Study of the fuzzy modelling based on parameters adjustable interpolation function

The modeling method based on fuzzy inference (MMFI) is applied to ordinary nonlinear system in this paper which interpolation function is once determined, the object model shall be determined. Because different systems apply different interpolation functions, the reestablishment of the system model is necessary, if adopting interpolation function to create the system model. A class of parameters adjustable interpolation function is designed in this paper. By changing the parameters, thereby changing the shape of the interpolation function so as to be close to such as type of the triangle and Gaussian membership function, and a more applicable model framework is deduced. PSO methods is adopted to optimize the model parameters. And by applying it to power system short-term load forecasting, it achieve -d great results.