A MIMU-based gait and pace detection system designed for human powered energy harvest devices

摘要

Energy harvesting from the human body is a field of promising research. To improve the efficiency of the harvester, there is a need to identify its motion accurately and control the harvester's time. This paper presents a human gait and pace detection system based on MIMU, which consists of a three-axis accelerometer, three-axis gyroscope, and the parameters of human gait detected by MIMU in real-time. An extended Kalman filter based on quatemions is designed for an attitude algorithm and getting gait information. Next the discrete Fourier transform is used to extract the frequency information and finally the k Nearest Neighbours classification algorithm is employed to identify the pace level. Through the testing and experimentation, it proves that this human gait and pace detection system is stable at working.