Design And Implementation Of Defects Forecasting Of Power Transformer Based On Multidimensional Power Data

As the development of power system, the security issues of power transmission and transformation equipment become more and more valued.Power transformer is one of the most important equipment of the power system.The defects of it not only seriously affect the daily life of residents, but also may cause huge losses to social and economic. At present, some approaches have been made to forecast the defects of power transformer by using data mining technology. Most of them are limited to finding the defects caused by the quality of power transformer,or make the forecasting macroscopically which does not suit for the individual. As the development of electric power technology, the real-time monitoring systems have been integrated in power system. These systems make the size and dimension of power data larger in recent years, and they also bring us a new idea of solving the defect forecasting of power transformer.Generally, the defects of power transformer have a great relevance with its quality and working environment. The impact caused by power transformer’s working environment is formed year after year. It could be seen as a time series in data mining process. Therefore, this paper designs and implements an approach to make prediction on power transformer’s defects.This approach divides the multidimensional power data into time-related data and non-time-related data. It builds model on the time-related power data by using time series forecasting approach. Then it combines the output of time series model, non-time-related power data and defect-related data as a new data set, which is used to build model of power transformer’s defects forecasting by using logistic regression. The prototype system works in a way that includes offline modeling and online prediction. It implements data preprocessing,modeling and online forecasting. This paper provides reference for defect forecasting of power transformer.First of all, this paper demonstrates the background and significance of power transformer defect forecasting. And the related technologies are discussed with advantages and disadvantages. Secondly, the requirements of defects forecasting to power transformer based on multidimensional power data are analyzed by declaring the application scenarios and overall goals of this prototype system. Then, this paper designs an approach which is used to predict the defects of power transformer based on multidimensional power data,including data preprocessing, time series prediction method and logic regression prediction methods. Next, this paper designs and implements the power transformer defect forecasting prototype system, including its architecture and the interactive interfaces of core modules. Last but not least,conduct unit testing and integration testing on this prototype system. The results of testing show that the proposed approach has availability.