Multi-Objective Parameter Optimization of Submersible Well Pumps Based on RBF Neural Network and Particle Swarm Optimization

In order to improve the hydraulic performance of a submersible well pump, steady and transient simulations were carried out based on ANSYS CFX software. The head and efficiency of the submersible well pump under standard operating conditions were taken as the optimization objectives, and the impeller outlet placement angle, outlet width, and vane wrap angle were selected as the optimization variables using the Plackett-Burman experimental design method. The RBF neural network training samples were constructed using the uniform experimental design method to build a hydraulic performance prediction model for the submersible well pump, and a multi-objective particle swarm optimization was used to solve the model and obtain the Pareto optimal solution set. Using the head and efficiency of the initial model as the boundary, the Pareto optimal solution and the corresponding structural parameters are sought. After the optimization, the head of the individual with the better head is increased by about 2.65 m, and the efficiency of the individual with the better efficiency is increased by about 2.3 percentage points compared with that of the initial model. The pressure gradient in the impeller flow channel is more obvious, the work capacity is significantly improved, the vortex area of the spatial guide vane is smaller, the flow line is more regular, and the pressure pulsation amplitude at the inlet and outlet of the impeller and the spatial guide vane is significantly reduced.