Multi-objective optimization design of a sewage pump based on non-dominated sorting genetic algorithm III

Abstract

Accumulation of sanitary refuse, such as flexible cloth-like structures or the so-called rags, inflows through sewage pumps are prone to tangling, ultimately leading to clogging and wear. To prevent this, the ability of sewage pumps to handle wet wipes, rags, and similar flexible materials is a key feature that must be considered as the pumps are designed. Therefore, this paper proposed a multi-objective optimization strategy based on the fluid–structure interaction simulation, Support Vector Regression (SVR), and non-dominated sorting genetic algorithm III (NSGA-III). First, the values of the optimization objectives were obtained by a Coupled Computational Fluid Dynamics-Discrete Element Method (CFD-DEM) approach. Then, SVR was utilized to establish an approximate model between the design variables and the optimization objectives. The NSGA-III was applied to search the Pareto front. Finally, the improved impeller model was selected by adopting technique for order preference by similarity to an ideal solution (TOPSIS) with entropy weight. The results show that the multi-objective optimization method is suitable for the optimization design of sewage pumps. Comparing the numerical calculations of the original pump and the optimized pump, the results show that the optimized head and efficiency increased by 9.7% and 7.13%, respectively. The optimized pump improves the passage rate of the rag and effectively improves the clogging behavior. The wear amount of the optimized pump is significantly reduced by 32.54%.