Sitting posture recognition by body pressure distribution and airbag regulation strategy based on seat comfort evaluation

Unbalanced body pressure distribution caused by poor sitting posture is a significant factor causing lumbar disease. Research on the dynamic adjustment strategy for comfort sitting posture is of great significance to the prevention and improvement of lumbar disease. In this study, based on the pressure sensor array, the body pressure distribution in different sitting postures are obtained, including: back bow, waist bow, body slip, body right leaning, standard sitting position, and body left leaning. The recognition of the human body posture was achieved using the support vector machine algorithm and the accuracy rate reach to 83.33%. Furthermore, the evaluation indicators for sitting comfort of the human body are proposed using the maximum pressure, average pressure, maximum pressure gradient, average pressure gradient and asymmetry coefficient, the accuracy of the above indicators were verified using semantic differential subjective evaluation method. Experimental results show that there is a strong positive correlation between objective indicators and seat comfort. In addition, the dynamic adjustment strategy using airbag base on seat comfort evaluation is studied.