Logo image
Study of Supervisor's Cognitive Decision Model in Automation
Dissertation

Study of Supervisor's Cognitive Decision Model in Automation

Cheng-Li Liu
Doctor of Philosophy (PHD), 國立清華大學, 工業工程與工程管理學系
2000

Abstract

情境認知 信任度 可靠度 警戒度 監控 自動化系統 決策 模糊控制 SA Trust Reliability Vigilance Supervisory Automation Decision-making Fuzzy Control
Automation system operating performance (ASOP) is a concept of how well an automation unit (AU) is monitored and performed by supervisors. The purpose of this research was to study and improve the ASOP for dynamic characteristics of human decision-making in automation. Previous researches have shown that situation awareness (SA) and trust are the important and influential factors of automation system performance. Firstly, a conceptual structure of relationship between SA and trust and a model of automated system performance with relation to SA, trust and automation reliability were developed. Secondly, a quantitative ASOP measuring model was proposed. Thirdly, a matrix experiment based on orthogonal arrays through a simulated system of Auxiliary Feed-Water System (AFWS) was conducted to verify the model specifically on ASOP. Finally, according to the results of the first experiment, a fuzzy logical vigilance performance alarm system was constructed to improve the operating performance. The first experimental results indicated that the quality-objective function (□ value) of human dynamic decision-making characteristics to measure the ASOP is easy and objective, a good situation awareness and correct decision-making calibration may have a greater likelihood of making appropriate decisions and performing well in automation, and keeping appropriated vigilance is an important focus. The second experimental results indicated that applying the □ value to design the fuzzy logical vigilance performance alarm system can improve the ASOP efficiently. The results of this study indicates that the quality-objective function could be regarded as measuring standard of ASOP and evaluating standard of selecting the adapted supervisors in automation. The function combined with fuzzy technique to design human-machine interface on improving cognitive decision and operating performance is a correct and important direction.

Metrics

1 Record Views

Details

Logo image