The feeder outage prediction model based on Optimization and Improvement of Random Forests Algorithm. Issue 6 (March 2020)

• Record Type: Journal Article
• Authors: Zhu, Jiajia; Zhou, Lanbo; Lu, Liming
• Is Part Of: IOP conference series. Volume 768:Issue 6(2020)
• Publication Date: 2020-03
• Subjects: Materials science -- Periodicals; 620.1105
• DOI: 10.1088/1757-899X/768/6/062029
• Languages: English
• ISSNs: 1757-8981

Abstract

Abstract: The accurate prediction of feeder monthly fault level is the key to distribution network operation and maintenance. Aiming at the problems of low prediction accuracy, a method based on improved random forest to predict feeder monthly fault level is presented. Compared with the similarity calculation method of original random forest sample, this one increases the measurement of leaf node path distance, and applies the improved sample similarity to classification problem. Through the experimental comparison in the fault database of distribution network, the improved method can achieve a better classification effect than the original one, thus proving the effectiveness of the improved method.