Abstract
An automatic estimation method of fuzzy rates using a pattern recognition technique has been developed.Three characteristics values(mean square error, peak ratio, and mean minimum distance)are used in the process to compare the fuzzy rates and target rates variations.In order to get the final numerical value for the objective estimation from these three values, fuzzy logic was used.The method was applied to the rate simulation during walking based on minute ventilation and oxygen saturation level as input guide parametersjEstimation results were very similar to those by the conventional subjective judgements.This method is found to be effective for the automatic tuning of membership functions and control rules.
