Title :
A PSOGSA method to optimize the T-S fuzzy neural network for displacement prediction of landslide
Author :
Ping Jiang ; Zhigang Zeng ; Jiejie Chen ; Huiming Tang
Author_Institution :
Sch. of Autom., Huazhong Univ. of Sci. & Technol., Wuhan, China
Abstract :
In this study, a modified method for the displacement of landslide prediction is presented. This method is based on Takagi-Sugeno fuzzy neural network (T-S FNN), with an efficient hybrid optimization algorithm based on the combination of particle swarm optimization (PSO) and gravitational search algorithm (GSA) applied to optimize the parameters for T-S FNN. Moreover, correlation analysis is an important analysis to look for the potential input variables for a predict model. Pearson cross-correlation coefficients (PCC) and mutual information (MI) are adopted in this paper. The performance of the obtained model is verified through two case studies in Baishuihe (BSH) and Liangshuijing (LSJ) landslide in the Three Gorges reservoir in China.
Keywords :
correlation methods; fuzzy neural nets; geomorphology; geophysics computing; particle swarm optimisation; search problems; BSH landslide; Baishuihe landslide; China; LSJ landslide; Liangshuijing landslide; MI; PSOGSA method; Pearson cross-correlation coefficients; T-S FNN; T-S fuzzy neural network; Takagi-Sugeno fuzzy neural network; Three Gorges reservoir; correlation analysis; efficient hybrid optimization algorithm; gravitational search algorithm; landslide displacement prediction; mutual information; particle swarm optimization; Indexes; Manganese; Reservoirs; Support vector machines; Terrain factors; Vectors; Fuzzy neural network; Gravitational search algorithm; Particle swarm optimization; Pearson cross-correlation coefficients; Takagi-Sugeno; mutual information;
Conference_Titel :
Systems, Man and Cybernetics (SMC), 2014 IEEE International Conference on
Conference_Location :
San Diego, CA
DOI :
10.1109/SMC.2014.6974080
