Study On BDS Fast Static Landslide Monitoring And Early Warning Algorithm

Every year,landslide geological disasters will threaten people’s lives and bring about great economic losses.In order to reduce the impact of landslide geological disasters,the study of landslide geological disasters has become the focus of geological disasters,and has great practical significance for the early warning of landslide geological disasters.With the improvement of China’s independently researched beidou navigation satellite system,BDS-related monitoring algorithms are applied in all aspects of life,which can promote the development of BDS system while improving the monitoring technology,thus forming a situation of mutual progress.Therefore,the paper based on the BDS related algorithm was studied for the landslide geological disaster monitoring,the discussion in the process of monitoring the ill-conditioned problem and puts forward corresponding solving algorithms,at the same time of landslide geological disaster monitoring data of related research,to carry on the related data to predict,discusses the related early warning of landslide is presented.The innovation points and main research contents of this paper are as follows:(1)Explore the advantages and disadvantages of several BDS monitoring technologies.Constructed a traditional double difference carrier phase observation equation with BDS fast static positioning,and its ill-conditioned problem is discussed.This paper improves on Tikhonov regularization algorithm and Grey Wolf optimization algorithm.Combined Tikhonov regularization algorithm with grey Wolf optimization algorithm to improve the accuracy of fast static positioning of BDS.(2)Aiming at the problem that the traditional grey GM(1,1)algorithm has too much redundancy,which affects the accuracy of the displacement prediction of landslide data,the rolling window design is adopted to determine the modeling rolling windown.At the same time,in order to better balance the accuracy and speed of data prediction,the precision and speed were balanced based on the residue-length ratio,so as to determine the size of the rolling windown,and obtained an improved GMGM(1,1)algorithm.(3)Study the advantages and disadvantages of intelligent ELM algorithm,compare the ELM algorithm and BP neural network algorithm,root mean square error of the training sample as objective function,the optimal threshold value method is used to select higher prediction precision and the lowest value of hidden layer neurons,one of the best in prediction accuracy and the algorithm to calculate the amount of calculation balance intelligent r ELM algorithm.(4)In view of the data characteristics of landslide displacement,based on the early-warning algorithm in chapter 2 and chapter 3 and chapter 4,the relationship between the induced variable and the de-trend variable in the displacement monitoring data is mined by using EEMD,and the displacement sequence is decomposed into the induced term displacement and the trend term displacement.The mapping relation between rainfall,water content and induced term displacement is constructed,and a new combined landslide displacement prediction algorithm is proposed.On this basis,the early warning values of multi-source landslide monitoring data are explored,and the landslide early warning mechanism is studied.