Structural Optimization of an Electromagnetic Actuator Based on Genetic Algorithm, Greedy Search and Their Combination

In the last years, optimal operation and usage of electrical machines leads to an increased attention for improving the operating behaviour and changes in the electrical machine's structure for industrial applications. In conventional optimization, it is common to use parametric design and geometric optimization, but this manuscript presents the structural or topological optimization. This method gives more freedom to the designers to adapt high performance electrical machines to the customer goals. The implementation of the proposed method for structural design and topology optimization is applied to the detailed design of an electromagnetic actuator. In general, electrical machinery is aimed to produce high torque and power at minimum weight to gain a high torque and power density. Therefore, high force or torque and low weight are the two main goals in this work for designing electrical machines, which can satisfy mechanical, structural and magnetic constraints. To show the validity and the opportunities of the proposed optimization method, the design of a simple magnetic actuator using a metaheuristic global optimization method (genetic algorithm (GA)) and a deterministic local search (greedy algorithm) is investigated at first and, secondly, a combination of both of these methods is presented for a highly nonlinear problem. The given design goals have been successfully achieved using the proposed structural optimization method to find the best suited topology.