Dynamic optimization for SP of control loops using adaptive APC techniques

Background

Advanced Process Control (APC) is widely applied in the industrial chemical sector. Dynamic Matrix Control (DMC) is favored for its ability to handle complex optimization control problems involving multiple variables and constraints. However, changes in working conditions over long periods of industrial operation can lead to a mismatch between the predictive model and the actual system. To achieve adaptivity to working conditions, this paper proposes a setpoint (SP) dynamic optimization method for control loops based on adaptive APC techniques.

Methods

Firstly, an identification algorithm with Nussbaum gain is employed to update the system model in real-time. Secondly, the predictive model is adaptively updated through digital testing after changes in working conditions. Subsequently, the weight coefficients and constraints of the performance index are adaptively adjusted. Finally, this paper employs the Memory-GA-PSO (MGAPSO) algorithm to solve the performance index with varying constraints and dimensions with the goal of optimizing the SP trajectory efficiently.

Significant findings

The experimental results validated the accuracy of the identification algorithm and the effectiveness of the adaptive APC techniques. The adaptive APC techniques proposed in this paper achieve self-adaptation to working conditions, ensuring the long-term effectiveness of advanced control schemes and effectively avoiding model mismatch.