Piezoelectric-electromagnetic motion monitoring device with identification capability and motion pattern recognition function

In order to improve the applicability of wearable devices in the field of human motion recognition and expand their applicability in the upper limbs. This paper proposed a Piezoelectric-electromagnetic motion monitoring device (PEMD) with identification capability and motion pattern recognition function. PEMD is worn on the runner’s arm to harvest the energy generated by swinging the arm, the relative motion between the magnetic ball and the coil generates the electrical signal for sensing, and the magnetic ball transmits the impact through the spring to the piezoelectric patch to generate electricity. The influence of the wire diameter of springs and the installation angle of piezoelectric patches on the output of PEMD is tested. Experiments yielded the springs wire diameter of 0.4 mm and mounting angle of 70° of piezoelectric patches to enable the prototype to achieve optimal output, and the maximum output power reaches 564 mW. Through a series of application experiments to verify the power generation performance and recognition ability of PEMD. Establish a neural network system to realize the identification of the runner’s identity, the trained neural network has a recognition accuracy of 91.7 %. The proposed PEMD can drive the development of wearable devices for human motion sensing and monitoring.