A many-objective optimization of industrial environmental management using NSGA-III: A case of China’s iron and steel industry

Abstract Under the restriction of multiple industrial environmental targets, the difficulty of industrial environmental management, as a many-objective optimization problem, has increased significantly. As traditional optimization methods such as bottom-up models and commonly used intelligent algorithms have drawbacks in solving many-objective optimization problems, we introduce the third edition of Non-dominated Sorting Genetic Algorithm (NSGA-III) to the environmental management problem in China’s iron and steel industry. We build a many-objective optimization model to plan the application of the four types of decision variables: process equipment, cleaner production technologies, end-of-pipe treatment technologies and synergic technologies. In total, 7 objectives including the minimization of energy consumption, 5 types of pollutant reduction and economic cost are considered. In addition, to formulate final decision schemes, we adopt the Fuzzy C-means Clustering Algorithm to cluster the Pareto-optimal solutions. The results show that NSGA-III performs well in center distance, spacing metric, and computational efficiency. The Pareto-optimal solutions reflect that SO2 reduction target, is too strict, while others, such as energy conservation and PM emission reduction are too loose. Besides, we obtain four final decision schemes based on different objective preferences. In sum, the proposed methodology is proved to be capable of solving many-objective optimization problems and helping decision making in industrial environmental management.