Comparison of different algorithms in typical process optimization of chemical industry
Jinze Li
Sinopec Petrochemical Science and Engineering Research Institute Co., Ltd.
Yi Zhao
Sinopec Petrochemical Science and Engineering Research Institute Co., Ltd.
Lei Zhang
Sinopec Petrochemical Science and Engineering Research Institute Co., Ltd.
DOI: https://doi.org/10.59429/esta.v12i2.10570
Keywords: Atmospheric and vacuum process; Heuristic algorithm; Optimization algorithm; Algorithm model
Abstract
Complex problems in chemical processes often need to be transformed into optimization problems for solution, and intelligent optimization algorithms provide efficient strategies for this. Based on the WilliamsOtto (WO) process and the crude oil atmospheric and vacuum distillation process, this study compares the performance of five optimization algorithms, including the Sequential Quadratic Programming (SQP) method, Particle Swarm Optimization (PSO) algorithm, etc., and evaluates them from dimensions such as running time, number of iterations, and the quality of results. The results show that the traditional gradient algorithm SQP is suitable for low-dimensional nonlinear problems, while heuristic algorithms exhibit better global search capabilities and robustness in the strongly coupled and multi-stable industrial process of crude oil atmospheric and vacuum distillation. It can be seen that different optimization algorithms have their own advantages and disadvantages, and a reasonable selection should be made according to specific problems.
References
[1]Huang Lingxiang, Ye Zhencheng, Shen Feifei, et al. Energy coupling modeling and global optimization of the demethanization system in the cold box[J]. Control Engineering, 2020, 27(11): 1873-1881.
[2]Hwang J, Roh M, Lee K. Determination of the optimal operating conditions of the dual mixed refrigerant cycle for the LNG FPSO topside liquefaction process[J]. Computers & Chemical Engineering, 2013, 49(11): 25-36.
[3]Mortazavi A, Alabdulkarem A, Hwang Y, et al. Novel combined cycle configurations for propane pre-cooled mixed refrigerant (APCI) natural gas liquefaction cycle[J]. Applied energy, 2014, 117(15): 76-86.
[4]Tian Peng. Research on the application of intelligent algorithms in dynamic optimization of chemical engineering[D]. Zhenjiang: Jiangsu University, 2020.
[5]Xu Yufei, Qian Feng, Yang Minglei, et al. An improved whale optimization algorithm and its application in parameter optimization of residue oil hydrotreating[J]. CIESC Journal, 2018, 69(03): 891-899.
[6]Osaba E, Yang X, Fister I, et al. A Discrete and Improved Bat Algorithm for solving a medical goods distribution problem with pharmacological waste collection[J]. Swarm and evolutionary computation, 2019, 44: 273-286.
[7]Bai Zonghan, Kang Qi, Wu Haihao, et al. An oil and gas field production optimization model based on sequential quadratic programming and machine learning algorithms[J]. Drilling & Production Technology, 2025, 48(01): 165-172.
[8]Wei H, Tang X S. A Genetic-Algorithm-Based Explicit Description of Object Contour and its Ability to Facilitate Recognition[J]. IEEE Trans Cybern, 2015, 45(11): 2558-2571.
[9]Ge Rui. Improvement and application of a multi-modal multi-objective particle swarm optimization algorithm[D]. Yangzhou: Yangzhou University, 2023.
[10]Xin Wenbin, Wang Zhongyuan, Gong Yuning, et al. Application of the simulated annealing algorithm in the inflow performance relationship of steam stimulation wells[J]. Petrochemical Industry Application, 2024, 43(7): 36-39.
[11] Xie Zhiwen, Wang Zheng, Wang Rui, et al. Research on path planning of substation inspection robots based on an improved ant colony algorithm[J]. Energy and Environmental Protection, 2021, 43(12): 212-216.
[12]Williams T J, Otto R E. A generalized chemical processing model for the investigation of computer control[J]. Transactions of the American Institute of Electrical Engineers. Part 1. Communication and electronics, 1960, 79(5): 458-473.
[13]Xiong Q, Jutan A. Continuous optimization using a dynamic simplex method[J]. Chemical Engineering Science, 2003, 58(16): 3817-3828.