Abstract:The relationship among the controlling factors of reservoir slope stability are highly non-linear. Meanwhile,the adaptive neuro—fuzzy inference system(ANFIS)has been widely recognized for its capability of nonlinear dynamic analysis and processing both certain and uncertain information at the same time.Hence, the employment of ANFIS to assess the stability of reservoir slope was proposed.With eight parameters including permeability coefficient,declining rate of the water level,pore pressure ratio,slope angle,slope height,cohesion,internal friction angle,and severity as the inputs,and the reservoir slope stability coefficient as the output,a ANFIS model has been constructed based on 21project cases.The training correlation coefficient of the model was 0.999 96,and the correlation coefficient of the validation was 0.977 48,which was significantly better than the BP neural network model.The successful prediction on the slope stability of a dam reservoir in Jiangxi Province illustrated a desirable forecasting ability of the established ANFIS model for coupling multiple impact factors