系统工程与电子技术
繫統工程與電子技術
계통공정여전자기술
SYSTEMS ENGINEERING AND ELECTRONICS
2014年
7期
1232-1237
,共6页
胡玉伟%马萍%杨明%陆凌云%王子才
鬍玉偉%馬萍%楊明%陸凌雲%王子纔
호옥위%마평%양명%륙릉운%왕자재
电磁发射%放电时序%多阶段优化策略%改进粒子群优化算法%脉冲成形网络
電磁髮射%放電時序%多階段優化策略%改進粒子群優化算法%脈遲成形網絡
전자발사%방전시서%다계단우화책략%개진입자군우화산법%맥충성형망락
electromagnetic launch (EML)%discharge sequence%multi-stage optimization strategy%im-proved particle swarm optimization (PSO)%pulsed forming network
针对电磁发射对电流波形的平稳性要求,提出了一种基于改进粒子群优化算法(particle swarm opti-mization,PSO)的电磁发射系统电源时序多阶段优化策略。根据脉冲电源时序放电特征,将发射过程分为多个阶段,确定各阶段的划分原则,建立了各阶段脉冲成形网络电源时序优化模型及整个发射过程的时序优化框架。将种群的平均信息引入到粒子速度更新过程中,提出了一种自适应动态调整惯性权重的 PSO。改进的 PSO 根据当前群体的进化状态自动调整惯性权重,使算法具有较强的动态适应性,提高了优化过程中的全局和局部搜索能力。考虑时序初始取值空间过大会导致收敛速度过慢,采用动态缩减搜索空间操作,在新阶段产生的电流波形不能改善时,提高时序取值区间的下限。最后将该优化策略用于某电磁发射系统电源时序的优化设计中,仿真结果表明放电电流曲线非常平稳,而且能够获得较高的出口速度。
針對電磁髮射對電流波形的平穩性要求,提齣瞭一種基于改進粒子群優化算法(particle swarm opti-mization,PSO)的電磁髮射繫統電源時序多階段優化策略。根據脈遲電源時序放電特徵,將髮射過程分為多箇階段,確定各階段的劃分原則,建立瞭各階段脈遲成形網絡電源時序優化模型及整箇髮射過程的時序優化框架。將種群的平均信息引入到粒子速度更新過程中,提齣瞭一種自適應動態調整慣性權重的 PSO。改進的 PSO 根據噹前群體的進化狀態自動調整慣性權重,使算法具有較彊的動態適應性,提高瞭優化過程中的全跼和跼部搜索能力。攷慮時序初始取值空間過大會導緻收斂速度過慢,採用動態縮減搜索空間操作,在新階段產生的電流波形不能改善時,提高時序取值區間的下限。最後將該優化策略用于某電磁髮射繫統電源時序的優化設計中,倣真結果錶明放電電流麯線非常平穩,而且能夠穫得較高的齣口速度。
침대전자발사대전류파형적평은성요구,제출료일충기우개진입자군우화산법(particle swarm opti-mization,PSO)적전자발사계통전원시서다계단우화책략。근거맥충전원시서방전특정,장발사과정분위다개계단,학정각계단적화분원칙,건립료각계단맥충성형망락전원시서우화모형급정개발사과정적시서우화광가。장충군적평균신식인입도입자속도경신과정중,제출료일충자괄응동태조정관성권중적 PSO。개진적 PSO 근거당전군체적진화상태자동조정관성권중,사산법구유교강적동태괄응성,제고료우화과정중적전국화국부수색능력。고필시서초시취치공간과대회도치수렴속도과만,채용동태축감수색공간조작,재신계단산생적전류파형불능개선시,제고시서취치구간적하한。최후장해우화책략용우모전자발사계통전원시서적우화설계중,방진결과표명방전전류곡선비상평은,이차능구획득교고적출구속도。
In view of the requirement for stable discharge current during electromagnetic launch (EML) process,a multi-stage optimization strategy of discharge sequence based on improved particle swarm optimi-zation (PSO)is proposed.According to the sequential discharge characteristics of pulsed forming network (PFN),the subdivision rules are given for dividing the EML process into multiple stages,and the optimiza-tion model is established.Then the whole optimization framework for the EML process is built.The average information of swarm is introduced into updating process of velocity,and an adaptive PSO with dynamically adjusting inertia weight is proposed.In this PSO,the inertia weight is adjusted according to the evolutionary state of current population with dynamical adaptability in the optimization process,which improves the glob-al and local search capability in every stage.To overcome the defect of slow convergence for unreasonable value space,the method of reducing search space is used which increases the lower limit of the value range when the current shape is not stable in the new stage.Finally the strategy is applied to optimize the discharge sequence of PFN for an EML system,the results show that the current shape is very stable and a higher exit velocity is obtained.