Abstract:In view of the lack of systematic methods and the difficulty of effectively evaluating combat capabilities in the combat test of land unmanned equipment clusters, the boundaries and capabilities of the test tasks are clarified by defining the core concepts, classifications and typical combat scenarios of land unmanned equipment clusters. Analyze the key capability requirements of cluster operations, refine the quantitative and qualitative test index system including collaborative decision-making efficiency and dynamic networking stability, and solve the problem of "what to measure". On this basis, the "requirements-indicator-scenario-method" linkage test system is constructed, and the design of virtual and actual hybrid test environments close to actual combat, such as complex terrain and electromagnetic countermeasures, and the collaborative linkage scheme between the semi-physical simulation platform and the actual test site are studied, so as to realize the planning of "where to measure". Then, a hierarchical and progressive test implementation method of "stand-alone verification - small-scale collaboration - large-scale complex tasks" is designed, and the experimental data collection and performance analysis technology based on big data and artificial intelligence algorithms are studied simultaneously. In addition, special test schemes and simulation methods for abnormal working conditions are proposed for test difficulties such as dynamic topology change and multi-objective conflict avoidance. After the typical combat mission test of a certain type of land unmanned equipment cluster, the feasibility and effectiveness of the proposed method are verified, and a generalizable test process specification is formed, and the combat test method of land unmanned equipment cluster is studied.