- 學位論文
穿戴式下肢輔具於脊損病人之上下樓梯及上下坡之意向偵測
Motion Detection for Paraplegic Patients with a Powered Lower Limb Orthosis in Ascending and Descending
摘要
本文藉由觀察一般健康人下肢活動方式，分析探討穿戴下肢輔具之脊髓損傷病患上下樓梯或上下斜坡的步態週期後，分解定義其步態週期之動作提出穿戴下肢輔具之脊髓損傷病患上下樓梯及上下斜坡之動作意向偵測演算法，發展對應之上下樓梯及上下斜坡的步態軌跡控制法則，最後整合主動式輔具以輔助脊髓損傷病患完成上下樓梯或上下斜坡的動作，提高患者的行動能力。整體主動式輔具融合加速度規對上半身行動之分析、放置於足底及拐杖下端之薄型壓力感測片對身體重量轉移之偵測及拐杖上所配置之陀螺儀感測拐杖之姿態角對使用者前方地勢高低之辨別等資訊，將其輸入類神經網路即時協助系統辨識使用者之動作意圖，並做為輔具上下樓梯與上下斜坡步態軌跡動作的控制依據，最後整合發展成一套具備人性、合理且安全之動作意向判斷演算法，以達成輔助脊髓損傷病患完成上下樓梯與上下斜坡的動作。
並列摘要
In this article, we present the ascending and descending gait cycle of spinal cord injured(SCI) patients by observing healthy people’s lower limb activities, and the algorithm to estimate human’s motion intention and descending for spinal cord injured patients. Previous studies focused on sit-to-stand and plane walking only. This article tried to assist spinal cord injured patients walking on ramps or stairs with a powered lower limb orthosis and a pair of forearm crutches, the system is designed to utilize three types of signals, one comes from the accelerometer to detect user’s upper body gesture, another comes from the flexi-pressure sensors which are placed under the crutches and a pair of shoes to measure weight-shift of the body, and a gyroscope is placed on the top of the crutch handle to detect the height of the terrain in front of the user. The collected signals are sent to several trained artificial neural network to classify user’s motion and trigger the corresponding gait trajectory. It is expected that spinal cord injured patients could walk on ramps or stairs with the proposed algorithm and incorporated gait trajectory as they wear a powered lower limb orthosis.