In this paper, for the unmanned cluster formation effectiveness evaluation issue, based on the ADC algorithm, we fully consider the influence of environmental factors and the devices’ own performance on the formation performance, and explore a set of unattended cluster formation autonomous cooperative ability test and evaluation system with comprehensive performance to adapt to the evaluation needs of the algorithm development under the complex environment. Second, because the traditional evaluation method is too dependent on the experience of experts and the BP neural network is easy to fall into the local optimum, the Sparrow Search Algorithm (SSA) is used to optimise the BP neural network. Finally, the improved BP algorithm and the extended ADC algorithm are verified by examples, and the simulation results show that the model is the most stable and closer to the actual combat capability of unmanned clusters than the traditional ADC-BP model, which can accurately and effectively evaluate the performance of unmanned cluster formations.