Machine Learning Based ID Identification for Linear Synchronous Motor Driven Cart with Magnetic Field Signal

The linear synchronous motor (LSM) driven industrial conveyor cart system has revolutionized modern material handling and conveyor systems, positioning them as fast, accurate, and cost-effective solutions in various applications. However, the LSM driven conveyor system lacks an accurate identification solution to retain cart IDs during power cycle or reset. In diverse applications, precise tracking of moving vehicles along the linear motors is always essential for monitoring job status and assessing the condition of running vehicles within the system. In our research, we delve into LSM magnetic field signal, and execute signal processing and feature engineering approaches to extract representative information from the data. We then explore four high-performance, low-computing-cost machine learning models, including XGBoost, LightGBM, K-nearest neighbors, and multinomial logistic regression. These models are well-suited for on-premise LSM cart systems.