| As the historical carrier and witness of the development of human civilization,immovable cultural relics have special historical,cultural and economic values.Most of the immovable cultural relics are open-air cultural relics,which have suffered from natural damage for a long time,and their health condition is deteriorating.The preventive protection of cultural relics is an important premise and basis for prolonging the life and value of cultural relics.Therefore,the research on disease prediction of immovable cultural relics is of great significance.Preventive protection of cultural relics is to make use of all the science and technology beneficial to the research and protection of archaeological heritage to predict the disease of cultural relics.With the development of cultural relics protection Internet of Things technology,the existing preventive cultural relics protection system has made some achievements in environmental monitoring,but there is a lack of utilization and analysis of big data on cultural relics,to solve this problem,in this paper,under the idea of multi-source information fusion,a least squares support vector machine regression method based on multivariate time series wavelet correlation analysis is proposed to achieve accurate crack opening prediction of immovable cultural relics.Firstly,the correlation of multivariate time series of immovable cultural relics are quantitatively analyzed and the wavelet correlation analysis combined with stepwise regression forward algorithm is proposed to effectively identify the environmental variables that affect the crack opening of immovable cultural relics;then,the least squares support vector machine prediction model is established according to the key factors affecting the crack opening of cultural relics;finally,the good performance of the method is verified by using the environmental monitoring data of the rock mass fracture in the Qianfo Cliff of Dafo Temple in Binzhou City of Shaanxi Province.The experimental results show that the proposed method is more effective than the traditional prediction methods based on support vector machine,back propagation neural network,and relevance vector machine for crack opening of immovable cultural relics.The method in this thesis is universal and easy to implement.It is suitable for multivariate time series prediction of immovable cultural relics diseases.It provides data basis for judging the preservation status of cultural relics and lays a theoretical foundation for the preventive protection of cultural relics. |
