Multi-objective optimization design in a centrifugal pump volute based on an RBF neural network and genetic algorithm

The hydraulic and acoustic performance of centrifugal pumps are interrelated contradictory in structure design. An optimization design method a volute based on radial basis function (RBF) neural network genetic algorithm (GA) is proposed to solve this problem. efficiency total sound pressure level the pump used as objectives. diameter base circle, width installation angle tongue, height diffuser tube variables. Latin hypercube sampling (LHS) establish sample space, RBF build agent model between variables objectives, GA multi-objective optimization. For Pareto solution set obtained, two extremal individuals initial selected for comparative analysis under different working conditions. result demonstrates that rated condition, compared with individual, optimal individual increases by 3.79%; internal noise decreases 5.5%, external 2.3%.