Modern warfare requires efficient integration of multi-source heterogeneous equipment data. To solve the problem of inconsistent equipment names of data from different sources, the aggregation model and aggregation process of equipment data are studied and designed. Based on the comprehensive analysis of existing algorithms and the characteristics of equipment data, a new similarity algorithm is provided for the model, the algorithm combines Jaro-Winkler and the longest common subsequence to improve the matching accuracy. Finally, experiments show that the algorithm has high adaptability and robustness compared with the traditional similarity algorithm, and can provide effective support for equipment data aggregation.