Estimation of lower limbs angular positions using Kalman filter and genetic algorithm

This paper presents an application of filtering in the field of robotic rehabilitation. The proposed system is being developed to estimate the angular positions of an impedance-controlled exoskeleton for lower limbs, designed to provide motor rehabilitation of stroke and spinal cord injured people. A Kalman filter based on genetic algorithm is used in a sensor fusion strategy for estimation of the angular positions, whereas Kalman filter fuses the data from inertial sensors and genetic algorithm tunes the weighting matrices of the filter. Also, to properly use accelerometers in a position estimation strategy, the measured acceleration must be close to the gravity acceleration. In this paper, we use the three components of the three-dimensional accelerometers to ensure that they are measuring only the gravity vector. We compare the proposed system with our previous sensor fusion system where force sensors located in an insole system was used for gait-phase identification, giving the periods where the foot was in full contact with the ground and the one-dimensional accelerometer measurements are suitable for position estimation. Simulation results validate the effectiveness of this proposal.