USE OF THE MULTI-START METHOD TO SOLVE A PROBLEM IN PLANNING LONG-TERM PREVENTIVE MAINTENANCE ORDERS

With technological advancement and population growth, industries operate with high demand and full production capacity, increasing equipment degradation and the risk of failures. In this scenario, the maintenance sector becomes strategic, as long-term planning reduces costs compared to corrective maintenance. Preventive Maintenance aims to keep equipment in a satisfactory state for production, increasing its useful life and reducing breakdowns. This work proposes the use of a gluttonous heuristic, refined by the Random Variable Neighborhood Descent (RVND) and optimized by the Multi-Start (MS) metaheuristic. The objective is to obtain viable solutions to reduce unperformed maintenance and team costs in a Long-Term Preventive Maintenance Order Planning Problem (PPOMPLP). The problem, classified as NP-Hard, is applied in an iron ore processing unit, aiming to optimize the operational and financial efficiency of the plant.