汽车整车质量缺陷水平的综合评价

On Comprehensive Evaluation of the Qualitative Defect Level of a Whole Car

在对汽车的审核中,其中一项十分重要的工作就是对汽车的缺陷进行分析和评价。目前,通常的做法是根据确定的审核项,组织审核员逐项检查,并针对缺陷状况给出评价等级或缺陷分值,然后再用审核批次中多辆汽车的单车缺陷分值之和的算术平均数或其简单变形式充当整车质量缺陷水平的评价值。但是,这种方式往往又存在种种的不足。针对传统评价方法的不足,本文提出了基于改进的熵权系数法和模糊综合评判的分层混合评价模型。该模型将汽车的客观质量缺陷状况和评审专家的主观倾向很好地结合起来,从而使对整车质量水平的评价更加合理,并且能够很好地克服基于算法平均数评价方法的种种不足,使评价结果更加科学。此外,文中还给出了基于蚁群优化算法的权值向量确定方法。实践表明,在历史数据充足的条件下,蚁群优化算法是求解权值向量的优秀解决方法。

In order to overcome the shortcomings of the traditional evaluation method for the defect level of a whole car, we present a layering mixed evaluation model based on the improved entropy weight coefficient method and fuzzy comprehensive evaluation. The model combines the objective information of the cars＇ qualitative defects and the subjective information of the audit experts＇ experience nicely, thus the model is more reasonable. In addition, we also introduce the method of calculating weight vector based on Ant Colony Optimization Algorithm which is proved by practice that this algorithm is an exceUent solution to calculating weights vectors when the historical data enough .