A Real-Time Scheduling Method for Flexible Job Shop Considering Machine Fault Prediction

Machine fault have a significant impact on flexible job shop scheduling, causing disruptions in the existing production schedule. This directly affects production progress and delivery times, substantially reducing production efficiency and economic benefits. To address this issue, we propose a real-time scheduling method for flexible job shops that incorporates machine fault prediction (FJSRS). Firstly, Industrial Internet of Things (IIoT) and various sensor technologies were utilized to capture real-time production information. Two key features, the standard deviation of vibration signals and spindle temperature, were extracted. Secondly, a machine fault prediction method based on Long Short-Term Memory (LSTM) was proposed to predict the operational state of the machine. Finally, a flexible job-shop real-time scheduling method that incorporates machine fault prediction was proposed and optimized using the Non-dominated Sorting Genetic Algorithm II (NSGA-II) with the objective of reducing completion time and total load. This method establishes a theoretical foundation for optimization decisions in the field of flexible job shop scheduling.