摘要
为了能够准确地判断出充放电机故障类型,找出其故障原因,提出利用BP神经网络识别故障类型的思路,并对此进行了分析和研究。获取故障类型所对应的特征参数;对特征参数归一化处理,并用BP网络对其进行训练。运用MATLAB软件进行仿真。实验结果表明,建立的BP神经网络对检测的效果准确可靠,可以用来模拟人工对充放电机故障做出正确的判断,同时也验证了BP网络的优越性和可行性。
In order to be able to accurately judge the charge and discharge machine fault type and identify the reasons for its failure,this paper proposes the idea of using the BP neural network to identify fault types and carries out the analysis and research. First of all,the fault type of the characteristic para-meters is acquired;secondly,the characteristic parameters are normalized and trained using BP network. Finally,the MATLAB software is used for simulation. The experiment results show that the BP neural network established has the accurate and reliable detection effects and it can be used to simulate the manual judgment for charging and discharging machine failures and the paper also verifies superiority and feasibility of the BP network.
出处
《电网与清洁能源》
北大核心
2015年第10期1-3,共3页
Power System and Clean Energy
基金
国家863项目(2011AA05A110)~~
关键词
电动汽车
充放电机
故障诊断:BP神经网络
electric vehicle
charging and discharging machine
fault diagnosis
BP neural network
