Abstract
This study proposes a method for evaluating relaxation during a tennis swing using a single IMU sensor. An IMU attached to the racket’s grip end recorded acceleration and angular velocity data, and racket head speed was calculated using rigid-body kinematics. Two features—the speed difference around impact and the swing speed at impact— were extracted, and a fuzzy inference system was constructed to estimate the relaxation level. The fuzzy system utilized triangular and trapezoid membership functions and five linguistic labels (“very low” to “very high”) for relaxation. Experimental results showed that relaxed swings exhibited higher head speeds and smaller post-impact deceleration compared to tense swings, consistent with the physiological knowledge that excessive muscle tension increases limb stiffness and reduces energy transfer efficiency. These findings demonstrate the feasibility of estimating swing efficiency using a single sensor. Future work will increase the number of participants and incorporate additional physiological sensors (e.g., EMG) to develop a movement-quality evaluation system applicable to sports coaching and rehabilitation support.
