Evaluation of the intelligent behaviors of unmanned ground vehicles based on fuzzy-EAHP scheme

In view of that both domestic and foreign intelligent vehicle competition adopts task driven evaluation method for intelligent vehicle, leading to a conservative attitude of participant teams for fulfilling task, and hence deviating from the initial objective of technical progress, an evaluation scheme, fuzzy-extended analytic hierarchy process (fuzzy-EAHP), for the intelligent behavior of unmanned vehicles is proposed by combining EAHP with fuzzy comprehensive evaluation method. It overcomes the ambiguity problem of expertise judgment and the defect of judgment matrix inconsistency in AHP, effectively avoiding massive trying calculation and being able to rationally calculate the weight factors for all indicators of unmanned vehicle's intelligent behavior. Its quantitative results can better reflect the actual situation, and can handily be converted to specific scores, facilitating the intuitive comparison and sorting of results. The results of applying the scheme to the evaluation of "Future Challenge 2011" competition show that the proposed scheme is pragmatic, reliable, convenient, reasonable, and can do both the evaluation on individual indicator and the overall evaluation on all indicators, as well as further find out the inadequacies of unmanned vehicle's indicators and point out the directions of improvements.