For large-scale complex combat scenarios, firstly, threat assessment method is studied to improve the selection of skyline method, and in designing a clustering method, a parameter adaptive density grid clustering method is proposed by the fuzzy theory to process large-scale targets, and facilitate the subsequent rapid threat assessment. Secondly, a multilevel skyline selection method is proposed, avoiding the complex process of traditional threat assessment methods such as the weight setting process, and avoiding the complex process of traditional threat assessment methods such as weight setting, and eliminating many interfering factors such as personal preference in the process of weight setting. Finally, a multi-indicator representation method of radar map is established, providing a basis for targeted decision-making by the commanders. The results of simulation and comparison experiments show that the subsequent target allocation and fire interception based on the threat assessment method ultimately achieve better combat effectiveness.