Dynamic Optimization for SP of Control Loops Based on Both Rapid Intelligent Optimization and UniSim/EPKS Environments

Advanced Process Control is frequently applied in industrial chemical processes, where Dynamic Matrix Control is widely employed for its capability to handle complex optimization control problems with multiple variables and constraints. However, during prolonged operation of industrial processes, factors such as changes in operating conditions and external environmental variations may lead to a mismatch between the predictive model and the actual plant in DMC. To address the Model-plant Mismatch, this paper proposes a dynamic trajectory rapid intelligent optimization method based on online reconstruction of the predictive model. Constructing the system model based on the known linear dynamic model of the control loop and PID parameters, the new dynamic matrix is obtained using the closed-loop step response of the system model. Then, updating the prediction model online and simultaneously updating the optimization-constrained conditions in real time ensures the long-term effectiveness of the advanced control scheme. To rapidly solve the optimization objective, this paper utilizes an improved GA-PSO algorithm to compute the Set Point trajectory of the control loop. Finally, this study conducted simulations by integrating Honeywell's UniSim Design simulation platform with the EPKS DCS control system, verifying the effectiveness of dynamic matrix online updating and demonstrating the long-term effectiveness of the advanced control scheme.