Predictive maintenance in multi-objective supply chains by combining machine learning and evolutionary optimization

Predictive maintenance, as a new approach to industrial equipment management, uses machine learning to predict failure probability and evolutionary optimization algorithms to determine optimal maintenance strategies. In this study, a hybrid model is presented that first estimates the failure probability of equipment using XGBoost and LSTM algorithms and then uses NSGA-II and PSO to optimize maintenance decisions. The results show that the NSGA-II algorithm performs better than PSO in reducing maintenance costs by ۴۲.۸% and reducing equipment downtime by ۵۵.۴%. The innovation of this research lies in integrating machine learning and evolutionary optimization into an intelligent and efficient framework that can reduce operating costs, increase equipment reliability, and optimize maintenance strategies. The findings of this research can be used in various industries to improve productivity and reduce unexpected failures.

کلیدواژه ها:

Predictive Maintenance ، Machine Learning ، Evolutionary Optimization ، NSGA-II ، PSO