【摘要】 该文将断路器机构动力学特性仿真分析和基于人工神经网络(ANN)的状态识别算法相结合,为高压断路器的状态在线检测方法的研究提供了一条新思路。利用所建立的VS1型真空断路器机构动力学模型对故障状态下的断路器机构动力学特性进行了仿真分析,分析结果表明:不同状态下的可检测参量主轴转角具有明显不同的特征。之后对主轴转角进行了参数化描述,从而为断路器的状态识别奠定了基础。最后,引入可信度的概念,提出一种基于ANN的断路器机械状态识别算法,该算法不仅能够识别已知的状态类型,而且具有新状态类型的识别功能。更多还原

【Abstract】 A new research method was proposed forhigh-voltage circuit breakers condition monitoring, whichcombined the mechanism dynamic features simulation andcondition recognition arithmetic based on artificial neuralnetwork. VS1 circuit breakers mechanism dynamic features in fault were simulated with the dynamics model, which was built by ourselves. The simulation results indicated that theparameter which can be monitored—main angle, has different character. The main angle was described by several parameters, which provided theory support for circuit breakers condition recognition. The concept of reliability was introduced, and anew condition recognition arithmetic based on ANN for circuit breakers was proposed. The arithmetic can recognize not only the known condition but also the new condition.更多还原