制造系统中不确定因素的行为决策分析与机理研究

The processes of performance assessment and production capability in the manufacturing systems are influenced by the behaviors of decision makers inevitably. Meanwhile, the transfer of incomplete and inaccurate information inside the production lines also renders the stochasitc production computation model improper, as to be consistent with the expectations of the decision maker. In this research proposal, a noval optimization model is built to measure the actual correlations between the decision makers' behavior and the production output of the manufacturing system, based on the rational decision making theory, which is introduced by the applicant recently. The fundamental goal of this theory is to make the production outputs resemble the decision makers' attitudes, via the reduction of redundant information and materials. When the production capability of the production system is calculated, the aggregation algorithm from the probability theory framework is employed as a prototype. The workstations in this research are described by additional fuzzy information, and the input conditions of the systems includesome linguistic information as well. A new aggregation algorithem based on credibility theory is constructed. Possible extensions are made to model more complex manufacturing systems in terms of structure. Further, as an important composite of the rational decision making framework, experimental studies are conducted to measure the similarities bewteen the decision results and the expectations of the decision makers. At the same time, computation experiments are conducted to measure the robustness and the performance of the new aggregation algorithm with mixed uncertain information.

制造系统的绩效评估和产能计算环节中不可避免地受到决策者行为的影响，同时，制造系统内部也存在着诸如不完整信息和不精确信息的传递，此类型的不确定性因素很大程度上干扰了系统的绩效评估结果，使得基于随机理论的产能计算模型不能很好地贴合决策者心理预期。本课题基于行为决策方法，通过建立决策者行为模式与制造系统产能之间的一致性测量优化模型，以减少生产资料投放冗余度，增加系统产出与顾客行为决策结果的相似度。在计算制造系统内部的产能时，将以随机系统绩效评估方法中的聚合算法为原型，通过在随机加工中心模型中引入模糊信息，并在系统的输入条件中糅合进语言信息，重新构建基于可信性理论的聚合评估算法。并以此为基础，探讨将聚合单元拓展为具有更复杂结构的制造系统的可能。实验研究作为行为决策模型的重要组成，将衡量决策结果同行为主体心理之间的相似度，并通过与比较实验测定混合不确定性信息聚合算法的鲁棒性和算法性能。

在中国制造2025的政策指引下，如何将制造系统理论同我国当前的制造业现状相结合，能够切实体现出当前我国制造业中自动化程度不足，人工作业同机器加工高度融合的实际，为定量分析该情况下的制造系统绩效，引入确定性调度算法、随机性系统评估模型以及不确定性决策相结合的理论框架。本项目旨在通过深入分析制造系统中存在的不确定性现象，对其应用可量化的研究方法，将传统制造系统中需要通过主观推断才能获取的决策指标替换为不确定性描述，并应用最新的为此类不确定性现象建模的方式，将制造系统中的确定性和不确定性因素结合在一起，形成统一的整体模型，从而实现更加准确合理地评估制造系统性能、绩效、行为等一系列关键要素的目标。本项目在执行期间按照既定的研究计划执行，从多个刻画制造系统中不确定性因素的角度出发，结合直觉模糊集、概率模型、模糊算子方法以及经典复合模糊评判准则等，设计了受到国内外同行广泛认可的多属性决策理论方法，为在制造系统中糅合确定性、随机性和基于模糊集理论的不确定性因素，使得理论模型更加贴近实际生产，模拟制造系统中同时存在机器（确定性、随机性）、操作工人（模糊性）的复杂情形，并以此为基础得到制造系统的评判标准和实例研究。

内容获取失败，请点击重试