Parallel Self Optimizing Control Framework for Digital Twin Enabled Smart Control Engineering

This paper presents a parallel Self Optimizing Control (SOC) framework that combines parallel intelligence, Digital Twin, and derivative-free optimization to enable smart capabilities in classic process control. The parallel SOC framework supports the interaction between the physical system and its Digital Twin via Simultaneous Perturbation Stochastic Approximation (SPSA) derivative-free optimization algorithm. The framework is tested using the Digital Twin of a thermoelectric heating system. Obtained results show that using parallel intelligence with Digital Twin and SPSA optimization method, the SOC can improve the system performance by introducing a developmental behavior on the classic process control system.