Design and Optimization of Tubular Linear Permanent Magnet Generator with Performance Improvement Using Response Surface Methodology and Multi-Objective Genetic Algorithm

Abstract

Linear generators are electric machines which generate electrical energy from linear movement. Since these machines can lift gear wheel or power train, they have begun to be used widely nowadays. Since their working areas differ according to speed and power characteristic, this study contains design and optimization of tubular linear generator for free piston practices. The design performed response surface optimization through design variables acquired as a result of sizing via interface. The association between the determined design variables and the sizes of generator output was examined. In addition, these sizes were used for objective functions of increasing efficiency, decreasing overall volume and increasing general performance and their optimum values were found by using Multi-Objective Genetic Algorithm. Initial and optimum design data were compared with Ansys Maxwell 2D. With overall performance increase, 22,78% decrease was seen in total mass, while 11,7% decrease was seen in cost. In addition, prototype linear generator was made in line with initial geometry data and it was applied with crank slider mechanism.