Research Of Wetness And Softness Foundation Sedimentation Forecast In The Loess Donga Area Based On Case-Based Reasoning Integrated With Neural Network

Highway is constucted in the wetness and softness foundation and sedimentation control after construction is the important problem that influences pavement quality, travelling speed and service life. Through final sedimentation caculation , sedimentation of design height and the planned construction filling height can be forecasted. Aiming at the uncertainty of wetness and softness foundation sedimentation,the thesis research home and the research content is as follows:(1) It analyses the distortion speciality of wetness and softness loess , expatiates by the numbers the sedimentation evaluation method of loess foundation and indicates the question.In view of the complex and uncertainty of sedimentation factors,a new method ofhighway wetness and softness foundation sedimentation forecast------the evaluation approachto wetness and softness foundation sedimentation based on case-based reasoning integrated with neural network is presented.(2) Case-based reasoning with neural network is the foundation. A model indexing wetness and softness foundation base case with neural network is set up.Making use of the fuzzy analogy concept,the similar series between the two cases with different affecting factor on wetness and softness foundation sedimentation is given.(3) In this model, the relationship of similarity between wetness and softness foundation cases is established by the neural network through training based on case-based reasoning; and the most similar base case to thewetness and softness foundation target case in the base cases of wetness and softness foundation is found out.Finally,the wetness and softness foundation stability of target case is evaluated.(4) It is shown from examples that the evaluation result is the same as the practical sedimentation result.The model forecast precision is high,generalization capability is stong.The approach have a good caculation precision ,convenient to use and is not prone to be influenced by human factor ,so it has engineer application foreground.