Phytic Acid/MXene@Polyurethane Sponge-Based Flexible Pressure Sensor With Assistance of MC-GRU Model for Motion Posture Recognition

In this work, a polyurethane (PU) sponge was employed as a flexible substrate to achieve surface modification of 2-D MXene nanosheets with phytic acid (PA), and a PA/MXene@PU (PMP) sponge sensor was successfully constructed through a self-assembly process. As a flame retardant and binder, PA effectively enhances the conductivity and mechanical properties of PMP. This sensor demonstrates high sensitivity (16.71 kPa ${}^{\text {-1}}\text {)}$ , a wide detection range (0–175 kPa), and stability exceeding 8000 cycles. It can effectively detect the pressure changes triggered by water droplets falling from a height of 2 cm and 20 mg NaCl aqueous solution impacts (0.01 N). Furthermore, its excellent flame retardancy enables self-extinguishing within 20 s after ignition. In addition, this article integrates PMP with a gated recurrent unit convolutional neural network to achieve recognition and classification of six joints—ankles, wrists, fingers, throat, knees, and elbows—with a classification success rate of up to 99.83%. The results demonstrate the immense potential of PMP sponge sensors in physiological signal detection and human motion monitoring. This work exhibited a promising application of wearable flexible pressure sensors for motion posture recognition with asistance of intelligent perception.