【作者】 夏春艳；

【导师】 颜声远；

【作者基本信息】 哈尔滨工程大学 ， 机械设计及理论， 2010， 博士

【摘要】 随着核电事业的迅速发展,安全已成为核电存在和发展的基础。良好的核电厂主控室人机界面设计对减少人因失误、保障核电厂的安全运行具有十分重要的作用。人机界面评价是保障人机界面设计质量的重要手段,科学、适用的人机界面评价理论和方法是正确评价的前提和基础,因此,建立和完善人机界面评价理论和相关技术具有重要的理论和现实意义。本文以核电厂主控室人机界面评价为研究对象,分析了人机界面评价研究的必要性,回顾了评价理论的研究现状,研究了具有多层次、多属性、不完全信息的复杂系统评价问题。针对评价指标的筛选、指标体系的构建、指标权重的分配、不完全信息系统的评价等几个关键的技术问题展开研究。具体的研究工作如下：首先,在回顾和分析评价指标筛选方法的基础上,运用系统分析的思想,综合考虑指标的相关性、重要性和有效性,提出了评价指标筛选的多因子综合算法；同时对评价指标的有效性检验进行了探索性的研究,提出了利用灰色关联分析进行评价指标的有效性检验,研究案例验证了所提出方法的有效性和可行性。在此基础上,分析了核电厂主控室人机界面评价的相关标准和影响因素,构建了核电厂主控室人机界面评价指标体系。其次,将灰色理论引入到层次分析法中,从整体相似性的角度构造专家判断的信度系数,提出了基于信度系数的指标权重分配方法,实现了认知特性的定量描述,解决了多专家权重信息的集结问题。同时利用灰色关联分析构造Ⅰ型和Ⅱ型判断矩阵,提出了基于灰关联层次分析的多指标权重分配方法,给出了权重分配实现的过程和算法,分析了权重分配结果的稳定性。实例研究表明所构造的Ⅰ型判断矩阵自然满足一致性要求,适合于指标过多时的权重分配问题。再次,探讨了人机界面不完全信息的评价方法,提出利用区间灰数处理人机界面不完全信息。研究了包含不完全信息的复杂系统多方案评价问题,以灰色理论和理想点法为理论基础,构造了灰区间绝对关联分析模型和灰区间理想点模型,提出了灰区间相近度指标,建立了基于相近度的人机界面不完全信息综合评价模型,通过案例分析验证了该模型的有效性和合理性。同时研究了具有不完全信息的多层次复杂系统评价问题,提出了基于灰区间聚类的多层次评价方法,建立了典型白化权函数的灰区间聚类多层次评价模型,给出了该方法实现的具体步骤和边界条件,研究案例表明了该方法的可行性。最后,研究了核电厂主控室人机界面评价软件开发过程中主、客观评价指标的量化方法,开发了核电厂主控室人机界面综合评价软件,利用该软件对某核电厂主控室人机界面后备盘中部分盘面设计情况进行评价,分析了评价结果,提出了改进意见,验证了本文提出的理论和方法。更多还原

【Abstract】 With the development of nuclear power, safety has become the foundation on which nuclear power can exist and develop. The well designed human-machine interface plays an important role in reducing human error and ensuring the safe operation of nuclear power. Human-machine interface evaluation is an important means to ensure the design quality of human-machine interface. The scientific human-machine interface evaluation theory and method is the premise and the basis of correct evaluation. Therefore, it has great theoretical and practical significance to establish and perfect human-machine interface evaluation theory and correlative technique.In this paper, human-machine interface evaluation for main control room of nuclear power plant was selected as object investigated. The necessary of human-machine interface study was analyzed. The status of evaluation theory was reviewed. The evaluation problem of complicated system with multi-attribute, multi-hierarchy and incomplete information was studied. The key technology issues include the selection of evaluation index, the construction of index system, the assignment of weight, the evaluation of incomplete information system and so on. The main contributions are summarized as follows:Firstly, the methods of index selection were reviewed and analyzed. On this basis, the method of index selection based on multi-factor arithmetic was put forward by using system theory, considering the relativity, the significance and the validity of index. At the same time, the validity examination of index was studied exploringly. Grey relation analysis was used to exam the validity of index. The investigation example proves the validity and feasibility of the proposed method. On this basis, the correlative standard and the influence factor of human-machine interface evaluation for main control room of nuclear power plant were analyzed; the evaluation index system of human-machine interface for main control room of nuclear power plant was constructed.Secondly, Grey theory was imported into analytic hierarchy process. The reliability coefficient was constructed by the similarity degree. The method of weight assignment based on reliability coefficient was proposed. The quantitative representation of the cognitive characteristic was carried out. The synthesis of multi-expert’s evaluation information was resolved. At the same time, two kinds of judgment matrix were constructed by using of grey relation analysis. The method of weight assignment for multi-criterion based on grey relation analytic hierarchy process was proposed. The process and the arithmetic of weight assignment were presented. The stability of weight assignment results was analyzed. The example proves that the first judgment matrix fits in with consistent requirements spontaneously, and the proposed method is suitable for the problem of weight assignment for multi-criterion.Thirdly, incomplete information evaluation method of human-machine interface was discussed. Interval grey number was used to deal with incomplete information of human-machine interface. The decision-making problem of complicated system with incomplete information was studied. Based on grey theory and TOPSIS, the grey interval absolute relation analysis model and the grey interval TOPSIS model were established. The grey interval nearness degree was presented. The incomplete information evaluation model based on nearness degree was established. The evaluation example proves its rationality and validity. At the same time, the evaluation problem of multi-hierarchy system with incomplete information was studied. The method of multi-hierarchy evaluation based on grey interval clustering analysis was proposed. The multi-hierarchy evaluation model of grey interval clustering based on representative white weight function was established. The process and the boundary conditions of the method were given. The evaluation example proves its feasibility.Finally, the quantification methods of subjective index and objective index were studied. On this basis, the human-machine interface comprehensive evaluation software for main control room of nuclear power plant was developed. The evaluation software was used to evaluate a part panel design of back-up panels’human-machine interface in the main control room of the nuclear power plant. By analyzing the evaluation results, the suggestions of design improvement were presented. The theory and method which have been proposed in this paper was confirmed.更多还原