To address the problems of high operation and maintenance costs in industrial equipment maintenance decision-making, this paper proposes an intelligent maintenance decision-making method that integrates improved kernel density estimation and improved genetic algorithm. Dynamic update strategies are employed to effectively adapt to changes in data density and improve prediction efficiency and accuracy. Experimental findings demonstrate that the root mean square error of this method is as low as 0.053, and its prediction performance is better than the comparative methods. Using wind turbine generators as the empirical example, this method converges in 18 iterations, with a maintenance cost rate as low as 550 EUR/h and a minimum maintenance cost of 472 EUR/h for a single component. Sensitivity analysis shows that the cost-saving effect is weakened by the increase in preventive maintenance costs, and the advantages of group maintenance are more prominent in scenarios with high installation costs. The above results demonstrate that the research method can provide a precise and efficient technical solution for intelligent equipment operation and maintenance, and has significant engineering application value.
Xuening Lei (Wed,) studied this question.