Design Of Intelligent Temperature Control System For 10 Liter Crystallizer

Cooling crystallization is the most common method used in industrial crystallization.In cooling crystallization,the temperature in the crystallizer is an important factor affecting the crystallization process.Accurate temperature control is not only related to the quality of crystal products,but also guarantee repeatable production for batch crystallization.Because the temperature of a crystallizer is a typical integral response process with large time delay and inertia,it is difficult to use the traditional linear system control methods to realize accurate heating and cooling operation.In order to solve this problem,this thesis designs a temperature control device for a 10 liter crystallizer,which can realize accurate control of the temperature regulation process.The temperature control device designed in this thesis takes PC as the master computer,and the monitoring interface is configured based on the LabVIEW software platform.The PLC controller is used as the slave computer.For the heating operation,an immersion heater with power of 2kW is adjusted by the pulse width modulation(PWM)and the solid state relay.For the cooling operation,a compressor with the refrigeration power of 1.5P is adjusted by a frequency converter.A platinum thermocouple(PT100)is used to measure the solution temperature in a crystallizer.The solution temperature in a crystallizer is modified by the heat exchange between the circulating oil and the crystallizer.The heater and compressor are taken as the main actuator of the temperature control device to heat and refrigerate the circulating oil.In order to identify the dynamic characteristics of the temperature response of a crystallizer,a step response experiment is adopted,and an attenuation factor is used to make Laplace transform of the step response,such that the system frequency response can be effectively evaluated.Accordingly,an identification algorithm is established for fitting a integral transfer function model with time delay parameter.A two-degree-of-freedom(2DOF)internal model control(IMC)scheme is designed to overcome the problem of large overshoot and oscillation usually involved with the traditional PID control methods for controlling integral plants with large time delay and inertia,which can realize decoupling regulation of set-point tracking and disturbance rejection.Combined with the real engineering constraints,tuning methods of the 2DOF controller parameters are given,so that the set-point tracking performance and disturbance rejection performance of the system can meet the operation requirements,respectively.Further,an iterative learning controller is designed using the historical batch operation information.By selecting the controller parameters to satisfy the asymptotic convergence condition,the desired output trajectory can be entirely tracked,therefore obtaining better control effect.Based on the above temperature control device,experimental tests are performed to verify the proposed model identification and control algorithms.Step response identification tests are carried out for the heating and cooling devices,and correspondingly,the transfer function models of the temperature responses in the above crystallizer are established.The experimental results verify the effectiveness of the proposed identification algorithm.Then,based on the proposed 2DOF IMC scheme,heating and cooling control experiments are carried out.The experimental results verify that the proposed control method can achieve better set-point tracking performance and disturbance rejection performance.Finally,through batch operation experiments,it is demonstrated that the iterative learning algorithm given in this thesis can effectively optimize the control effect along the batch direction,and therefore realize batch operation optimization of the control system.