With growing air traffic demands, one proposal to reduce overall controller workload is to delegate responsibility for maintaining separation to the cockpit. One suggested implementation of this concept is a control law that can automatically maintain separation in a string of in-trail aircraft, with, or without pilot intervention or oversight. In this paper we explore the development of control laws that can aid in the maintenance of separation in strings of aircraft. A string of aircraft is assumed to consist of a lead aircraft which is following a nominal route and flight profile. The nominal route includes descent, speed changes and turns. Following aircraft are required to maintain separation from aircraft ahead in the string. Each following aircraft in the string has identical control law. The controller used is based on Proportional-Derivative (PD) action. The effect of feedback from different aircraft is studied to determine the requirements to maintain safe separation. A simulation of an Automatic Dependence Surveillance – Broadcast (ADS-B) signal is used to communicate state vectors between aircraft in the string. The numerical approach was to simulate the dynamics of aircraft within MATLAB using linearized aircraft equations of motion. Using these dynamics, the control law aimed at maintaining the aircraft on their course while maintaining a fixed separation distance. Additional nonlinear logic is added to maintain speed and thrust commands within acceptable limits. The program was then run to test strings of aircraft in descending and turning flight trajectories using several feedback schemes. The results show that the control algorithm and ADS-B transmission to determine state information are sufficient to allow following aircraft to maintain separation in a descending, variable speed string of aircraft. Generally, in a long string of aircraft, feedback of separation from only the immediately preceding aircraft may be insufficient to maintain separation. Further tests are required in a total nonlinear simulation are required to validate these conclusions, to investigate other simulation scenarios, and to optimize the control schemes chosen.
The different versions of the original document can be found in: