Evaluation Method of Residential Air Conditioning Load Time Series Adjustment Potential Based on Data Mining

Due to the diversity of air conditioning load types and the decentralized access points, it is challenging to conduct direct dispatching control. This paper discusses the exploration of air conditioner load response potential. Firstly, the K-means clustering algorithm is utilized to obtain the typical daily residential air conditioning load curve, and then the period is divided based on the characteristics of residential electricity consumption. Subsequently, a multi-period analysis model of air conditioning load during each key period is established using curve fitting technology, and a multi-period evaluation model of adjustable air conditioning load capacity considering multiple factors in residential areas is obtained. Finally, the effectiveness and applicability of this method are verified by using actual data.