基于改进遗传算法的电磁轨道炮电源时序优化  

Optimization for Discharge Sequence of Pulsed Power Supply for Electromagnetic Railgun Based on Improved Genetic Algorithm

作　　者：马萍[1] 胡玉伟[1] 杨明[1] 刘志钊[1] 王子才[1] 

机构地区：[1]哈尔滨工业大学控制与仿真中心,哈尔滨150001

出　　处：《弹道学报》2014年第3期104-110,共7页Journal of Ballistics

基　　金：国家自然科学基金创新研究群体科学基金项目(61021002);中央高校基本科研业务费专项资金资助项目(HIT.NSRIF.2014036);重点实验室开放基金资助项目(HIT.KLOF.2013.081)

摘　　要：针对电磁轨道炮电流波形平稳性的特殊要求,提出了一种基于改进遗传算法的电源时序优化方法。根据脉冲电源时序放电的特点,将电磁轨道炮的工作过程分成相互关联的多个阶段,采用多阶段优化策略。各阶段运用引入了精英保留、自适应变异和交叉3种策略共同作用的改进遗传算法产生新种群,克服早熟及收敛速度慢的缺陷;优化过程中逐步动态更新缩小设计变量的搜索空间,以进一步提高搜索能力。以使用金属电枢的电磁轨道炮多个脉冲电源时序设计为例进行分析比较,结果表明,该方法不但增强了搜索能力,改善了求解的质量,而且电流波形非常平稳,提高了优化结果的实用性,是确定电磁轨道炮脉冲电源时序的一种有效方法。Aiming at the stable requirement of current waveform during launch, an optimization approach on the basis of improved genetic algorithm was proposed to design discharge sequence of pulsed power supplies. According to the discharge characteristics of powers with delay time, the launch process of electromagnetic railgun was divided into multi-stage, and the optimization of discharge sequence was performed. The strategies of elitist selection, adaptive crossover and adaptive mutation were adopted to overcome the defects of premature convergence and slow con- vergence in simple genetic algorithm. During the optimization process, the design variable space was dynamically reduced to further enhance the search capability. By the improved genetic algorithm, the optimization of discharge sequence was performed in each stage. In a case of railgun with solid armature supplied by multiple pulsed power supplies, the optimization results were obtained from different methods. The results show that the proposed approach has better search capabilities and gives better optimization solution, and moreover it increases the usability of optimization results for the realization of reliable current, which indicates the effectiveness for design discharge sequence of power.

关 键 词：电磁轨道炮 放电时序 脉冲电源 遗传算法 

分 类 号：TM153[电气工程—电工理论与新技术]