本實驗藉靈敏度g/1333 為1count單位、測量計錄頻率為0.005～50Hz之高精密三軸加速規於角力選手執行靜坐、走路、慢跑、快跑等身體活動，以頻率20.Hz精確計錄身體活動於三軸產生之加速情況。算得身體活動之能量消耗設為自變項，同時與效標呼吸儀測得之能量消耗定為依變項，求得皮爾森相關係數（Pearson correlation）之實驗設計；結果顯示高精密三軸加速規與呼吸儀之R平方＝.924, 估計標準誤（SEE）＝.084,P < .05達顯著水準。並與呼吸儀測得之代謝當量（MET）依線性回歸統計得能量消耗預測方程式METs= .055counts/分＋ .240；之前學者研究步頻、體重將是影響能量消耗之因子之一，本實驗首度將步頻依多重回歸之統計，步頻取加速規計錄身體活動之垂直軸之十秒算出步頻，得能量消耗預測方程式METs＝13.137 count/min＋1.371Sf－3.120，Sf為步頻（步/秒），估計標準誤（SEE）＝1.302，F（2,7）＝140.815，P＝.000呈顯著差異，更高之R平方＝.943,較簡單線性回歸多解釋變異數2%；不過，體重的多重線性回歸中在本實驗之統計結果並未達顯著標準，故本實驗在角力選手之能量預測方程式中不納入體重因子。結論：本實驗之樣本只針對角力選手完成實驗設計之身體活動，的確，高精密三軸加速規與呼吸儀具有相當高之統計關係，試值得信賴之能量消耗評估工具之一；希望能量預測方程式，對之後研究有相當大的助益。

　　The experiments by sensitivity g/1333 for 1count unit, the gauge recorded a frequency of 0.005 ~ 50Hz high-precision three-axis accelerometers wrestling players perform sit-ins, walk, jog, run physical activity frequency 20.Hz accurate count recorded of physical activity accelerated in triaxial generated. Regarded physical activity energy consumption is set to the independent variable, while respirometer measured criterion of energy consumption as the dependent variable, obtained Pearson correlation coefficients （Pearson correlation） experimental design; results show that high-precision three-axis accelerometer regulation respirometer R-squared = .924, the estimated standard error （SEE） = .084, P <.05 significant level. And breathing instrument measured the metabolic equivalent （MET） is obtained by linear regression statistics energy consumption prediction equation METs = .055 counts / min + .240; before the scholars stride frequency, weight will be one of the factors that affect energy consumption, this experiment the first step frequency step frequency to take physical activity accelerometer count recorded ten seconds of the vertical axis calculated by multiple regression statistics, step frequency, in energy consumption prediction equation METs = 13.137 count / min +1.371 Sf-3.120, Sf steps frequency （step / sec）, the estimated standard error （SEE） = 1.302, F （2,7） = 140.815, P = .000 was a significant difference, the higher the R-squared = .943, relatively simple linear regression model explained variance the number of 2%; However, the weight of multiple linear regression, the statistical results of this experiment did not reach significant standard in the experiment are not included in the weight factor in the the wrestling athletes' energy prediction equation. Conclusion: In this study, samples only for wrestling players to complete the experimental design of physical activity, indeed, high precision triaxial accelerometers and breathing instrument has a very high statistical relationship trial worthy of the trust of the energy consumption assessment tools; hope the energy prediction equation after study quite helpful.