SIMULATION AND ANALYSIS
A general comparative study of Conventional PID, Model predictive control through servo and regulatory response of CSTR model will be analyzed using MATLAB. Various temperature sensors have been installed to measure the temperature of the reactor (CSTR) and the opening and closing of the PSV (Pressure control valve) is done as per the control algorithm. The equation (1) given below, governs the simulation using the transfer function model of the CSTR plant.
MATLAB is used for tuning the PID parameters rather than manual tuning as it offers better results and accurate results. The table below shows the various PID parameters.
The transfer function is simulated for the servo response and compared with the servo tracing ability. The set points are varied in negative steps so as to generate the servo response of the system
Figure 4 shows the servo responses for a particular set of MPC controller points.
Figure () shows the output of the conventional PID an MPC in servo responses. The transfer function model is used to simulate the Control response for PID controller and MPC controller is simulated by using predetermined values of Control horizon as 0.5 units and 2 intervals along with a prediction horizon of 10 intervals.
Table 1 shows the time domain response of the PID and MPC controller. It is clearly evident that the MPC controller is better than PID controller in terms of settling time and time domain response. Similar results were also observed for overshoot. From all the results obtained MPC controller produced stable and better response.
The CSTR process is a highly nonlinear process. The modelling of the CSTR process is identified and implemented, the model is found out by empirically determining the system deriving from the data the real process. The MPC controller and PID controller are implemented to control the temperature inside the reactor. The simulations are implemented to track the servo response and regulatory response. The simulation results prove that the MPC control method is an easy tuning and more effective way to enhance the stability of time-domain performance of the temperature of CSTR process. The simulation results demonstrate the capability of the proposed identification strategy to effectively identify the compact and approximate accurate model for the CSTR process. It is shown that the proposed model predictive control design yields better improvement with significantly better response time for both servo and regulatory control objectives.