Improved distributed optimization algorithm and its application in energy saving of ethylene plant

This paper investigates an energy optimization problem for ethylene production process without system dynamics. Fundamentally different from the conventional centralized optimization strategies, the energy optimization in the ethylene production process is transformed into a distributed optimization problem with input constraints due to multiple interconnected units involved. Based on the combination of consensus mechanism and gradient tracking technique, a distributed algorithm with adaptive step-sizes is derived to optimize the coil outlet temperature (COT) and steam hydrocarbon ratio (SHR) in the thermal cracking process as well as the temperature and duty in the product separation process. Besides, considering the mechanism limitation, a projection operation is adopted to deal with input constraints. The experimental results show that the proposed distributed algorithm is superior to the centralized method in terms of computational efficiency and robustness with a faster convergence speed. (C) 2022 Elsevier Ltd. All rights reserved.

Automated summary

This paper investigates the energy optimization problem in the ethylene production process and proposes a distributed algorithm based on consensus mechanism and gradient tracking technique to solve this problem. Experimental results demonstrate the superiority of the proposed algorithm in terms of computational efficiency and robustness compared to traditional centralized methods.