Kobayashi Akane, Nakamura Kenji, Ono Takahito
Department of Mechanical Systems Engineering, Tohoku University, Sendai 980-8579, Japan.
Department of Management Science and Technology, Tohoku University, Sendai 980-8579, Japan.
Sensors (Basel). 2025 Jul 18;25(14):4471. doi: 10.3390/s25144471.
This study aimed to monitor the operating state of an induction motor, a type of electromagnetic motor, using a highly sensitive magnetic sensor, which could be applied for anomaly detection in the future. Monitoring the health of electromagnetic motors is very important to minimize losses due to failures. Detecting anomalies using the changes compared with the initial state is a possible solution, but there are issues such as a lack of training data for machine learning and the need to install multiple sensors. Therefore, an attempt was made to acquire the various operating states of a motor from magnetic signals using a single magnetic sensor capable of non-contact measurement. The relationships between the magnetic flux density from the motor and the other motor conditions were investigated. As a result, the magnetic spectrum was found to contain information on the rotor rotation frequency, torque, and output power. Therefore, the magnetic sensor can be applied to monitor a motor's operating conditions, making it a useful tool for advanced data analysis.
本研究旨在使用高灵敏度磁传感器监测感应电动机(一种电磁电动机)的运行状态,该传感器未来可用于异常检测。监测电磁电动机的健康状况对于将故障导致的损失降至最低非常重要。利用与初始状态相比的变化来检测异常是一种可行的解决方案,但存在诸如机器学习缺乏训练数据以及需要安装多个传感器等问题。因此,尝试使用能够进行非接触测量的单个磁传感器从磁信号中获取电动机的各种运行状态。研究了电动机的磁通密度与其他电动机状态之间的关系。结果发现,磁谱包含有关转子旋转频率、扭矩和输出功率的信息。因此,磁传感器可用于监测电动机的运行状况,使其成为高级数据分析的有用工具。