Method for training to cope with a fault affecting one powertrain of a hybrid propulsion system

Abstract

A method for training a pilot to cope with a fault affecting one powertrain of a hybrid propulsion system for an aircraft. The aircraft includes, connected in parallel to a transmission unit, n powertrains (where n?2), including a first and a second powertrain that are heterogeneous in nature. It involves, during a flight of the aircraft, simulating a fault affecting the first powertrain while, at the same time as performing the simulation, checking the status of the n powertrains of the propulsion system. If a fault affecting one of the n powertrains is detected, the simulation is halted and the instantaneous power delivered by at least one of either the first or the second powertrain is increased so that the sum of the instantaneous powers delivered by the n powertrains is ? a minimum total instantaneous power required for the aircraft to continue its flight.

Claims

1. A method for training a pilot to cope with a fault affecting a powertrain of a hybrid propulsion system for an aircraft comprising n powertrains connected in parallel on a transmission unit, n being an integer greater than or equal to 2, including first and second powertrains that are heterogeneous in nature, the method comprising, during a flight of the aircraft, simulating a fault affecting the first powertrain by implementing the following steps: decreasing the instantaneous power P.sub.M1nst delivered by the first powertrain down to a training power P.sub.M1Ecol and maintaining this power P.sub.MIEcol until the end of the simulation, with: P.sub.M2max_OEI>P.sub.M1Ecol>P.sub.M1min P.sub.M2max_OEI being the instantaneous maximum power that could be delivered by the second powertrain when not in the training mode and P.sub.M1min being the instantaneous minimum power that could be delivered by the first powertrain; and increasing the instantaneous power P.sub.M2inst delivered by the second powertrain up to a power that is lower than or equal to an upper limit power P.sub.M12lim_Ecol applicable to the second powertrain in the training mode, and regulating the power PM2inst during the simulation so that the instantaneous total power P.sub.tot_Ecol delivered by the first and second powertrains in the training mode is lower than or equal to P.sub.M2max_OEI, with: $P_{\mathrm{tot}\_\mathrm{Ecol}}=P_{M1\mathrm{Ecol}}+P_{M2\mathrm{inst}}{P}_{\mathrm{tot}\_\mathrm{Ecol}}?P_{M2\max \_\mathrm{OEI}}{P}_{M2\mathrm{inst}}?P_{M2{\lim }_{-}\mathrm{Ecol}}<P_{M2\max \_\mathrm{OEI}}{P}_{M2{\lim }_{-}\mathrm{Ecol}}+P_{M1\mathrm{Ecol}}=P_{M2\max \_\mathrm{OEI}}$ P.sub.M2lim_Ecol being the maximum power that could be delivered by the second powertrain in the training mode so that P.sub.tot_Ecol does not exceed P.sub.M2max_OEI; the method further comprising, at the same time as performing the simulation, checking the status of the n powertrains of the propulsion system and, if a fault affecting one of the n powertrains is detected, stopping the simulation and increasing the instantaneous power delivered by at least one amongst the first powertrain and the second powertrain, so that the sum of the instantaneous powers delivered by the n powertrains is higher than or equal to P.sub.Rmin_OEI, P.sub.Rmin_OEI being a minimum total instantaneous power required for the aircraft to continue its flight.

2. The method according to claim 1, wherein the second powertrain is selected from among a hydraulic or electrical type powertrain, and the first powertrain is selected from among a gas turbine type powertrain.

3. The method according to claim 1, wherein, the second powertrain being reversible, the step of increasing the instantaneous power P.sub.Minst delivered by the second powertrain is preceded by a step of drawing, by the second powertrain, a portion of the instantaneous power P.sub.M1 delivered by the first powertrain to the transmission unit, whereby a faster drop in the instantaneous total power P.sub.tot_Ecol delivered by the first and second powertrains during the simulation is obtained.

4. The method according to claim 1, wherein the step of decreasing the instantaneous power P.sub.M1inst delivered by the first powertrain includes a transient reduction in the power of the first powertrain below P.sub.M1Ecol, followed by an increase in the power of the first powertrain to P.sub.M1Ecol.

5. The method according to claim 1, wherein the triggering of the step of increasing the instantaneous power P.sub.M2inst delivered by the second powertrain is delayed and/or the increase of the instantaneous power P.sub.M2inst delivered by the second powertrain is slowed, whereby a transient power hole is created.

