Co-Regulation of Computational and Physical Effectors in a Quadrotor Unmanned Aircraft System

Traditional control strategies rely on real-time computer tasks executing in fixed intervals providing periodic sampling upon which discrete controllers are designed. But emerging trends challenge this fixed resource allocation strategy by sampling at the "right" time rather than at fixed intervals. We propose a strategy in which a model representing the sampling rate is augmented to the state-space model of a quadrotor unmanned aircraft system, coupled controllers are designed for this holistic system, and computational and physical effectors are co-regulated in response to system performance. We investigate a new discrete-time-varying control strategy by gain scheduling a discrete linear quadratic regulator controller at a series of sampling rates, and co-regulating the sampling rates using a cyber controller whose gains are optimized via a strategic cost function. We then show step responses of the quadrotor to demonstrate how rapid changes in physical system gain at discrete sampling rates negatively impacts system performance. To solve this we introduce a new cyber control strategy that reduces these negative impacts and show how the response can be improved. Since most multicopters employ waypoint tracking planning and guidance, we also evaluate our strategy by assessing performance of the quadrotor in following a waypoint trajectory giving a much better indication of how a control strategy affects mission performance. We develop cyber-physical metrics for assessing waypoint following performance and use them to improve controller design. Results show that our proposed coupled cyber-physical system model and controller can provide physical system performance similar to fixed-rate optimal control strategies but with less control effort and much less computational utilization. Our strategy allows cyber and physical resources to be dynamically allocated to system demands as needed.