Estimation of ground reaction forces using forearm crutches instrumented with pressure sensors and accelerometers

Seylan, Çağlar
The use of crutches is critical for successful restoration of walking mobility through lower-body robotic orthoses that externally support and replace the functions of knee and hip joints. In this context, crutches can also provide useful sensory data, allowing the estimation of system state, postural stability, controller performance as well as user intention to regulate controller actions. In this thesis, we describe design and analysis of a crutch system instrumented with accelerometer and pressure sensors to estimate ground reaction forces on their point of contact, providing a well-defined sensory output for such applications. We propose an angle-dependent quadratic model to map pressure data to force components, which we identify using least-squares methods. First, we show performance of the model for specific crutch angles. Then, we evaluate the model and show the results for crutch angles other than those used for training. Finally, we present the evaluation and analysis of the model under dynamic conditions in which the crutch angle is varied by time.