| 基于视线追踪技术的工艺操作人员人为失误识别研究 |
| |
| 引用本文： | 胡瑾秋,张来斌,胡静桦.基于视线追踪技术的工艺操作人员人为失误识别研究[J].中国安全生产科学技术,2019,15(5):142-147. |
| |
| 作者姓名： | 胡瑾秋 张来斌 胡静桦 |
| |
| 作者单位： | (中国石油大学（北京） 安全与海洋工程学院 油气生产事故预防与控制基础研究室，北京 102249) |
| |
| 基金项目： | 基金项目: 国家自然科学基金项目(51574263);中国石油大学（北京）教育教学改革项目（2017-2019） |
| |
| 摘 要： | 为了对生产加工过程中工艺操作人员的不安全行为进行有效监控,及时避免因操作失误带来的工艺安全事故,开展基于视线追踪技术的工艺操作人员人为失误识别研究。通过设计眼动实验方案,利用视线追踪技术采集操作者在工艺流程控制过程中各类失误模式的眼动数据,并对采集的数据进行统计,提取失误行为的眼动特征,并建立人为失误智能识别方法,利用距离函数聚类实现现场操作人员失误状态的准确识别。研究结果表明:以被试样本在不同区域的视线停留时间百分比作为特征参数,并运用欧氏距离函数分类法判别人员操作状态,对操作人员的不安全行为能够准确识别。
|
| 关 键 词： | 视线追踪 眼动分析 人为失误 模式识别 |
Study on human error recognition of process operators based on eye tracking technology |
| |
| HU Jinqiu,ZHANG Laibin,HU Jinghua.Study on human error recognition of process operators based on eye tracking technology[J].Journal of Safety Science and Technology,2019,15(5):142-147. |
| |
| Authors: | HU Jinqiu ZHANG Laibin HU Jinghua |
| |
| Affiliation: | (Laboratory of Basic Research on Accident Prevention and Control for Oil and Gas Production, College of Safety and Ocean Engineering, China University of Petroleum (Beijing), Beijing 102249, China) |
| |
| Abstract: | In order to effectively monitor the unsafe behavior of the process operators during the production process, and avoid the process safety accidents caused by operation error in time, the human error recognition of process operators based on the eye tracking technology was studied. The scheme of eye movement experiments was designed, and the eye movement data of various error patterns for the operators in the process procedure control were collected by using the eye tracking technology. The statistics on the collected data was carried out, and the characteristics of eye movement for the error behavior were extracted. An intelligent recognition method of human error was established, and the accurate recognition on the error status of on site operators was realized by using the distance function clustering. The results showed that the unsafe behavior of the operators could be recognized accurately by taking the percentage of time of sight fixed on different regions for the tested samples as the characteristic parameter and judging the operation status of operators through using the Euclidean distance function classification method. |
| |
| Keywords: | eye tracking eye movement analysis human error pattern recognition |
| 本文献已被 CNKI 维普 等数据库收录！ |
| 点击此处可从《中国安全生产科学技术》浏览原始摘要信息 |
|
点击此处可从《中国安全生产科学技术》下载全文 |
|
