PROBLEM TO BE SOLVED: To provide an awakening state determination method having high determination accuracy of an awakening state.SOLUTION: The awakening state determination method uses a respiration sensor for acquiring respiration data which is varied by respiration of a seating person, and a controller having a calculation part for calculating the respiration data acquired from the respiration sensor and a determination part for determining a state of the seating person. The method comprises: a step in which the respiration sensor acquires the respiration data of the seating person (step S11)； a step in which the calculation part calculates P_vel of the acquired respiration data by the calculation (step S16)； and a step in which the determination part sets probability in which drowsiness occurs on the seating person about P_vel as likelihood, uses a Bayesian filter in which the likelihood and the prior probability in which drowsiness occurs are multiplied, for determining the awakening state of the seating person (steps S18, S19, S20, S21).SELECTED DRAWING: Figure 2【課題】覚醒状態の判定精度の高い覚醒状態判定方法を提供する。【解決手段】覚醒状態判定方法は、着座者の呼吸によって変動する呼吸データを取得する呼吸センサと、呼吸センサから取得された呼吸データを演算する演算部、及び着座者の状態を判定する判定部を有するコントローラと、を用いる。呼吸センサが、着座者の呼吸データを取得する呼吸データ取得工程（ステップＳ１１）と、演算部が、取得された呼吸データのＰ_ｖｅｌを演算により算出する演算工程（ステップＳ１６）と、判定部が、Ｐ_ｖｅｌについての着座者に眠気が生じる確率を尤度とし、尤度に眠気の生じる事前確率を乗じるベイズフィルタを使用して、着座者の覚醒状態を判定する覚醒判定工程（ステップＳ１８，Ｓ１９，Ｓ２０，Ｓ２１）と、有する。【選択図】図２
