Conflict(free(trajectory(op0misa0on(for(complex( - PowerPoint PPT Presentation

Conflict(free(trajectory(op0misa0on(for(complex( departure(procedures(( 6 th (ICRAT(Conference,(Istanbul,(Turkey( San$%Vilardaga ( Xavier%Prats% san0.vilardaga@ctae.org( xavier.prats@upc.edu( ASCAMMFCTAE,(Aerospace(Research( UPC,(Technical(University(of(Catalonia,( and(Technology(Centre,(Barcelona( Barcelona( Taking research further

Introduc0on( • SESAR(and(NextGen(aim(at(improvement(of(air( traffic(efficiency( • Hence(new,(enhanced,(opera0ons( – Con0nuous(Climb(Departures((CCD)( – Con0nuous(Cruise(Climb((CCC)( – Con0nuous(Descent(Approaches((CDA)( • However,(…(

Problema0c( Huge(variety(of(aircraT( and(configura0ons( Uncertainty(along( conflic0ng(areas( Impact(on(Air(Traffic( Capacity(

Proposed(solu0ons(in(the(literature( • Dynamic(speed(requests( • Mul0ple(different(FPA(phases( • Requested(Time(of(Arrival((RTA)(at(a(specific( point( • etc.( • All(accep0ng(subFop0mal(trajectories(

Objec0ve( • Dynamic(4D(trajectory(op0miser( • Complex((and(realis0c)(lateral(and(ver0cal( trajectories( – Waypoints( – Flight(phases( • Self(separa0on(from(surrounding(traffic(

Methodology(

Methodology( Ini0al(climb( Clean(config.( Constant(Mach(

Methodology( • Op0mal(Control(Theory( – Dynamics( – State( – Control( – Objec0ve( – Constraints( – Solved(with(colloca0on(methods(and(NLP(

Dynamics( • PointFMass(model( • Calm(winds( • Flat(nonFrota0ng(earth( • Specific(A320(performance(model( – Aerodynamic(and(propulsive(forces( – Fuel(consump0on(

State(Variables( x = [ v γ χ n e h ] 1 = v ˙ = m ( T − D − mg sin γ ) g = ˙ = v ( n z cos φ − cos γ ) γ g sin φ = ˙ = cos γ n z χ v = n ˙ = v cos γ cos χ = e ˙ = v cos γ sin χ ˙ h = v sin γ = s ˙ = v cos γ

Control(Variables( u = [ n z φ π ] . n z = L mg

Objec0ve(Func0on( Z t f J ( t ) = FF ( x , u , p , t ) d t. t 0 ✓ ◆ X X 3 3 ◆ i √ ✓ N 1 X X c F F M j FF = n e δ θ √ ij θ i =0 j =0

Opera0onal(Constraints( Constraint Definition Operating airspeeds V MCA ≤ v CAS ( t ) ≤ V MO No deceleration allowed v ( t ) ≥ 0 ˙ ˙ No descent allowed h ( t ) ≥ 0 Procedure Design Gradient (PDG) h ( t ) ≥ 0 . 033 s ( t ) Load factor 0 . 85 ≤ n z ( t ) ≤ 1 . 15 − 25 � ≤ φ ( t ) ≤ 25 � Bank angle

General(problem(formula0on( • Ini0al(guess((C++)( • Discre0sa0on(with(direct(colloca0on(method( (GAMS)( • NLP(solver(interface((GAMS)( • Results((C++)(

General(problem(formula0on( !% Phase(1( Phase(2( Phase(3( 30(s( 60(s( 50(s( 40(s( 70(s( 67(s( WP1( WP2( WP3(

Separa0on(assurance( • Self(separa0on(strategies((ASAS)( • Collabora0ve(scenario( – Fully(coopera0ve( – Semi(coopera0ve( – Non(coopera0ve( • Target(modelling( • Separa0on(geometry(

Separa0on(assurance( Flight Plan Intent information initial state n o e o Trajectory IntruderTrajectory REAL Position ADS-B Out h o Optimiser Speed t o Out Intruder Ownship IntruderTrajectory PRED In n o Spline fitting e o Trajectory Trajectory Intent info. ADS-B Out h o Predictor Position Optimiser t o Speed initial state Estimated Intruder OwnshipTrajectory Performance Flight Plan

Separa0on(formula0on( !%

Intruder(representa0on( • Separate(a(moving(vehicle(from(a(moving( target( • Separa0on(of(ownship(4D(posi0on(with( respect(to(target(4D(geometry(…( • …(at(specific(ownship(0mes(of(sample!( • Use(of(polynomial(fifng((splines)(

Intruder(representa0on( • n intruder (t)( 0me( • e intruder (t)( 0me( • h intruder (t)( 0me(

Separa0on(geometry( • Horizontal(vs(Ver0cal(disjunc0on( (Cylinder( g h ( t ) ≥ 0 ∨ g v ( t ) ≥ 0 (( (Superegg( ✓ ∆ n 2 + ∆ e 2 ◆ p ◆ p ✓ ∆ h 2 + ≥ 1 2 2 d h d v

Fixed(lateral(routes( • Due(to(constraints(in(ATC(or(strategic( deconflic0on(

Results(

Results( 10( 9( 8( 7( 6( 5( 4( 3( 2( 1( 0( 260( 280( 300( 320( 340( 360(

Results(

Results(

Conclusions( • Generic(op0misa0on(framework( • Op0mal(control(formula0on( • Flexible(defini0on(of(complex(routes( • SemiFcoopera0ve(separa0on(assurance( • Can(be(integrated(in(many(scenarios(

Further(work( • Conformance(monitoring( In (from intruder) Ownship Time to Compute conflict ADS-B Threshold Expected system error Pos Aircraft category Pos Pos+Vel Intent info. Pos+Vel Intent info. Estimated Residuals Trajectory Conformance Continue Intruder Trajectory No intruder > Predictor Monitoring as planned performance Threshold Yes Ownship Continue No Intents Yes Replanning performance change? as planned data New Ownship Trajectory

Further(work( • Conformance(monitoring( 600 Prediction at t=95s Vertical error (ft) Initial prediction (t=0) Prediction at t=171s 450 Aggregate prediction Prediction at t=273s 300 150 0 5 10 15 20 25 30 Along path distance (NM)

Further(work( More(info(in(ATIO(Conference( June(2014,(Atlanta,(Georgia(

Thank(you(for(your(aken0on( Any(Ques0ons?( San$%Vilardaga ( Xavier%Prats% san0.vilardaga@ctae.org( xavier.prats@upc.edu( ASCAMMFCTAE,(Aerospace(Research( UPC,(Technical(University(of(Catalonia,( and(Technology(Centre,(Barcelona( Barcelona( Taking research further