看看國際先進的超讚邏輯與觀念
Michael Havbro Faber
Risk and Safety in Engineering
Lecture 1 Introduction and overview
風險的基準
幾個風險的來源
1.客觀未知/人類知識經驗不足
2.當事人主觀上不理解/沒有意識到
3.當事人理解與意識到可能有問題(=風險)，但是忽視
4.理解與意識到有風險，但是選擇接受(冒險)
5.有風險，但採取錯誤/無效方式來處理風險
Lecture 2 Review of probability theory and
statistics
懂統計機率，才能談論風險
不確定性=難以判別機率的正確性
一本體論，二是認識論，三是人腦神經認知侷限
Lecture 3 Engineering decision analysis
關鍵在於成本效益的推估與取捨
A decision
is: a committed allocation
of resources
Analysis: Prior, Posterior and Pre-posterior
欸，腦洞大開
工安的東西做了太多事前prior推測與猜想(e.g.,防爆區域等級劃分，健康風險評估)，缺乏事後posterior的驗證與反省；難怪給人感覺裝神弄鬼、假鬼假怪
工安的東西很需要進行流言驗證與終結(=花大錢做實驗)
資訊與驗證結果的價值
Hint:
事前多花一些小錢進行假設條件的驗證，可以避免決策錯誤事後花大錢收尾（只是很多時候，主事者腦筋簡單）
Lecture 4 Risk Assessment in civil engineering
所謂的評估，有很多面向(不同單位與屬性，其實可以用DEA來比較)
JCSS-Joint Committee of Structural Safety
"Acceptance
criteria in risk assessment".
• Which objectives are to be
included in the utility function for comparison/decision regarding risk
acceptability and how?
• Which constraints secure
that the risk of the best alternative with the maximum expected utility is acceptable
with respect to the safety of the individual, legislative rules of different
kinds etc?
• How are the objectives
measured?
• How is the mutual weighing
of objectives? Are all objectives transferred to a common scale?
風險要有比較才有意義
欸，間接的影響(Integrity/Robustness)要能夠具體被呈現(讓人有感)其實不簡單
Lecture 5 Classical reliability analysis
有名的浴缸曲線
呵呵，這部份的計算要來去請教品保的專家
故障失效率參考資料
結構安全的觀念
系統餘裕>使用負荷
Lecture 6 Structural reliability analysis
Lecture 7 Software tools for reliability
analysis
Lecture 8 The JCSS probabilistic model code
Lecture 9 Systems reliability analysis and robustness
Lecture 10 Time variant reliability
analysis
Lecture 11 Reliability based structural design
and assessment
Lecture 13 Optimal decision making and risk
acceptance criteria
沒有留言:
張貼留言