The Research For The Application Of The Adaptive Network Fuzzy Inference Systems Model In Slope Deformation Prediction

Because the slope deformation tends to cause great impact on people’s daily lives and theconstruction, the slope deformation prediction is an important research direction indeformation prediction. Therefore, while we monitor the slope deformation, analyze themonitoring data, and then make a timely and accurate prediction. It can largely reduce thelosses of national economic and people’s life safety which are caused by disaster occur. In thepaper, on the basis of studying some of the existing deformation prediction models, usingfuzzy set theory, combining BP neural network model, proposing Adaptive Neuro-FuzzyInference System, also known as Adaptive Network-based Fuzzy Inference System,abbreviated ANFIS. This model combines the advantage of fuzzy system language reasoningability and neural network learning mechanism, I use the different types of data in an slopeengineering example to verify the feasibility of this model in the slope deformation prediction.In this paper, the results of research and contents are summarized as follows:1. Simply introducing some of the deformation forecast models which are commonlyused. Focusing on the basic theory、structure、algorithms and process of BP neural networkmodel.2. The fuzzy system theory is introduced in detail, including the selection of fuzzymembership and fuzzy rules. Simply describing the application of each parameter in fuzzysystem toolbox, comparing the advantages and disadvantages of fuzzy system and BP neuralnetwork, integrating the two theories, putting forward the applicability of the AdaptiveNetwork-based Fuzzy Inference System (ANFIS), and making a detailed description of thismodel..3. Based on adaptive fuzzy neural network model, this article selects the existing slopedeep horizontal displacement data to verify this model, proves the feasibility of the model inthis slope deformation prediction.