Research On Electical Power Of Meteorological Early Warning Algorithm Based On Decision Tree

In the global scope, many of electric equipment are attacked by adverse weather frequently, including strong winds, thunderstorms, high temperature and other horrible weather conditions, so that the electric power safety accidents occur frequently in the event of natural disasters at the same time that it will endanger human life and health.Based on the monthly safety operation of JingGu County from2012to2014in110K.V substation, thesis analyzes the practical significance of electric power meteorological early warning. And based on the classical ID3algorithm of decision tree algorithm, by in-depth research and implement the time efficiency of construction of decision tree, the ID3algorithm has been optimized. In order to obtain the predominant performance of electric power meteorological early warning algorithm, this paper presents a new method for describing information chaos, namely, average degree. Based on the average degree, the normalized average degree has been defined at the same level, for optimize the ID3algorithm, thesis using the calculation of normalized average degree instead of the calculation of the information gain for the selection criteria of the root node of decision tree. Finally, using the historical data of weather in Jinggu County and the data in the monthly safety operation statements of110KV substation in Jinggu County for data cleaning and integration to create the decision tree model of electric power meteorological early warning, and then the experimental analysis has been done, the results shows the performance of electric power meteorological early warning algorithm is better than ID3algorithm. In addition, in order to make the decision rules are easy to use, the paper defines a structure of XML tree which has infinite depth for implement the persistence of decision rules. Those research results partly shows that it is controllable that the electrical power unit affected by the weather in the future, and it can provide a kind of basis for power supply company and decision makers.