摘要
针对难以精确实现对永磁真空断路器合分闸过程动态特性控制,从而导致影响其机械与电气寿命的问题,提出了一种模糊神经网络全闭环智能调控方法.在分析永磁真空断路器合分闸过程硬件结构设计模型的基础上,建立了运动机构的实时位移x和输出量PWM占空比的隶属度函数与模糊控制规则.构造了1-3-1的三层BP神经网络结构,建立了合分闸过程BP神经网络数学模型,结合Matlab仿真软件训练了模糊控制规则,训练结果能较好满足系统合分闸过程精确控制要求.实验数据对比表明,相较于全导通控制策略和模糊控制策略,模糊神经网络控制策略能更有效地实现对永磁真空断路器合分闸过程的动态控制,提高了永磁真空断路器的机械与电气寿命.
It is difficult to realize precise dynamic control of opening and closing processes of the permanent magnet （PM） vacuum circuit breaker, which influences its mechanical and electrical life- time. A fuzzy neural network full closed-loop intelligent control method is proposed to solve this problem. Based on the analysis of the hardware structure design model for the opening and closing processes of the PM vacuum circuit breaker, membership function between real-time displacement and output PWM （pulse width modulation） duty ratio and fuzzy control rules are established. A 1-3- 1 three-layer BP（ back propagation） neural network structure and the BP neural network mathemati- cal model of the opening and closing processes are constructed. Matlab simulation software is em- ployed to train and obtain fuzzy control rules, and the training results can well meet the precise con- trol requirements during the opening and closing processes. The test data shows that, compared with fully-conducting control strategy and fuzzy control strategy, the fuzzy neural network control strategy can more efficiently realize the dynamic control of opening and closing processes of the PM vacuum circuit breaker, and further improve the mechanical and electrical lifespan of the PM vacuum circuit breaker.
出处
《东南大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2015年第4期683-689,共7页
Journal of Southeast University：Natural Science Edition
基金
国家自然科学基金资助项目(51307011)
安徽省自然科学基金资助项目(1308085QE97)
江苏省博士后科研资助计划资助项目(1301173C)
安徽省高校省级优秀人才基金资助项目(2013SQRL093ZD)
滁州学院科研资助项目(2012kj009B)
关键词
断路器
模糊神经网络
全闭环调控
动态特性
circuit breaker
fuzzy neural network
full closed-loop control
dynamic characteristic
