Comprehensive load model dominant parameter screening method based on activation subspace and terminal

The invention discloses a comprehensive load model dominant parameter screening method based on an activation subspace and a terminal, and aims to solve the problems of low identification efficiency and poor accuracy caused by high-dimensional nonlinear characteristics of complex comprehensive load model parameters. The method comprises the following steps: firstly, constructing a comprehensive load model parameter vector group, calculating an average functional matrix of a model output gradient, performing characteristic decomposition on the matrix, and screening out an activation subspace described by a characteristic vector according to the size of a characteristic value; and according to the activation subspace, determining a feature value with the maximum gap and a corresponding feature vector, and determining the importance degree of elements in the feature vector relative to model output, thereby screening out dominant parameters of the comprehensive load model. In this way, the dominant parameters inf