Method to achieve compressor characteristics maps based on SVM and PSO

An algorithm for support vector regression is improved and implemented as in computer programming. And this new algorithm, together with the particle swarm optimization, is used to regress the nonlinear function about how would compressor efficiency or corrected inlet air flow vary when the rotor speed, pressure ratio and compressor inlet flow angle changed. Here the inlet flow angle is considered as one of the independent variables, so this is the improvement for the method to achieve the compressor characteristics maps only depending on rotor speed and pressure ratio. The nonlinear relationship between the dependent variables and independent ones described above was regressed by support vector machine, and its degree of approximation was optimized by particle swarm optimization. Finally, the method was tested to be effective to achieve the nonlinear function describing the compressor characteristics maps.