An Accurate and Robust Comparison Method of the Intelligence for Two Unmanned Swarms Based on the Improved CRITIC and Hypothesis Test

This paper proposes an accurate and robust comparison method of the intelligence for two evaluated UMSs specialized in performing the same specific task. In the proposed method, an improved criteria importance through intercriteria correlation (CRITIC) method was adopted to assign the objective weights of intelligence evaluation indexes established for the same specific task, and the hypothesis test method was adopted to obtain the central intelligence scores for two evaluated UMSs according to their respective accidental intelligence scores obtained on two similar sets of tasks and further judge whether their central intelligence scores are statistically equal. The obtained evaluation result can also be used for determining whether these two evaluated UMSs belong to the same intelligence level. To verify the effectiveness of the proposed method, this paper presents a case study comparing the intelligence for two simulated unmanned ground vehicle (UGV) swarms specialized in performing the same specific pursuit task. The result of the case study shows that the central intelligence score for the simulated UGV swarm A are larger than that for the simulated UGV swarm B from the statistical point of view, and therefore they cannot be classified into the same intelligence level.