Multisize Fitting Method of Attributed Scattering Center Based on POMP-GA for RCS Prediction

The attribute scattering centers (ASCs) model offers a concise and clear physical depiction of the target. In this study, we propose a novel method for extracting ASC to predict radar cross section (RCS) at varying sizes. This method is categorized as a joint estimation approach based on perturbation orthogonal matching pursuit (POMP) and genetic algorithm (GA) (POMP-GA). The proposed POMP-GA employs an enhanced perturbation method to iteratively calculate the phase and angle information of each ASC feature in the radar frequency-angular domain, achieving high accuracy without the need for complex dictionary matrix settings. The GA is used to estimate the amplitude and frequency dependence. Furthermore, we propose an RCS prediction formula that incorporates fitting correction terms to enhance the prediction accuracy for targets of various sizes. Numerical results based on both simulated and measured data show that, for five metallic point targets, the proposed POMP-GA method reduces the computation time by 83.5% compared to the GA. Meanwhile, for the SLICY model, the multisize fitting strategy leads to a 54.6% reduction in the root-mean-squared error (RMSE) of the RCS prediction. The accuracy and superiority of the proposed POMP-GA method and RCS prediction formula have been demonstrated, confirming their effectiveness in predicting radar target RCS.