In order to effectively extract feature information of target radiated noise in complex and ever-changing environments, a target radiated noise feature extraction method based on optimized variational mode decomposition(VMD) and slope entropy (SlEn) was proposed. Firstly, a parameter optimization VMD algorithm(BOA-VMD) based on BOA was proposed to achieve adaptive selection of the optimal parameter combination for VMD. Then, the four types of target radiation noises were decomposed to obtain a certain number of IMF. The SlEn of each IMF component was calculated as the feature value, and was compared with three types of features: dispersion entropy, fluctuation dispersion entropy, and permutation entropy to demonstrate the effectiveness of SlEn through experiment. The results show that the proposed feature extraction method based on BOA-VMD and SlEn can achieve classification and recognition of different types of targets, and has the highest recognition rate under single and multiple feature conditions. Additionally, as the number of extracted features increases, the highest recognition rate will also increase accordingly.