基于FCM的内河一类船长实操考试效能评估

为解决内河一类船长实操考试效能评估问题,提出了一种基于模糊认知图的评价模型(FCMES).在建立影响因素层的不同评价等级基础上,通过问卷调查辨识出了25个影响因素;据专家意见将收集到的定性数据转化为三角模糊数,得到不同的模糊矩阵;利用模糊推理规则的方法聚合各模糊矩阵,再解模糊化,获取邻接矩阵.采取输入端数、输出端数及中心度3个指数的方法,分析FCMES中各因素的重要程度,选出核心变量.利用Matlab实现对FCMES认知图的推理仿真,检验核心变量的准确性.结果表明,绩效评估、评估员素质、模拟器效能、考试流程设计合理性,这4个影响因素对内河一类船长实操考试的效能有重要影响. 

A Performance Evaluation of Practice Examination for Inland First-grade Captains based on Fuzzy Cognitive Maps

In order to conduct the performance evaluation of practice examination for inland first-grade captains, this paper proposes a model of fuzzy cognitive maps based performance evaluation method (FCMES).After identifying the 25 influencing factors by using questionnaires, the evaluation grades are first defined.Then, the qualitative expert judgments are transformed to triangular fuzzy numbers, through which the fuzzy matrixes are obtained.Moreover, adjacent matrix is obtained after integrating fuzzy matrixes using fuzzy rule based reasoning and defuzzification methods.Second, the importance of influencing factors are analyzed to select the key factors by using three parameters, which are input, output, and centrality.Finally, Matlab simulation is used to verify the proposed FMCES model, the result indicates that there are four significant factors for practice examination of inland first-grade captains, specifically, they are evaluation of performance, professional quality of evaluator, efficiency of simulator, and rationality of examination process.