top of page
Iterative Learning Control (ILC) algorithm to study how the user of a Powered Lower Limb Orthosis (PLLO) could learn to coordinate with a robust finite-horizon LQR driving the torque at the hips of the robot for safely performing a sit-to-stand (STS) movement, regardless of parameter uncertainty, and factors deliberately introduced to hinder learning.
Compute over-approximations for the reachable sets of the state, input, and output of a three-link planar robot used to model the sit-to-stand (STS) movement of a Powered Lower Limb Orthosis (PLLO) in the presence of constant parameter uncertainty.
Technique for assessing the robustness against parameter uncertainty of a batch of finite time horizon LQR controllers designed for tracking the sit-to-stand (STS) movement of a powered lower limb orthosis (PLLO).
Matlab implementation of a finite time horizon Linear Quadratic Regulator (LQR) for the sit-to-stand (STS) movement of a minimally actuated Powered Lower Limb Orthosis (PLLO) at the hips.
The motion planning strategy relies on a transformation that maps the desired trajectory for the Center of Mass (CoM) of the system into reference trajectories in the space of the angular positions of the links of a three-link planar robot. The resulting sit-to-stand movement is illustrated with the implementation of a tracking controller based on feedback linearization, and control allocation.
Random sampling within an n-dimensional L-p ball of radius r, centered at c.
Samples within L-2 balls are obtained by mapping values from Latin Hypercube Sampling (LHS) into spherical coordinates of n-dimensional hyperspheres. Samples of L-1 balls come from applying a rejection method on the points sampled from an L-2 ball of the same radius. Samples of L-Inf balls are given by isotropic scaling of vectors from LHS.
Matlab tools for performing the Kriging interpolation of a function f: R -> R. A quadratic regression is used to fit the sample points, and a Gaussian function for defining the covariance.
Matlab tools relying on dynamic programming for obtaining a closed-loop controller that aims to minimize the Total Travel Time (TTT) of the traffic flow in a three-link merge junction, an elemental component in transportation networks.
Fractal Generator in C++ written together with Miguel Barousse-Ordonez.
Resume
bottom of page