Power line communication power consumption data analysis method and system based on clustering analysis

Abstract

The invention provides a power line communication power consumption data analysis method and system based on clustering analysis. The power consumption data analysis is realized based on a power consumption information acquisition system. The data acquisition is conducted continuously according to a preset acquisition time, including a power consumption data acquisition process and a power consumption data feature extraction and saving process. For a current acquisition time, abnormal data discrimination is performed in real time, including a power consumption data reading process at the current acquisition time and an abnormal data discrimination process based on K-Means clustering analysis. The invention proposes the technical scheme of abnormal power consumption data identification based on the clustering validity for the first time by using a power line as a communication carrier to request and transmit the power consumption data, so as to achieve the purpose of improving the operation reliability of the power consumption information acquisition system.

Description

technical field [0001] The invention belongs to the field of electrical secondary systems, in particular to a method and system for analyzing electricity consumption data of power line communication based on cluster analysis. Background technique [0002] The National Power Grid Corporation vigorously promotes the coverage of the electricity consumption information collection system. Through the data communication between the main station and the field terminal, functions such as data collection, cost control management, line loss analysis, remote meter reading, and load monitoring management have been realized. see figure 1 The power consumption information collection system includes a server, a concentrator, a smart energy meter and a power line. Each energy meter is connected to the concentrator through the power line, and the concentrator is connected to a remote server through Ethernet, which can be provided by a network cable or a router. Wireless data transfer connec...

AI Technical Summary

Problems solved by technology

[0003] For abnormal data mining, the existing method based on artificially setting the threshold is simple but highly subjective; the method based on statistics has human factors to set the error identification interval; the method based on neural network intelligently finds the internal relationship of the data and realizes identification, but some parameters in the network need to be set manually, and it is difficult to determine the optimal parameters; based on the data change rate method, by finding the historical maximum change rate as a threshold and comparing the current collection time change rate, the method is simple and efficient, but continuous abnormal data Unrecognizable; clustering-based methods classify data by distance, but cannot know the optimal clustering result in advance

The basic method of identifying electricity consumption data is to use the rate of change to identify the threshold value set by the historical rate of change, but there is a possibility of exceeding the historical range

Images