Distribution transformer failure rate prediction model based on multi-source information

For present distribution transformer failure rate prediction model, load and weather forecast information in the future have been not considered. In this paper, distribution transformer failure rate prediction model with parameters is proposed. Firstly, the characteristics of several common failure rate models are analyzed. The failure rate model combined the health index and operating time is established. The integrated correction factor by calculating the appropriate index system is used to modify the health index using guide method. It is the health index correction model of the initial point. Then load and weather forecast information is taken into account to describe the change trend of the health index. At last, it is easy to predict the distribution transformer failure rate by take the health index forecast trend into the failure rate model. The example of 5 kV/100 KVA distribution transformer failure rate prediction has shown that the improved model is based on comprehensive consideration of the future failure rate influence factors and it can improve the accuracy of the predicted failure rate effectively. In addition, the predicted result is more close to the actual transformer failure rate curve.