Intelligent Recommendation Of Demand Response Combined With Knowledge Graph And Neural Tensor Network

With the rapid development of human society,the demand for energy is increasing day by day,and the pattern of multi-source complementary is gradually formed.How to realize the efficient use and scheduling of energy has attracted more and more attention,and demand response plays an important role in power grid scheduling.Demand response is not only a response to the national strategy of "three types and two networks",but also an important measure to ensure social and economic development.Intelligent recommendation of demand response can effectively promote the implementation and promotion of demand response among ordinary users,and let more private enterprises and ordinary users participate in the reform of power grid.With the development of the Internet,everyone is surrounded by massive information every day,and only a small amount of information is what we are interested in.Intelligent recommendation can effectively help people filter information and put different information to interested people.Intelligent recommendation has been introduced into the field of power consumption in the early days and achieved remarkable results.However,with the increasingly complex data scene of power grid,traditional recommendation algorithms are faced with problems such as data sparsity,cold start,increased calculation delay and decreased recommendation accuracy.In view of the above problems,this paper proposes a demand response intelligent recommendation model based on knowledge map and neural tensor network.The model effectively overcomes the cold start problem of intelligent recommendation system by introducing knowledge map,and effectively solves the problems of data sparsity and recommendation accuracy decline by introducing representation learning.Finally,the recommendation delay is reduced by using preference propagation algorithm.At the same time,the fusion model and the collaborative filtering based recommendation model are compared based on the public electricity data set.The experiment makes a comparison from the recommendation accuracy,recall rate,coverage rate,recommendation list diversity and recommendation novelty.The results show that the demand response recommendation algorithm based on NTN model can effectively overcome the problems of data sparsity,cold start and the decline of recommendation accuracy.Compared with the existing demand response intelligent recommendation algorithm based on collaborative filtering,the recommendation accuracy and recall rate are significantly improved.