多目标量子进化算法及在无刷直流电机优化设计中的应用

发布时间：2018-05-18 15:33

本文选题：量子进化算法 + 多目标优化　；参考：《沈阳工业大学》2017年硕士论文

【摘要】：永磁无刷直流电机目前已应用于航天、军事、工业自动化控制、机械、医疗器械、机床加工等行业。目前研究、开发、生产无刷直流电机已经成为一种新的趋势。大数据分析显示,若未来几年全部用高效电机取代内燃机,平均每年二氧化碳的排放量将减少数亿吨。因此设计高效率的电机符合国家节能减排的政策及节约型社会的要求。实数编码量子进化算法在优化前期种群中的个体随机性较大、对优化方向的指导能力弱,导致算法在前期寻优进展缓慢。针对这一问题,本文采用特征矩形编码方式,利用多种适应度的矩形基因叠加有效地提高了个体的灵活度。同时针对特征矩形编码,改进了算法的种群更新机制,提出了特征矩形实数编码量子进化算法。利用大量基准测试函数对该算法进行验证,对比结果表明,特征矩形实数编码量子进化算法即使在种群规模小、优化代数少的前提下依然快速收敛,结果优于其他算法。基于该算法的优良性能,本文进一步提出改进的多目标实数编码量子进化算法并利用多目标基准测试函数证明其有效性、先进性。电机优化这类问题一般具有非线性、多约束、变量离散的特点。设计方案既要满足国家标准、达到用户要求,又要使电机性能达到最优。本文将提出的多目标量子进化算法应用到无刷直流电机的优化问题中。以电机效率、电机重量为目标函数,以定子直径、气隙磁感应强度、电流密度等五个变量为优化变量确立无刷直流电机优化模型。利用罚函数法对约束条件进行处理,得到增广目标函数。优化后的设计方案在满足约束的前提下,提高了电机效率并降低了电机重量。
[Abstract]:Permanent magnet brushless DC motor has been used in aerospace, military, industrial automation control, machinery, medical equipment, machine tool processing and other industries. At present, the research, development and production of brushless DC motor have become a new trend. Big data analysis shows that if all efficient motors replace internal combustion engines in the next few years, the average annual carbon dioxide emission will be reduced by hundreds of millions of tons. Therefore, the design of high-efficiency motor meets the national policy of energy conservation and emission reduction and the requirements of a conservation-oriented society. The real coded quantum evolutionary algorithm has a large randomness in the pre-optimization population and a weak ability to guide the optimization direction, which leads to the slow progress of the algorithm in the early stage of optimization. In order to solve this problem, this paper adopts the method of feature rectangle coding, and makes use of multiple fitness rectangular gene superposition to improve the flexibility of individual effectively. At the same time, aiming at the feature rectangle coding, the population updating mechanism of the algorithm is improved, and the feature rectangle real number coding quantum evolutionary algorithm is proposed. A large number of benchmark functions are used to verify the algorithm. The comparison results show that the algorithm converges rapidly even if the population size is small and the optimization algebra is less. The results are superior to those of other algorithms. Based on the excellent performance of the proposed algorithm, an improved multi-objective real number coding quantum evolutionary algorithm is proposed and its effectiveness and progressiveness are proved by using the multi-objective benchmark function. The problems of motor optimization generally have the characteristics of nonlinear, multi-constraint and discrete variables. The design scheme not only meets the national standards and user requirements, but also makes the motor performance optimal. In this paper, the multi-objective quantum evolutionary algorithm is applied to the optimization of brushless DC motors. The optimization model of brushless DC motor is established with five variables, such as stator diameter, air gap magnetic induction intensity and current density, as the objective function of motor efficiency and motor weight. The penalty function method is used to deal with the constraint conditions and the augmented objective function is obtained. The optimized design can improve the motor efficiency and reduce the motor weight on the premise of satisfying the constraints.
【学位授予单位】：沈阳工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TM33

【参考文献】