Genetic algorithm based co-design of wave energy converter power take-off, floater geometry, and control

Abstract

This work presents the development of a novel control co-design model for the simultaneous optimisation of both the power take-off and floater of a heaving point absorber wave energy converter (WEC). Using genetic algorithms, both the internal geometry of a permanent magnet synchronous generator (number of pole pairs, rotor radius, etc.) and the scale of the floater are optimised for
a given location – using historical wave climate data – and a user-selected control strategy – impedance matching or linear damping. Controllers are tuned for each WEC and sea state pairing, accounting for the efficiency of the WEC’s drive train – as a function of force and speed – to maximise the electrical power output in those specific wave conditions. The model is tested for different control strategies finding that at the same location, the genome of the optimal WEC design varies by 22% . Using the same control strategy at different locations resulted in a similar change of 24% between sites in the North Sea and the Bay of Biscay.