An efficient scheme to optimize the superconducting levitation via genetic algorithm

Abstract

The superconducting levitation consisting of high-temperature superconductors (HTSs) and permanent magnet guideway (PMG) is deemed promising technique for the advancement of the maglev transit. To improve the cost-efficiency and thus reduce the investment of this superconducting levitation transit, the optimization of the PMG is the most critical issue of practical interest since it serves as the continuous rail to generate the magnetic field by the rare-earth magnets. By the use of a generalized vector potential within the quasistatic approximation as the state variable to mathematically describe the HTS as well as the surrounding medium, an efficient scheme for optimizing the superconducting levitation has been developed with the genetic algorithm as a strategy to perform the global search of the PMG. This scheme directly describes the HTS element without simplification of its intractable nonlinearity of constitutive law, which renders this study stand out from the existing efforts. The testing of the proposed scheme on a typical optimization of the superconducting levitation has proven its robustness and efficiency, i.e., the time cost is merely 3.6 hours with 3000 individuals evaluated on a moderate desktop. Taking a HTS over the Halbach-derived PMG as a practice, a set of case studies were carried out to understand how the working condition, geometrical and material characteristics of the HTS affect its maximum levitation force achievable at different constraints of the cross-section of the PMG. The findings attained by the case studies, being inaccessible from the experiments, are aimed to provide useful implications for the optimization of a superconducting levitation system for the transit and analogous purposes.