K-means clustering algorithm-based power enterprise user rework condition monitoring method

Abstract

The invention discloses a K-means clustering algorithm-based power enterprise user rework condition monitoring method, and belongs to the field of enterprise rework monitoring. The method comprises the steps: 1, obtaining the power utilization data of an enterprise in the year where an event happens through a power system; 2, cleaning the power utilization data, and reserving available data; 3, clustering the available data through a K-means clustering algorithm, generating a clustering center, and continuously updating the clustering center until the clustering center is ended; 4, judging whether the enterprise is shut down or not through a class cluster formed by the clustering center, if so, obtaining the shutdown time of the enterprise, and turning to the step 5; 5, judging the rework time of the enterprise by calculating an outlier closest to the power consumption class cluster before shutdown; and 6, comprehensively obtaining shutdown and rework information of the enterprise in combination with the obtained time data. According to the invention, the enterprise rework and re-production time can be accurately monitored according to different enterprise operation conditions.

Description

technical field [0001] The invention relates to the field of enterprise resumption monitoring, in particular to a method for monitoring the resumption status of electric power enterprise users based on a K-means clustering algorithm. Background technique [0002] At present, the scheme of identifying the resumption of work and production of enterprises through power data is mainly to compare and analyze the electricity consumption of enterprises in the same period of the previous year, and determine the resumption time of enterprises by setting the threshold ratio of resumption of work and production. [0003] The existing threshold ratio method does not take into account the annual increase or decrease in the electricity consumption of enterprises due to the speed of development. For example, some technology companies showed a multiple increase in electricity consumption last year due to business expansion, but Due to the particularity of their companies, the power consum...

AI Technical Summary

Problems solved by technology

[0003] The existing threshold ratio method does not take into account the annual increase or decrease in the electricity consumption of enterprises due to the speed of development. For example, some technology companies showed a multiple increase in electricity consumption last year due to business expansion, but Due to the particularity of their companies, the power consumption of such companies will remain at a certain level during the shutdown period (mainly generated by power-consuming devices such as servers of technology companies), and this type of power consumption will exceed the set value. The threshold ratio, which leads to misjudgment of the resumption of work of such enterprises

At the same time, although some enterprises have resumed work, due to the impact of the incident on their business, their production capacity or efficiency is far lower than that of the same period last year. Therefore, the power consumption of such enterprises will not meet the set threshold ratio, resulting in such enterprises being misjudged as Did not resume work

[0004] The method of setting the threshold ratio by comparing the data of the same period last year is similar to "one size fits all". The policy changes and immediately shuts down, and does not take into account the differences in electricity consumption characteristics of different enterprises

Images