The Technique Of Information Fusion With Immunization Algorithms Applied To The Prediction Of Landslide

Landslide is a geological mutations phenomenon which occurred on the rock of the peripheral crust. Landslide is a kind of geological disaster happened frequently, In recent years, as energy development and infrastructure in-depth, our country faces unprecedented geological disasters. Because of the absence of adequate economic and technological, we can't govern the landslide overall. So it is very significant to find new methods of predicting the landslide disaster and improving the precsion to reduce the disaster losses continuously.Currently, we mainly judged the landslide-shaped by monitoring its displacement. Usually there are several monitoring station with a landslide, and they will produce multiple monitoring information of displacement in time series correspondingly. But each monitoring information has not been sufficient to reflect the whole deform of landslide, and existing landslide prediction model is one-dimensional displacement parameters prediction mode. If bring certain monitoring information into the prediction model, there will have a larger deviations between the result and the actual. So if we can extract comprehensive displacement information from multiple monitoring information, and this information not only integrated each individual monitoring information ,but also reflects the whole deform of landslide ,and thus not only the monitoring information has been strengthened in space and time, but also the complementarity and reliability of the monitoring data has been enhanced. it laid a solid foundation for us to do the landslide predicting according to the prediction model which based on displacement in the latter part.To solve the above problem , this paper first raise a new technology on the basis of artificial immune algorithm at the core of multi-sensor information fusion, and use this technology to integrate a comprehensive displacement information in a number of monitoring stations and solve actual engineering problems . In this paper, we take typical landslide as the examples to verify the various algorithms, and compare with takend for the actual engineering problems, as typical for this slide body to the the preliminary results.The research results have enriched the prediction landslide theory and has important practical significance.