Dynamic Systems Simulation Development And Parameters Tuning For Model Predictive Control

The identification model is commonly used in the model predictive control system's simulation at present. Due to its disablity of emulating the nonlinearity of the actual system and the non-real outcome of system affected by the parameters-tuning, we bring out an OPC-interfaced dynamic simulation system based on the first principal modeling, also analyze the effects and rules of Parameters Tuning for Model Predictive Control.The main content of the article and its innovations are as following:1. The dynamic simulatin system based on the non-linear model is developed by using the S-Funciton in MATLAB/Simulink, which obeyed the first principle. In order to do some research with dynamic models on the steady state, we use system identification and linearization ways respectively.2. The parameters tuning of Dynamic Matrix Control (DMC) is researched by using the simulation method with the dynamic simulation system. The performace of setpoint tracking and load disturbance is analysed under both model-match and model-mismatch situations, then summarize the effects and rules of parameters tuning. Finally, we make the conclusion that the DMC performance of setpoint tracking and load disturbace are independent.3. There are several factors affecting the weighting factor selection, which are unit differences, gain relation, cost window sizes and dynamic behavior of the plant. The weighting normalization is used to solve this problem, which can simplify the weighting selection.4. The OPC-interfaced dynamic simulation system is developed, by using the OPC interface, the dynamic simulation system can communicate with the origianl predictive control software FrontApc. That makes the system more useful and features enhanced.