Optimization of runner and vane blade angle of an oscillating water column based on genetic algorithm and neural network

Abstract

Oscillating water column (OWC) converts ocean wave energy into electrical energy based on a reversible vane-runner-vane device. The blade profile is crucial and requires optimization design for a better performance. In this study, computational fluid dynamics (CFD) and genetic algorithm (GA) are used for optimization design. A weighted objective function is constructed by using multi-condition maximum efficiency ηmax and high efficiency range Rwid. Flow rate of maximum efficiency point φEmax is also monitored. GA helps to find the optimal 01 solution after 10 generation's iteration. Considering the time-consuming of GA in the local optimization, artificial neural network (ANN) is trained to assist the optimization. An optimal 02 solution with better performance was found by ANN and validated by CFD. The angle of runner and guide vane of optimal 02 is changed from 60° to 30°–55.739° and 35.082° with a flatter profile. ηmax is maintained at a high level over 43% and Rwid increases from 1.5 to 1.89. Compared with initial solution, the energy loss of optimal 02 is greatly reduced especially at high flow points by reducing the vortex scale. Widening of the high efficiency range enhances the ability of the unit to adapt to wave energy utilization.