Application of frequency-domain integral algorithm to the processing of sport signals based on MTi sensor

作者：Yu Qian^*; Ge Yunjian; Ma Tingting; Lu Zeyuan

来源：Sensors and Transducers, 2012, 138(3): 14-21.

• Acceleration signals
• Attitude sensors
• Frequency domains
• High precision
• Human bodies
• Integration accuracy
• Low frequency
• Low-frequency components
• Low-Frequency Noise
• Low-frequency signals
• MEMS sensors
• Mti
• Time-domain integrations

MTi sensor is a new kind of MEMS sensor that can output high precision values of three dimensional acceleration and angular velocity. The acceleration signal of human body is low frequency signal. There exist some noises of low frequency in the data acquired by MTi sensor. Integration is sensitive to low frequency components, thus such noises have great impact on the integration result. Over time the cumulative small low-frequency noises will be amplified. The lower the frequency is, the more obvious will the impact be. Therefore, time-domain integration has large errors. Undertaking this task in frequency domain by Fourier transform can avoid such amplification. In this paper, the frequency-domain integral algorithm can also filter out some low frequency noises thus improve the integration accuracy by selecting appropriate cut-off frequency. Successful application to simulation and real sport signals show that the algorism is effective.

• 出版日期2012