With the wide application of power load monitoring in production and life, nonintrusive load identification technology has been developed and has good prospects for development. In this paper, a behavior recognition method of home load based on feature matching degree is proposed, which is difficult to optimize the recognition feature and the convergence of the recognition algorithm. The DTW algorithm is used to calculate the similarity between the test template and the reference template, which can effectively identify the load data of home use, and simulate the load identification process by three cases, and verify the feasibility of the algorithm.