Expert Decision System for Robot Selection
Expert Decision System for Robot Selection
复制标题
机器人选型专家决策系统
DOI：
10.1002/9780470050118.ecse359
复制
发表时间：
2008
期刊：
影响因子：
--
通讯作者：
E. Karsak
中科院分区：
文献类型：
--
作者：
E. Karsak
在过去的二十年中，由于全球竞争对更高质量，生产力和灵活性以及健康和安全的需求，工业机器人的使用呈上升趋势。在这篇文章中，基于知识的决策框架，它集成了专家系统和决策支持系统，提出了提高机器人选择决策的质量和效率。一个多准则决策（MCDM）的方法是在专家决策系统中使用，因为专家系统通常提供了一个简短的列表的机器人替代品的技术方面的基础上，和一个适当的MCDM技术是必需的，以评估候选人名单的替代品，并确定最符合用户要求的机器人。所开发的决策支持系统集成了用户需求与基本的机器人属性，采用质量功能配置（QFD）和模糊线性回归。与以前提出的机器人选择技术相比，所提出的决策框架具有优势。所提出的框架的优点可以注意到，将专家知识结合到一个困难的问题，使用户的要求，一般是定性和机器人的特点，以考虑在机器人选择过程中，通过采用QFD原则，也考虑到机器人的特点之间的关系，从而忽视不切实际的优先独立性假设经常遇到的早期机器人选择研究，以及通过适于考虑高系统模糊性的模糊回归来执行上述函数关系的参数估计。一个机器人选择的例子来说明集成的决策框架。
保留字：
机器人选择;
多准则决策;
质量功能展开;
模糊线性回归;
基于知识的决策支持系统;
专家系统;
工业机器人;
先进制造系统的合理性
Over the past two decades, an upward trend has been observed in the use of industrial robots because of the global competitive needs for higher quality, productivity, and flexibility, as well as for health and safety. In this article, a knowledge-based decision framework, which integrates an expert system and a decision-support system, is proposed to enhance the quality and efficiency of the robot selection decisions. A multicriteria decision making (MCDM) methodology is used in the expert decision system because the expert system usually provides a short list of robot alternatives based on the technical aspects, and an appropriate MCDM technique is required to evaluate the shortlist of alternatives and determine the robot that best meets the user requirements. The developed decision-support system integrates user demands with essential robot attributes that employs quality function deployment (QFD) and fuzzy linear regression. The proposed decision framework possesses advantages compared with the techniques previously proposed for robot selection. The merits of the proposed framework can be noted as incorporating expert knowledge to a difficult problem, enabling both user requirements that are generally qualitative and robot characteristics to be considered in the robot selection process by adopting the QFD principles, taking into account also the relationships between robot characteristics and thus disregarding the unrealistic preferential independence assumption frequently encountered in earlier robot selection studies, and performing the parameter estimations of the abovementioned functional relationships by fuzzy regression that is suitable for considering high system fuzziness. A robot selection example is presented to illustrate the integrated decision framework.
Keywords:
robot selection;
multicriteria decision making;
quality function deployment;
fuzzy linear regression;
knowledge-based decision-support system;
expert system;
industrial robots;
justification of advanced manufacturing systems