Few‐sample multi‐objective optimisation of a double‐sided tubular machine with hybrid segmented permanent magnet

Double-sided tubular machine (DSTM) is very suitable for wave energy conversion but easily suffers from high thrust ripple. In order to get the minimum cogging force with maximum force, a new DSTM hybrid segmented permanent magnet array proposed and optimised by novel iterative few-sample multi-objective optimisation method. The method based on an Taguchi framework obtain optimal design only few samples. To solve low precision problem of method, surrogate-model algorithm that uses general regression neural network, speed-constrained particle swarm exponentially weighted moving average are embedded into this framework. result compared other alternative topologies methods, prototype manufactured testing experiment.