Measurement in a field offers very valuable information for analyzing sports performance. It is, however, difficult to acquire well-conditioned data in a filed due to artifact noise during dynamic movement. We propose a field experiment assist system using moving telecommunication and a notebook computer. The notebook computer is connected to the engineering workstation (EWS) network via the modem and the moving telecommunication. EWS network provides artifact noise reduction and spectrum analysis at everywhere in a filed. Thus, we can obtain high performance service even in a field. Artifact noises are sometimes unavoidable in measurement of Myoelectric (ME) signals during dynamic movement. Artifact noise is caused by mechanical friction between electrodes and the skin. The frequency components of artifact noise are similar to those of ME signals. Thus, it is difficult to reject artifact noise using linear filters. We have proposed a nonlinear filter based artifact elimination method that consists of an autoregressive (AR) filter, a nonlinear filter, and the inverse-AR filter. These filters are also adaptive filters to perform in time-varying ME signals. To improve the performance for abrupt artifact noise, the filter coefficients adjustment using a fuzzy rule was employed. Customizing fuzzy rules, we achieved a better elimination performance.
　The experimental results showed that our system is very useful to proceed with field experiment in terms of well-conditioned data acquisition. The only problem was a low transmission rate. This will be overcame by developing a high speed modem or developing the digital moving telecommunication service area.

DESCENTE SPORTS SCIENCE Vol.15/THE DESCENTE AND ISHIMOTO MEMORIAL FOUNDATION FOR THE PROMOTION SPORTS SCIENCE