An inertial neural network approach for loco-manipulation trajectory tracking of mobile robot with redundant manipulator (Q6077643)

Consider the combined control and optimization of a mobile robot, consisting of a mobile platform and the manipulator. The main objects of the paper are: i) design of a trajectory tracking model based on the loco-manipulability of the mobile robot, ii) the construction of a control law, having an improved performance compared with known ones. Here, the objective function of the resulting optimization problem is based on the loco-manipulability and kinetic energy of the manipulator and mobile platform. Constrains are given for the velocity in \(z\)-axis of the mobile platform, which can move only on the horizontal plane. iii) For solving the resulting augmented Lagrangian primal-dual optimization problem, a neural network approach is applied. Moreover, the convergence behavior of the method is evaluated by simulation experiments.