjust the compressor revolving speed.The PT100 measures the temperature of the solution inside of reactor with high accuracy and transfers it to upper machine for further computation.The LabVIEW program executes compiling the control algorithm,data communicating and data processing.The convenient human-computer interaction interface makes it easier for use's operation.For identifying the dynamic temperature response performance of 4L jacketed crystallization reactor,this thesis proposes a frequency domain step response model identification algorithm.By using damping factor to estimate the process frequency domain step-response and applying a 2-order integrating process plus time delay to fit the process,we can obtain a good identification result.Aiming at tackle the negative effect of long-time delay,this thesis proposes a simplified predicting control scheme based on the recent literature about generalized predicting control structure.This thesis adapts two stable filter to predict delay-free output response and gives a two-degree-of-freedom controller design method,which can obviously improve the control performance.Finally,for optimizing the control performance,this thesis combines simplified predicting control structure with iterative learning algorithm,utilizing batch-to-batch process data to modify the set-point reference curve,and then improved temperature tracking performance is obtained.