Multi-target threat assessment in unmanned cluster confrontation scenarios

Abstract

Threat assessment is a key module of the unmanned intelligent platform decision-making system. This paper mainly studies the target threat assessment in multiparty unmanned cluster confrontation scenarios. First, a three-level threat assessment functional model is established, and the target intent is regarded as a tactical threat, the threat assessment index system has been improved. Then the Adaboost algorithm is introduced into the threat assessment problem, and on this basis, the Adaboost algorithm is improved by building a various meta-learner algorithm library, which reduces the algorithm’s error rate. Finally, the threat assessment based on improved Adaboost is simulated, and the results show that the algorithm can be used in the threat assessment of multi-party unmanned cluster confrontation scenarios.