This paper proposes an anomaly detection model based on deep belief network ensemble learning. The model solves the problem of unbalanced positive and negative samples of multi-source operation and maintenance data, and at the same time uses the good feature extraction function of the deep belief network. Under the Markov assumption, the least squares method is used to establish a mathematical model to obtain the 2019-2027 year of Sanya City. The orderly degree of tourism industry structure. The numerical experiment results show that the order degree of Sanya's tourism industry structure is increasing year by year. However, economic fluctuations are not a significant factor affecting the evolution of the industrial structure, and there is no obvious mutual feedback relationship between the two. It is proposed to build a characteristic industrial structure system that adapts to the construction of an international tourism island.