Development of a Weighted Human Error Assessment Model and Its Application in Maritime Safety

瀏覽統計 Email 通知 RSS Feed

簡歷

基本資料

Project title

Code/計畫編號

NSC102-2410-H019-010

Translated Name/計畫中文名

權重型人為疏失機率模型之建構及其在海事安全之應用

Project Coordinator/計畫主持人

Shuen-Tai Ung

Funding Organization/主管機關

National Science and Technology Council

Department/Unit

Department of Merchant Marine

Website

https://www.grb.gov.tw/search/planDetail?id=3109290

Year

2013

Start date/計畫起

01-08-2013

Expected Completion/計畫迄

31-07-2014

Bugetid/研究經費

465千元

ResearchField/研究領域

交通運輸

"降低人為疏失在海事安全而言一直是非常重要的課題。人為可靠性評估主要為辨識成因及分析人為疏失機率，其分析模式依據資料資訊多寡而定。目前許多研究係以模糊理論為基礎在資料不完整或不足之情況，評估人為疏失機率。然當情境資料資訊相近時，該類型研究會有風險結果鑑別度過低之缺點。此外亦有以模糊認知可靠性與失誤分析評估人為疏失機率，然因該類型研究未考量變數權重、變數資訊遺漏及法則後項語言詞轉換過於主觀之故，其評估結果與實用性受到質疑。本研究之目的為發展一套權重型人為疏失量化機率評估架構，能於資料不完整之情況下，改善上述缺失，提供高鑑別度之風險分析。本研究將建立創新的模糊法則建構機制，法則後項語言詞之取得與轉換將以變數及前項語言詞之權重為基礎，該機制將具有考量變數相對重要性之特色，且能在真實反映專家專業判斷條件下進行模糊資訊的結合，其亦能克服法則後項語言詞轉換過於主觀之問題；並藉由發展新模糊推論法，在將變數資訊遺失的程度減至最低的前提下，求取模糊結論及求取量化數值，最後則依據不同類型人為可靠性評估轉換為機率。本研究架構將會在不同人為可靠性分析架構下，針對海難事故進行人為疏失風險評估，以驗證本研究之實用性。""The mitigation of human errors has been an important task in enhancing the safety in the maritime domain. Human reliability assessment (HRA) is accordingly developed. The objective of such research is to identify the causes and predict human error probabilities (HEPs) and the approaches applied vary depending upon the sufficiency degree of data. A number of studies based on fuzzy set theory are also developed to evaluate HEP results in circumstances where the lack or incompleteness of data exists. However, the degree of discriminability may be low when the data of each scenario are close to each other. In addition, some HRA studies based on fuzzy cognitive reliability and error analysis method (CREAM) are also conducted to evaluate HEPs. Nevertheless, some critics are often raised with regard to such an HAR approach and the outcomes generated are sometimes questioned. These include the incapability of considering the relative weight of inputs, the loss of information from variables and the logicality problem of acquiring the linguistic terms in the control modes. The objective of this study is to develop an HAR methodology that is capable of providing reliable HEP results while resolving the drawbacks aforementioned when the scarcity or deficiency of information exists. A novel mechanism for constructing fuzzy rules is proposed. The acquirement of the linguistic terms in consequents of each rule is based on the combination of the weights of each input and the relative importance of the linguistic term associated. Thus, the fuzzy base developed will have the feature of taking into account the relative weight of each factor. The membership functions of linguistic terms and the data generated are appreciated without distortions. Therefore, the logicality problem of acquiring the linguistic terms is also overcome. A new fuzzy reasoning is proposed to obtain fuzzy conclusions and crisp values without the concern for the loss of information from inputs. Such crisp values are finally converted into HEPs depending upon the HRA structure applied. The methodology will be verified and validated by marine case studies under different HAR framework. "

Keyword(s)

人為可靠性分析
人為疏失機率評估
海難事故
模糊推論
Human Reliability Assessment
Human Error Probability Assessment
Marine Accidents
Fuzzy Reasoning

DSpace CRIS

Development of a Weighted Human Error Assessment Model and Its Application in Maritime Safety

基本資料

Description