Brushless Dc Motor To Optimize The Design Of The Study

The brushless DC motor is a DC motor using electronic communication in place of mechanical communication, so it has excellent characteristic of DC motors, for example, good starting and adjustable speed performance. It simplifies the motor's structure and overcomes the friction disadvantages of ordinary DC motor. Because it doesn't use mechanical communication, it is extensively used in aviations, robots, digital machine tool and so on.Because of the high cost of permanent magnet, optimizing for brushless DC motor design is an inevitable choice, which is a comparatively mature and multiplicate technique after these year's developing. Each optimizing method has its strong point. On the point of view of math, optimizing for motor design is a process for extremum point for a nonlinear, restricted recessive function. Genetic Algorithms (G.A.), simulated anneal and Tabu are extensively applied in the field, in which, Genetic Algorithms (G.A.) is most general one.In the former study of the optimizing for motor, people often focused on cross operator and variation operator, which was based on the G.A.'s nature, so a lot of .mixed G.A. came forth. In this paper, exterior factor was introduced to make improve on G.A. which was cluster analysis.G.A. usually convergent fast at the beginning, but slow down when it reach about the extremum region. It will waste a lot of data which are around the extremum point if the generation number is quit so many, when people only take the last one generation data as their want. It will be a very useful way to do some analysis to these useful data with cluster analysis to find out which one or a few of them is up to the mustard. It not only can get the optimizing result satisfied with efficiency and cost require, but also apply the idea of data-mining on the optimizing for motor design.At the beginning of this thesis, theory and design process for brushless DC motor was introduced, then a small power brushless DC motor was designed and compared with the result of experiment. After that, the design result was optimized with the method of GA. combined with cluster analysis. Optimizing result was discussed.At last, the application perspectives of this mixed optimizing method and the future work direction are also described.