Intelligent algorithm based multi-objective optimization of impeller parameters for short-term axial flow ventricular assisted pump

This study aimed to optimize a percutaneous interventional axial blood pump by considering hemolysis, coagulation indices, and hydraulic performance as objective functions. Key impeller parameters were selected, and a sensitivity analysis identified the three most influential factors: shroud-side chord height (L1), hub-side chord height (L2), and hub-side wrap angle (θ3). A Radial Basis Function neural network model was constructed using Latin Hypercube Sampling data, and the optimal solution was determined through the Non-dominated Sorting Genetic Algorithm-II. The primary thrombogenesis risks are concentrated in the narrow blade-tip clearance and the mid-blade pressure surface on the impeller rim, where high scalar shear stress and prolonged blood residence time persist. After optimization, an overall reduction in turbulent kinetic energy was achieved. The hydraulic performance (Head increased by 9.6%), hemolysis (NIH decreased by 7.3%), and coagulation indicators (Vr decreased by 72.7%) of the blood pump all showed improvement.