Wyniki wyszukiwania - BazTech

1

A new approach to reliability importance analysis of complex technical systems

Chybowski L.

Journal of Polish CIMAC

|

2011

|

Vol. 6, no 2

65-72

EN

The paper presents new approach to the reliability importance analysis which is main topic of the author’s last research. The concept of comprehensive qualitative-quantitative method for the importance analysis for complex technical systems components has been shown. Extension of importance criteria of components to: safety, maintainability, spare parts waiting time and other factors has been proposed. Research conception, plan and final conclusions have been shown. The application of expert knowledge and reliability databases in case of limited knowledge about the system have been proposed.

2

Miękkie metody estymacji prawdopodobieństwa w modelowaniu bezpieczeństwa systemów.

Brandowski A.

Zagadnienia Eksploatacji Maszyn

|

2005

|

Vol. 40, nr 1

181-190

PL

Dokonany został przegląd metody estymacji prawdopodobieństwa na podstawie sądów ujawnianych przez ekspertów, stosowanej w modelowaniu bezpieczeństwa systemów technicznych. Wyróżniono cztery grupy metod uznanych za odpowiednie w tym obszarze zastosowań. Są to metody: Mendela i Sheridana, klasyczna Cooke'a, porównywania parami oraz metody nazwane "prostymi". Zostały one opisane i poddane ocenie ze względu na możliwe obszary zastosowania. Zwrócono uwagę na niepewność wyników uzyskiwanych za pomocą miękkich metod estymacji prawdopodobieństwa, szczególnie w sytuacji, gdy sądy ekspertów nie są kalibrowane. Na niepewność tę ma istotny wpływ doświadczenie oraz wiarygodność ekspertów.

EN

Methodology of probability estimation on the base of expert judgments applied in the modelling of technology system safety has been presented. This methodology is used to safety model parameter estimation when lack of objective data for estimation. Two phases of estimation procedures have been described: elicitation the judgments from experts and processing the obtained data. Application of scoring rules for expert credibility evaluation has been focused too. Various kinds of expert errors have been depicted and analysed. Four groups of methods have been focused as especially appropriate in this case. These are methods Mendel and Sheridan, classic method of Cooke, paired comparison method and group of methods named "simple". The first method needs experts whose are very skilful in probabilistic estimations. The latter are appropriate in situation when experts have no experience in probability opinions expressing. Attention was paid on uncertainty those methods, which generally depend on experience and credibility of experts. The expert's judgments should be calibrated by objective data if they are available.

3

On General Theory of Risk Management and Decision Support Systems

Kulikowski R.

Bulletin of the Polish Academy of Sciences. Technical Sciences

|

2003

|

Vol. 51, nr 3

213-226

EN

The paper extends the risk management theory based on the two factors utility concept introduced by Kulikowski [5],[6]. The extension concerns, in particular, the rate of return for financial as well as the human capital. The simple success - failure model is used for construction of the utility function. The utility function depends on the subjective probability, which is an explicite function of the objective probability of success. It depends also on the fear of risk quelling parameter. Using that function one can show when the utility of an unfair economic gamble or lottery can be accepted or rejected. Such an approach enables one to take into account the behavioural aspects in the decision support problems.

4

Extension of the Portfolio Models with an Aspect of Purchase of Information

Banek T., Kozłowski E.

Bulletin of the Polish Academy of Sciences. Technical Sciences

|

2003

|

Vol. 51, nr 3

227-243

EN

An operational model of portfolio selection is presented in this paper. The model describes the situation of an investor who constructs a portfolio of assets according to the specific criterion and purchases information about the distribution of a random vector of returns from particular assets.