Typical Load Identification-Based Calculation for Air-Conditioning Load

As a primary issue in air conditioning load (ACL) management, the calculation of the power of air conditioning loads aids in quantifying the power consumption characteristics and regulation potential of the ACL, providing a guidance for ACLs to participate in demand response. In this work, a typical load identification-based ACL calculation method is proposed. First, a ridge regression-based temperature sensitivity analysis method of the power load is proposed to extract the load curves that are insensitive to the environmental temperature. Then, a Gaussian mixture model (GMM) is established on these extracted load curves to estimate the probability density distribution at each sampling time. Next, a cloud model-based benchmark load determination method is presented to identify the benchmark load for ACL calculation from the probability density distribution. Finally, the ACL is calculated as the difference between the total load and the benchmark load. The effectiveness of the proposed method is verified using air conditioning load data from a city in Zhejiang province, China, and the results show that the proposed method earns higher accuracy.