Constrained optimization of test intervals using a steady-state genetic algorithm

Probabilistic Safety Assessment and Risk-Informed Decision-Making

“…Regulatory applications of probabilistic safety assessment include monitoring and assessing the effectiveness of rules and requirements, training of the regulatory body staff, risk follow-up, risk-based safety indicators, analysis of operational events, assessment of deviations, response to emergency conditions, ranking of safety issues, ranking of importance of plant equipment, risk-informed inspection, safety guidance and prioritisation of regulatory research. Utility applications of probabilistic safety assessment include: -optimizations of technical specifications, including surveillance requirements optimization, changes and exemptions to technical specifications (Yang et al, 2000;Čepin & Martorell, 2002), -support for modification of licensing basis and assessment of plant changes, -management of in-service inspection and testing, optimization of maintenance, which includes preventive and corrective maintenance (Martorell et al, 2000;Čepin, 2002), -configuration control and planning of maintenance at outages, prioritization of activities and scheduling of the activities (Harunuzzaman & Aldemir, 1996), -improving training for operators and operational support stuff (Čepin, 2007a; Čepin, 2008), -improving of plant procedures (Prošek & Čepin, 2008), -improving plant vulnerability and security questions (Čepin et al, 2006; Čepin, 2009). In addition, probabilistic safety assessment is used for the design of new plants and it represents a chapter of the final safety report.…”

Modyfing an attribute control chart: Daudin's methodology

“…The parameters of the Genetic Algorithm have been selected following the rules showed in Martorell et al [4].…”

05/00897 Optimisation of ISI interval using genetic algorithms for risk informed in-service inspection

“…The main feature of the SSGA [4] is the utilization of overlapping populations, as it can be observed in Fig. 3.…”

“…Real encoding has been used [4], since there are large numbers of decision variables. Real codification is performed by organizing the parameters to be optimized inside an array of real and independent variables x: It is in the definition of this array where the explicit constraints that apply to each real variable, x; can be established.…”

“…Penalty functions artificially create an unconstrained optimization problem, which cope with infeasible solutions by weighting a penalty based on judgment [4]. A typical penalty functions, based on traditional calculus-based methods, is used, which is formulated as follows pða; b; iÞ ¼ Kðg; aÞSVCðb; iÞ ð 3:2Þ…”