A Perspective on Prioritized Tracking Control for Robotic Systems with More Outputs than Inputs

Robot trajectory tracking is a pervasive requirement in many robotic applications and is an active area of research. There are a number of different approaches and algorithms available for tracking problems. However, for systems with more outputs than inputs, it is not possible to track all outputs commands independently of each other. In such a situation, one may require a subset of the outputs to be tracked more closely than others. In this paper, we explore the possibility of including an additional dynamic prioritization feature to existing linear tracking controllers. With the help of time-varying weighting matrices, one can prioritize a subset of outputs or a combination of outputs as per the needs of the application. A simple metric to quantify the expected tracking performance (incorporating the effect of prioritization) before actual implementation and measurement is also presented along with simulation results. Some preliminary results of using a similar approach of prioritized tracking in nonlinear systems using feedback linearization controllers is also presented. Finally, the use of prioritized tracking in case of quadcopter station-keeping problem under the influence of wind disturbances is demonstrated to highlight its practical utility.