CONSTRAINED OPTIMIZATION OF AN INTEGRAL SECURITY INDICATOR FOR ADAPTIVE MANAGEMENT OF HAZARDOUS FACILITIES
Elshan Hashimov, Ramil Akhundov, Aziz Talibov, Islam Islamov
Abstract
The subject matter of the article is quantitative assessment and adaptive decision-making for managing the safety state of potentially hazardous facilities using an integral security indicator. The goal is to develop a mathematical framework for selecting an optimal management action that maximizes the indicator under resource, time, and acceptable risk constraints. The tasks are: (i) to formalize the facility state as a normalized vector of indicators and construct the integral indicator A(t); (ii) to formulate a constrained optimization model for selecting the optimal action v*(t) based on maximizing A(t); (iii) to define a periodic update and feedback mechanism and an effectiveness measure η for evaluating actions across decision cycles. The methods used are multi-criteria aggregation in a normalized indicator space, constrained optimization, argmax-based decision rules, and iterative feedback evaluation. The following results were obtained: an indicator-based state model and integral security indicator were constructed for compact decision support; a constrained selection problem was defined for a discrete action set with cost, implementation time, and risk admissibility constraints; and an operational loop of reassessment, action application, and feedback was specified, enabling adaptive refinement of decisions and visualization of A(t_k) dynamics. Conclusions. Scientific novelty: 1) an integrated constrained-optimization scheme that maximizes an integral security indicator while jointly accounting for cost, time, and risk constraints; 2) a feedback-enabled adaptive control loop with η and periodic reassessment to improve action selection under changing conditions.