Applied receding horizon control of the Caltech Ducted Fan

Abstract

This paper details the application of a constrained receding horizon control strategy t o an indoor vectoredthrust flight experiment known as the Caltech Ducted Fan. The strategy is used to stabilize the experiment about one operating point, and step response and disturbance rejection are examined with different configurations and in comparison t o a gain-scheduled LQR controller. Issues related to non-zero computation times, choice of horizon length and terminal cost are discussed. IntroductionIn receding horizon control, an open-loop trajectory is found by solving a finite-horizon optimal control problem starting from the current state. The controls of this trajectory are then applied for a certain fraction of the horizon length, after which the process is repeated. See [6] for a good review of recent work in this field. This approach has been used in the process control industry successfully for some time, where dynamics are relatively slow, but the computing power required and the tendency for naive implementations to unstabilize a system have prevented its use on fast, stability critical nonlinear systems. Recently, however, theoretical results regarding stability properties of receding horizon controllers and ever increasing computing power have revived interest in the scheme.

A model-predictive approach to formation control of omnidirectional mobile robots

“…By our implementation, at initial time the internal clock of each agent has to be synchronized using clock synchronization. Furthermore, the sampling time is calculated based on the average past results [17]. Thus sampling time is varying.…”

Experimental Demonstrations of Real-Time MILP Control

“…The matrix A p propagates the state forward by a short time period to account for the computation delays between the measurement of x and the execution of the first plan step [11]. The requirement for y to start at 1 and end at 0 requires that the states pass through the target region at some step of the plan, thereby completing the mission.…”

“…However, application to systems with faster dynamics, such as those found in the field of aerospace, has been impractical until recently, with the on-going improvements in computation speed. Recent hardware work in this field includes the online use of NTG nonlinear optimization for control of an aerodynamic system [11,12] and of a formation of highly-nonlinear vehicles [13,14].…”

Online Control Customization via Optimization‐Based Control

“…The real-time trajectory generation framework, and the corresponding software, was developed by Milam and co-workers and has appeared in [21,23,25]. The implementation of model predictive control given in this chapter is based on the work of Dunbar, Milam, and Franz [8,10].…”