基于BP神经网络的设备故障率获取

Failure-rate Data Acquisition of Power Equipments Based on BP Neural Network

电力系统可靠性基本参数的获取靠长期的统计计算。对于新工程的可靠性评估,需要相应的可靠性基本参数,在缺乏此类参数的情况下,都只能靠盲目的借用其他工程已有的统计参数。文章在考虑系统中影响设备故障率的各种因数的基础上,提出了利用神经元模拟训练的方法,获取电力系统设备故障率参数。利用该方法,以现有可信的设备故障率为蓝本,在各因数对故障率影响相差不大的地区,获得对新工程可靠性评估需要的设备故障率。

Acquisition of the basic parameters of power system＇ s reliability depends on long - term statistic and calculation. The stability evaluation for a new project needs relative basic stability parameters. In the case of lacking such parameters it is obliged to use the existing statistical parameters of other projects indiscriminately. This paper gives a method to obtain failure - rate parameters of power system equipments using BP neural network, taking into account various factors that have influence on failure rate. This method can be used to obtain the failure - rate parameters of power equipments needed for stability evaluation of a new project on the basis of the existing credible failure - rate parameters of power equipments in the areas where difference in influence exerted by various factors on failure rate is not significant.