Manyetoreolojik (Mr) Sönümlendiricili Diz Üstü Protezlerin Salınım Fazının Modellenmesi

Abstract

Above-knee leg prostheses are devices developed for individuals who have above knee amputations of their legs. Above knee prosthesis with semi-active MR damper consists of four parts: socket, knee, calf and foot. These prostheses control the MR damper and enable the prosthesis to move synchronously with the individual.

In the thesis, gait modeling and control of the gait of the subject using above knee prosthesis with a MR damper was performed. Within the scope of the study, MR damper and gait model were developed separately and combined to create a prosthetic gait model. In order to evaluate the application of different current values to the MR damper, the force, displacement and velocity of the damper under various current conditions were recorded using the force meter test device of ROEHRIG Company. Bingham, Dahl, LuGre and Bouc-wen parametric models were used to understand the MR damper behavior with a hysteresis-containing characteristic. Utilizing the set of data collected from the MR damper used for the prosthesis, the best parameter sets for each model under different current conditions were determined by pattern search optimization method. Moreover, the model parameters were composed as a function of the current for further analysis of the variation of the parameters according to the current levels. This is particularly important for calculation of damper forces according to current, displacement and velocity information. Based on the comparative study carried out in this part, the modelling error rates observed for Bingham, LuGre, Dahl and Bouc-wen models were 17.0%, 6.1%, 5.8%, 8.0%, respectively.

In the second part of the study, the gait model including the trunk - head, hip, calves and feet represented as solid segments is extensively examined. The dynamic equations of the segments were calculated using the Lagrange formulation. Assuming that the MR damper is connected between the calf and the knee section, the moment applied by the MR damper to the knee joint is added to the gait model equations. The developed model was operated using measured gait motion data and optimized MR damper parameters under Bingham model assumptions. This model was used to determine the current patterns required to be applied to the MR damper during the swing phase of the gait.

Through the simulations it is possible to run different control scenarios of the gait with a MR damper based prosthesis. The model gave expected reactions to different currents and different calf weights.