A Model Predictive Control Based Iterative Trajectory Optimization Method for Systems with State-Like Disturbances

摘要

Robots performing aggressive maneuvers in fluids must contend with disturbances caused by the dynamics of the fluids themselves. In this paper, we investigate the effects on system dynamics under disturbances that evolve in a state-like fashion, i.e., based on some internal state. Such a configuration has implications for the stability and controllability properties of the system, and has an effect on the optimality of any trajectories. We introduce an iterative method based on linear time-varying model predictive control to help deal with systems where the internal state of the fluid is not observable, but the disturbance is predictable. We then demonstrate the applicability of this method to optimal trajectory generation by planning a landing trajectory for a simplified planar quadcopter model.