Multi-Objective Optimization Approach Based on Robustness Index and Its Application to Maintenance of Nuclear Power Plants
Juksanee VIRULSRI, Shinsuke SAKAI and Satoshi IZUMI
Abstract:This paper presents a new index to evaluate the sensitivity of the variation in one objective value to the variation in the other in the multi-objective optimization framework. This index is named as Robustness index, RI and is applicable to determine the decision-making point in the multi-objective problem. The decision-making point determined by the conventional method does not necessarily satisfy the robustness of the solution. However, the point determined by RI satisfies the robustness since this index shows the property of sensitivity of solutions. In this paper, the application of RI to the decision-making process in the multi-objective optimization problem is shown according to the userfs requirement. The index is examined using a system in nuclear power plant and the effectiveness to the multi-objective problem is confirmed. Key Words:Multi-objective optimization, Robustness index, Maintenance activities, Unavailability, Maintenance cost, Probabilistic risk analysis