Transient-feature-clustering-based household load identification method

Abstract

The invention discloses a transient-feature-clustering-based household load identification method. The method comprises: an electric load feature database is established; an electric parameter of an electric load is collected, a transient event of the load is detected by using a dual-side accumulated CUSUM algorithm of a sliding window, and transient data are recorded; a clustering analysis is carried out by using a sum-of-squared-error criterion function, and a transient feature is extracted; and the extracted transient feature is compared with the electric load feature database and a household load is identified. On the basis of the collected information like the electric parameter of the electric appliance, the clustering analysis is carried out based on using a K-mean-value sum-of-squared-error criterion function by using the load transient feature as a feature quantity, thereby identifying a load with a similar or overlapped steady feature, improving the load identification precision, and realizing load identification of a household appliance or electrical equipment. Moreover, the method is simple; the calculation load is low; the electricity using behavior of the user is regularized; the power transmission losses are reduced; and the energy utilization rate is improved.

Description

technical field [0001] The invention relates to the technical field of household electricity identification, in particular to a household load identification method based on transient feature clustering. Background technique [0002] Classified metering of household electricity consumption is an inevitable requirement for smart energy meters. The electric energy meter currently used only realizes the time-of-use measurement of the total electricity, but does not have the function of classifying and measuring according to the nature of the load. Household load identification is to realize the classification and measurement of electricity bills, so that the user's electricity bill list is like a telephone bill list, and the electricity consumption of various household appliances is clear at a glance. This can enable home users to know their own electricity consumption in a timely manner, and provide reference for users to reasonably allocate electricity consumption time and c...

AI Technical Summary

Problems solved by technology

Although the load steady-state features are easy to extract and can identify most of the household loads, the disadvantage is that when the steady-state features of household appliances are similar or overlapped, they cannot be accurately identified and classified

Images