MIT ICAT MIT ICAT T RAJECTORY P REDICTION FOR S ELF S EPARATION DURING D ECELERATING A PPROACHES T RAJECTORY P REDICTION FOR S ELF S EPARATION DURING D ECELERATING A PPROACHES IN A D ATA- L INK E NVIRONMENT Alexander in ‘t Veld John-Paul Clarke D EPARTMENT OF A ERONAUTICS AND A STRONAUTICS M ASSACHUSETTS I NSTITUTE OF T ECHNOLOGY
MIT ICAT MIT ICAT T RAJECTORY P REDICTION FOR S ELF S EPARATION DURING D ECELERATING A PPROACHES Framework Airport noise impact problems New approach procedures Three Degree Decelerating Approach (TDDA) Continuous Descent Approach (CDA) Advanced Continuous Descent Approach (ACDA) Benefits Significant reduction of noise footprint Pilots like it But…….. Dramatic reduction of runway capacity Reason Increased minimum separation Difficult for controllers to assess necessary initial separation
MIT ICAT MIT ICAT T RAJECTORY P REDICTION FOR S ELF S EPARATION DURING D ECELERATING A PPROACHES TDDA procedure Three Degree Decelerating Approach Initial approach 250 kts at 7000’ Idle thrust on glide slope intercept Flap scheduling to reach V ref at 500’ allowing for wind conditions Add thrust at 500’ to maintain V ref Continue for normal flare 3° 7000’ V ref Idle thrust 500’
MIT ICAT MIT ICAT T RAJECTORY P REDICTION FOR S ELF S EPARATION DURING D ECELERATING A PPROACHES Problems with in trail separation 3° 7000’ 500’ Dist to runway [NM] Time [sec] Separation [NM] Aircraft decelerating at different rates Closing or widening gap Difficult for ATC to assess required separation Source: Ho and Clarke, 2001
MIT ICAT MIT ICAT T RAJECTORY P REDICTION FOR S ELF S EPARATION DURING D ECELERATING A PPROACHES New technology State vector broadcast Identification Position (lat/long) Altitude Air-vector Ground-vector Intent information 1 sec update cycle Info available to both ATC and FMS Implemented after 2005 Automatic Dependant Surveillance Broadcast (ADS-B) Data-link between aircraft, ATC and ground vehicles Source: RTCA, 2002
MIT ICAT MIT ICAT T RAJECTORY P REDICTION FOR S ELF S EPARATION DURING D ECELERATING A PPROACHES Proposed scenario After passing the FAF Pilot A Monitors trajectory aircraft B Maintains separation from aircraft B ATC Remains responsible Monitors conformance Monitors separation Issues separation warnings Information flows State vector Air- and ground-vector Intent information (V ref ) Separation compromise warning A B
MIT ICAT MIT ICAT T RAJECTORY P REDICTION FOR S ELF S EPARATION DURING D ECELERATING A PPROACHES Requirements / Scope No unacceptable increase in pilot workload Separation monitoring algorithm Own trajectory prediction (Koeslag, 2001) Trajectory prediction of preceding aircraft Separation is extra constraint for flap scheduling Display for separation monitoring In the cockpit On ATC workstation Based on ADS-B data Avoid the need for a complete aerodynamic database of all aircraft Detailed wind information
MIT ICAT MIT ICAT T RAJECTORY P REDICTION FOR S ELF S EPARATION DURING D ECELERATING A PPROACHES Boeing model Point mass dynamic model Validated on B research simulator (Koeslag, 2001) Automatic flap scheduling To ensure V ref at 500’ To cater for varying wind conditions flap [deg] V [kts] x [NM] Time [sec]
MIT ICAT MIT ICAT T RAJECTORY P REDICTION FOR S ELF S EPARATION DURING D ECELERATING A PPROACHES Wind model V [m/s] h [m] V nom Nominal profile: 60 kts at 4500’ Turbulence White noise variation Max 0,1 m/s 2 Source: Huber, 2002
MIT ICAT MIT ICAT T RAJECTORY P REDICTION FOR S ELF S EPARATION DURING D ECELERATING A PPROACHES Least Square Error V [kts] x [NM] Time [sec] V [kts] x [NM] Time [sec] no windwith wind Timestep = 25 sec
MIT ICAT MIT ICAT T RAJECTORY P REDICTION FOR S ELF S EPARATION DURING D ECELERATING A PPROACHES x [m] x dot [m/s] (x ref, V ref ) Intent information 7000’ h ref x ref Knowledge of V ref and h ref yields x ref (x ref, V ref ) forms an end constraint in the xV plane Solving the ODE gives expressions for x(t) and V(t)
MIT ICAT MIT ICAT T RAJECTORY P REDICTION FOR S ELF S EPARATION DURING D ECELERATING A PPROACHES Approximating x dot = f(x) x [m] x dot [m/s] (x ref, V ref ) Time
MIT ICAT MIT ICAT T RAJECTORY P REDICTION FOR S ELF S EPARATION DURING D ECELERATING A PPROACHES LSQ + end constraint V [kts] x [NM] Time [sec] V [kts] x [NM] Time [sec] no windwith wind Timestep = 25 sec
MIT ICAT MIT ICAT T RAJECTORY P REDICTION FOR S ELF S EPARATION DURING D ECELERATING A PPROACHES Preliminary results Based on ADS-B state-vector and intent information alone it is possible to make satisfactory trajectory predictions during decelerating flight Predictions based on only the state vector history are not accurate enough for separation purposes Better results may be obtainable with nominal flap schedule information
MIT ICAT MIT ICAT T RAJECTORY P REDICTION FOR S ELF S EPARATION DURING D ECELERATING A PPROACHES Future research Algorithm Incorporate flap schedule Sensitivity analysis Runway capacity estimates Displays design Primary Flight Display Navigation Display Piloted simulation Pilot rating (Harper-Cooper / TLX) Runway capacity estimates
MIT ICAT MIT ICAT T RAJECTORY P REDICTION FOR S ELF S EPARATION DURING D ECELERATING A PPROACHES T RAJECTORY P REDICTION FOR S ELF S EPARATION DURING D ECELERATING A PPROACHES IN A D ATA- L INK E NVIRONMENT Alexander in ‘t Veld John-Paul Clarke D EPARTMENT OF A ERONAUTICS AND A STRONAUTICS M ASSACHUSETTS I NSTITUTE OF T ECHNOLOGY