Effective Optimization of Deployment for Wearable Sensors in Transfemoral Prosthesis

Abstract: Transfemoralor above-the-knee amputees face discomfort in their prothesis primarily due to irregular distribution of pressure and shear forces in the Socket-stump interface (SSI). To quantify this discomfort it is necessary to first determine the pressure distribution in the SSI using sensors. However, knowledge of how sensors should be deployed is necessary to support the testing of said pressure on a test-rig or amputee. Previous methods used to determine sensor placement include discretization of the SSI into several regions or the use of a reiterative method based on pressure readings from sensors to determine the optimal placement of sensors. The former fails to identify high regions of pressure as the regions covered by the sensors may not have high pressure whereas the latter is time consuming and may cause further trauma to amputees as it requires repeated experimentation. With the advances in pressure sensor technologies, biomechanical simulations, and Finite elementanalysis(FEA)simulations it is now increasingly possible to determine an accurate estimate of dynamic pressure distribution occurring in the SSI during the gait cycle. The thesis investigates the dynamic pressure distribution in the SSI and determines an effective method of locating the optimal positions for the sensors using two different algorithms. The first is a Genetic Algorithm whereas the second is Pattern Search.