New common mathematical models of fanners performance based on the partial least squares regression method

In this paper, the partial least squares regression (PLS) method is adopted for the first time to solve the multivariate and nonlinear surface fitting problem of fanner regulation performance. With this new method, two new types of optimizing mathematical models (PLS1 and PLS2) are developed, and the detailed steps for doing so are also given. By taking a centrifugal fan and an axial flow fan for example, the paper develops and evaluates the common simulation models of regulation performance with high precision and comprehensive applicability. In addition, the applications of these two types of models are discussed.