The invention discloses a radiation source individual characteristic enhancement method based on a time-varying filtering theory, and relates to a radiation source individual characteristic enhancement method. The objective of the invention is to solve the problems of low individual identification accuracy of an existing radiation source and identification failure caused by change of main signal parameters. The method comprises the following steps of: 1, performing time-frequency distribution calculation on a multi-component signal, and extracting time-frequency information to obtain time-frequency information of each signal component in a main signal component; 2, recovering and separating signal components in the main signal components one by one based on a time-varying filtering algorithm of order time-varying short-time fractional Fourier transform to obtain an estimation result of the sum of the main signal components; 3, subtracting an estimation result of the sum of the main signal components from the radiation source signal to obtain a residual component of the multi-component signal; 4, performing feature extraction on the residual component, and constructing a feature vector; and 5, inputting the constructed feature vectors into a classifier, and outputting a classification and recognition result. The method is applied to the field of radiation source individual characteristic enhancement.