文章基本信息

标题：Human Skill Evaluation Based on Control Engineering Approach
本地全文：下载
作者：Kazushige Koiwai ; Yuntao Liao ; Toru Yamamoto 等
期刊名称：IFAC PapersOnLine
印刷版ISSN：2405-8963
出版年度：2016
卷号：49
期号：19
页码：633-638
DOI：10.1016/j.ifacol.2016.10.636
语种：English
出版社：Elsevier
摘要：Although progress is being made on automation technology for the production field, some equipment still requires a human operator with experience and skills to judge the conditions. These operators have the professional skill needed to work efficiently. However, the social issues of a decreasing and aging population create difficulties in transferring professional skills in a field where human skill is needed for the operation of some equipment. In this paper, operators and pieces of equipment that require human skill are described as controllers and systems in the control engineering field, respectively. The controller representing a human consists of a proportional-integral-derivative (PID) controller and cerebellar model articulation controllers (CMACs), which are used to tune the PID parameters. The CMAC is a mathematical model based on a cerebellar information processing mechanism that uses a mapping function. The CMACs learn the human skill using data that include the system output and control signal. To evaluate the skill, the data of professionals and beginners are compared as the trajectories and rates of change of the PID parameters. As an application example, an evaluation of the human skills needed for an excavator swing operation will be performed in this paper.
关键词：Neural networksPID controllersCMACSkill based systemsNonlinear systems