Development of IMU (Inertial Measurement Unit ) for human foot motion tracking using MEMS inertial sensors

Abstract

The analysis of human foot movement is still a field for research and development to serve several goals, such as building models of robots, the mechanism of moving them and controlling them, as well as tracking of firefighters and miners. Due to the complexity of the indoor environment, signal occlusion problems could lead to the failure of certain positioning methods. The most important applications are in telemedicine, i.e. monitoring Alzheimer's patients from falling. This research deals with introducing the micro-electromechanical systems MEMS and inertial sensors that used to track human foot movement and studying the inertial measurement unit IMU and the inertial navigation system INS. We have modeled the  human foot movement when walking in a straight way and going up or down stairs. Where a non-linear complementary filter was used with a slight improvement to correct the gyroscope Through pi controller and calculation of rotation correction within the body frame depending on both the rotation matrix, accelerometer data and magnetometer data. The mathematical models used are elaborated. In detail with an algorithm that show the work stages. We used a zero-velocity update ZVU  with a simplified method of removing integrative velocity drift. . Finally, we used LSTM, a type of neuron network, to track the position of a human's foot when it was moving straight on flat ground and gave close results.