A Method of Ferromagnetic Object Identification Based on Magnetic Sensor and Machine Learning

Magnetic sensor is a magnetic inductance sensor, which has the advantages of high sensitivity, wide measurement range, and low energy consumption compared to other magnetic field sensors, such as magnetic gate sensors and magnetic inductance sensors. Magnetic sensors are widely used in various domains such as metal detection, safety inspection, and drone navigation. This paper describes a ferromagnetic Object detection method based on magnetic sensor. This method uses a magnetic sensor to obtain the changes in spatial magnetic field intensity caused by the continuous movement of the measured object over a certain distance within the magnetic field range, thereby identifying the type of object. The principle of this method is to carry out feature extraction, parameter training and Generative model generation for the data of varying magnetic field intensity generated by the continuous movement of various ferromagnetic objects to be measured through machine learning algorithm in advance, so as to realize Object detection. The three-axis data obtained by sensors need to be preprocessed through three steps: step 1 data filtering. Step 2 is data processing. and step 3 is data fusion before it can be used for machine learning. In the process of machine learning, this article horizontally compares three mainstream learning algorithms: SVM(Support Vector Machine), RandomForest, and LightGBM(Light Gradient Boosting Machine), all of which can achieve high recognition rates. Finally, two sets of experiments verify that this method can be applied to the identification of multiple ferromagnetic Object detection with different shapes and different ferromagnetic individuals with the same shape.