Structural Optimization of Jet Fish Pump Design Based on a Multi-Objective Genetic Algorithm

Jet fish pumps are efficient hydraulic machinery for transportation. Yet, the complex flow phenomenon in it is major potential risk damage to fish. The dangerous phenomena fish, such as radial pressure gradient and exposure strain rate, usually controlled by structural parameters of jet pumps. Therefore, injury rate can be theoretically decreased optimization design However, there a nonlinear relation between key parameters. To solve this problem, present paper established mapping parameters, based on computational fluid dynamics back-propagation neural network. According mapping, an NSGA-II multi-objective genetic algorithm was used optimize structure results showed that optimized could reduce internal gradient, danger zone 40%, 12.5% 50% pre-optimization level, respectively. pump significantly injuries keep efficiency at high level. provide certain reference relevant problems.