edxmorgan/diff_uv

Implements 6-DOF underwater vehicle dynamics as fully symbolic CasADi expressions, enabling automatic computation of gradients, Jacobians, and Hessians for optimization and control. Includes integrated tooling for parameter estimation via convex optimization (CVXPY/MOSEK), filtering (EKF), and reinforcement learning via Gymnasium environments, with code generation for C++/MATLAB/Python backends. Models added mass, hydrodynamic damping, Coriolis, and restoring forces under low-speed, symmetric vehicle assumptions with analytic forward and inverse dynamics solvers.