Performance Parameters of Speed Control

The effect of speed control to resolve aircraft conflicts has not been well described in literature. Previous investigations on speed control have suggested some performance parameters such as resolution rates. In this paper additional performance parameters such as resolutions per flight, per conflict cluster and constraining aircraft are investigated by the comparison of speed control with other resolutions. This will help to give a better understanding of the effects of speed control on the overall air traffic management system regarding capacity, feasibility, safety, economy and ecology. I Speed Control And Automation Speed control is one of the measures that air traffic controllers may use in conflict resolution. However, clearances involving speed are almost unused in European upper en-route sector control (<<1%) [1], whereas transfers, level changes, direct routing and vectors are most commonly used. Speed control has not been studied for the purpose of automation yet, or at least there is no significant literature. The literature that treats algorithms for automatic conflict resolution mainly discusses horizontal vectors and, to a lesser extent, vertical resolutions, but very few suggest the use of speed adjustments. Only some project documentation treats speed manoeuvres, e.g. the CORA project [2]. Also, the operational concept of Airborne Separation Assistance (ASAS) investigates the possibility of chaining and 1 Presented at the 24 Digital Avionics System Conference, October 2005, Washington D.C., U.S.A. spacing aircraft in sequence for easier handling of (arrival) flows [3], so-called station keeping, with a consequence of similarities in speed of the targets, but is not explicitly giving speed instructions. The same spacing principles apply to the procedures over the Atlantic Ocean that space aircraft on in-trail sequences on similar speeds or with a specific radar separation buffer. All Arrival Manager tools set time constraints on arrival fix points, and aircraft can accomplish the time contract with speed adaptations [4][5]. This concept of metering fixes can further be generalised into the en-route phase to organise aircraft into flows, e.g. [6][7]. Those flow organisers do not include a conflict resolution function using explicitly speed, at least this has not been found documented. The procedures of sequencing aircraft with speed, however, do differ from generic speed instructions, because they only use one dimension of the geometry (along-track conflicts). If, instead, speed-instructions are used for all kind of conflict resolutions, they will be used in all dimensions of the possible conflict geometries. The Automated En-Route Air Traffic Control (AERA) concept [8] in its very early versions mentions speed restriction measures where the automation system tags traffic that should not be touched by the controllers and pilots, because it is recognised as not being in a conflict situation if nothing happens. This could be regarded as a “maintain speed and heading” clearance, hence a speed instruction that is given by the system. J. Villiers [9] introduces the notion of speed adjustments as a means to make what he calls “subliminal” control, which could be considered as an automation system that creates “lucky traffic” so that controllers, who would still be part of the system, would not have to intervene on the traffic. The automation system would use speed adjustments for this, and the controllers would neither see nor have to know about hidden automatic operations, because they would not perceive the speed changes. The previous study on the potential of speed control [10] has given initially promising results. More knowledge has to be sought on this topic in order to better understand the effects of speed control. This paper analyses additional performance parameters with the help of model simulations and compares speed control with other resolution manoeuvres. II Simulation Setup The Reorganised ATC Mathematical Simulator (RAMS) was used in its version RAMS-Plus5.04 and 5.20. The scenario was reused from the 5 States Fast-Time Simulation [10] with an area extending from London/Paris in the west to Berlin/Prague/Vienna in the east and from Copenhagen/Malmö in the north to Lyon/Milan in the south. More than 140 sectors from 24 ATC Centres were simulated. The measured centres were limited to en-route and to Karlsruhe, Maastricht and Reims, which corresponds to 36 en-route sectors above flight level 245. The traffic baseline simulates a traffic sample from 12 Sep. 1997, which corresponds to 100%. This was increased to 150, 200 and 300%, which emulates roughly traffic loads for the year 2005, 2010 and 2025 for optimistic traffic growth predictions. 0 100 200 300 400 500 600 700 800 900 00 :00 02 :00 04 :00 06 :00 08 :00 10 :00 12 :00 14 :00 16 :00 18 :00 20 :00 22 :00 KARLSRUHE MAASTRICHT REIMS Figure 1: Number of aircraft per hour for three measured centres (150% scenario). Figure 1 shows the distribution of flights over the day for the three measured centres for the 150%-2005 traffic sample. 0.0% 2.0% 4.0% 6.0% 8.0% 10.0% 12.0% 10 -19 40 -49 70 -79 10 0-1 09 13 0-1 39 16 0-1 69 19 0-1 99 22 0-2 29 25 0-2 59 28 0-2 89 31 0-3 19 34 0-3 49 37 0-3 79 40 0-4 09 Figure 2: Distribution of aircraft over flight levels Figure 2 presents the distribution of flight levels for the entire scenario. 0 100 200 300 400 500 600