Optimization of Bipedal Walker Gaits with Constant Knee Stiffness

Russcher, K.J. (2014) Optimization of Bipedal Walker Gaits with Constant Knee Stiffness.

Abstract:Energy efficient bipedal locomotion is one of the research areas of the Robotics and Mechatronics (RAM) group at the University of Twente. This MSc project continues that research with the optimization of bipedal walking gaits and the application of that optimization to a bipedal walker with constant knee stiffness. Previous students investigated bipedal walking gaits with the spring loaded inverted pendulum (SLIP) model. A swing leg was added to the model [1] and the swing leg trajectories were chosen by an educated guess. The hip and swing leg trajectories of the segmented-SLIP (S-SLIP) model were used as references for the control of a bipedal robotic walker [2] [3]. Simulations showed that the gait cycle behaviour of the SLIP and S-SLIP model was not similar to that of the robotic walker. Therefore, using SLIP or S-SLIP trajectories as references for the control of a robotic walker did not seems like a viable idea. One of the recommendations in [3] indicated the need for higher quality reference trajectories. The paper, presented in section 2, proposes a novel optimization method for the bipedal walker gaits with constant knee stiffness. Power of the used optimization method is that it optimizes over the state variables and inputs for the complete gait at once [4]. The bipedal walker has two legs, both with an upper and lower leg. There is a torsion spring on the knee. The parameters of the spring, the spring constant and the equilibrium position, are constant during walking. However, the spring parameters are part of the optimization. The walker is optimized in three steps, the distinction between the three is how the knee spring is used in the optimization. First, the knee spring constant is set to zero. Second, the knee spring constant is part of the optimization. Third, the knee spring constant and equilibrium position are part of the optimization. The optimizations show that the torque that is needed for a walking gait is reduced by adding the knee springs to the bipedal walker. Remarkable is that it also reduces the input torques to the hip joint that has no spring. The optimizations also show that the shape of the swing foot trajectories change with increasing gait velocities. So, imposing trajectories that have the same shape for different velocities, as was done before, will thus not result in optimal gaits with regard to energy efficiency. Section 3 presents details of the optimization that are needed for the implementation of the optimization. The conclusions of the MSc project are in section 4. Final recommendations are reported in section 5.
Item Type:Essay (Master)
Faculty:EEMCS: Electrical Engineering, Mathematics and Computer Science
Subject:50 technical science in general
Programme:Systems and Control MSc (60359)
Link to this item:http://purl.utwente.nl/essays/66499
Export this item as:BibTeX
HTML Citation
Reference Manager


Repository Staff Only: item control page