Font Size: a A A

Research On Forecast Model Of Debris Flow Time Series Data And 3D Dynamic Display Method

Posted on:2022-04-28Degree:MasterType:Thesis
Country:ChinaCandidate:Y ZhangFull Text:PDF
GTID:2480306563961899Subject:Computer technology
Abstract/Summary:PDF Full Text Request
Debris flow is a common geological disaster.The research of the debris flow prediction model and the realization of the dynamic demonstration system have important practical significance for the prevention and control of geological disasters.The key work of this paper includes:extracting important indicators and key indicators from useful information hidden in debris flow geological monitoring data,establishing prediction models and methods of debris flow time series data,and realizing the dynamic evolution function of debris flow.The main work of this paper is as follows:(1)A prediction model(AMSR-net model)of multi-sequence residual network based on attention mechanism is proposed.Select the index flow velocity that has a greater impact on the debris flow disaster as the prediction object.And select the mud width,bulk density,slope,volume ratio of sediment concentration and mixed sediment concentration as the influencing factor of the debris flow velocity prediction model based on the analysis of the influencing factors of the debris flow.The coupling dependence of debris flow velocity and the influencing factors of debris flow are verified.Combined the principle of attention mechanism in the prediction model to evaluate and analyzed the risk elements of debris flow.This method designed a comparison experiment between empirical formulas and multiple benchmark debris flow prediction models,the results show that compared with the empirical formula,the AMSR-net model can significantly reduce the prediction error,realize dynamic prediction,and enhance the prediction accuracy and generalization ability of the model.(2)This paper designed and implemented 3D modeling and dynamic display algorithms for debris flow which fusing the application results of predictive models.The main work includes as follows: implemented the construction of the 3D model of the debris flow mountain peak based on the Blender open source software;realized the design and realization of the 3D scene,by combining the rendering effects of earthquake,rain,and lighting in the debris flow 3D scene rendering process,and then improved authenticity of the simulation effect;Designed and implemented a 3D dynamic display algorithm of debris flow based on picking model,coordinate transformation and flow field texture.And a visualization module of debris flow analysis based on the principle of PNPOLY algorithm is implemented which can be used for risk assessment and analysis of the occurrence of debris flow.The method is helpful for researchers visual analysis of the formation of debris flow disasters and provides theoretical basis for the prediction and prevention of geological disasters.
Keywords/Search Tags:Time Series Data, Prediction Model, Data Visualization, Debris Flow, 3D Dynamic Display
PDF Full Text Request
Related items