6. The method according to claim 1, wherein, the second powertrain being reversible and P.sub.M1Ecol being selected so as to be higher than or equal to P.sub.Rmin_Ecol (P.sub.Rmin_Ecol being the minimum total instantaneous power required for the aircraft to continue its flight in the training mode), a step of drawing a portion of the power delivered by the first powertrain to the transmission unit is carried out, by the second powertrain, at least once during the step of increasing the power delivered by the second powertrain, the maximum portion P.sub.M2min_Ecol that could be drawn being a negative value and being equal, in absolute value, to the maximum power that the second powertrain could draw from the transmission unit in the training mode, with P.sub.M1Ecol+P.sub.M2min_Ecol?P.sub.Rmin_Ecol.

7. The method according to claim 6, wherein the power P.sub.M1Ecol of the first powertrain and the power limit of the second powertrain P.sub.M2lim_Ecol are adapted in real-time during the simulation, so that an average of the power of the second powertrain during the simulation is equal to a reference power PM2ref selected so as to guarantee a margin for piloting the aircraft, with P.sub.M2min<P.sub.M2ref<P.sub.M2lim_Ecol and P.sub.M2lim_Ecol(t)+P.sub.M1Ecol(t)=P.sub.M2max_OEI.

8. A device for training a pilot to cope with a fault affecting a powertrain of a hybrid propulsion system for an aircraft comprising n powertrains, n being an integer greater than or equal to 2, including first and second powertrains that are heterogeneous in nature and connected in parallel on a transmission unit, the device comprising control means configured to implement a training method according to claim 1.

9. An aircraft equipped with a hybrid propulsion system comprising n powertrains, n being an integer greater than or equal to 2, including first and second powertrains that are heterogeneous in nature and connected in parallel on a transmission unit, and a training device according to claim 8.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Other aspects, aims, advantages and features of the invention will appear better upon reading the following detailed description of preferred embodiments thereof, given as a non-limiting example, and made with reference to the appended drawings wherein:

[0053] FIG. 1 schematically shows the architecture of an example of a hybrid propulsion system with two powertrains in parallel and its regulation system according to the invention;

[0054] FIG. 2 shows an example of power profiles in the event of an actual fault affecting one of the powertrains of the hybrid propulsion system of FIG. 1;

[0055] FIG. 3 shows an example of power profiles in the event of a simulated fault according to the invention affecting one of the powertrains of the hybrid propulsion system of FIG. 1;

[0056] FIG. 4 shows an example of the power loss profiles in the event of a simulated fault affecting a powertrain according to two variants of the invention in comparison with a real fault;

[0057] FIG. 5 shows an example of power profiles of a simulation of a fault affecting the first powertrain according to the variant 1 of the invention;

[0058] FIG. 6 shows an example of power profiles of a simulation of a fault affecting the first powertrain according to the variant 2 of the invention;

[0059] FIG. 7 is a detailed view of the regulation system 5 of FIG. 1, according to the variant 3 of the invention;

[0060] FIG. 8 shows an example of power profiles of a simulation of a fault affecting the first powertrain according to the variant 3 of the invention.

DETAILED DISCLOSURE OF PARTICULAR EMBODIMENTS

[0061] The propulsion system used in the context of the invention is a system for generating and supplying propulsive power that is hybrid and redundant. In other words, the propulsion system is hybrid, because it includes at least two powertrains that are heterogeneous in nature and it is redundant, because these at least two powertrains are arranged in parallel on the transmission unit. This enables landing of the aircraft under satisfactory safety conditions in the event of a partial fault affecting one of the two powertrains. By partial fault, it should here be understood a fault that affects only one of the powertrains in parallel. Hence, the propulsion system as a whole is partially faulty since at least one of the redundant powertrains is functional.

[0062] A characteristic example of application is a parallel hybrid propulsion system of a helicopter, composed of a turboshaft engine and of an electric motor driving, both, the main and tail rotors.

[0063] In the following illustrative examples, the invention will be applied to a hybrid propulsion system composed of two independent powertrains in parallel (a so-called two-engine situation), namely a first powertrain with a gas turbine type engine and a second powertrain with an electric motor.

[0064] FIG. 1 shows the architecture of a hybrid propulsion system and its regulation system 5. In the particular application case illustrated in FIG. 1, the hybrid propulsion system includes two redundant powertrains, namely a main powertrain and a secondary powertrain, that we will respectively call first powertrain 1 and second powertrain 2. The two powertrains are independent and different in nature and they are redundant (it is also said that they are in parallel), i.e. they deliver mechanical power to the rotor 4 via a transmission unit 3 which adapts and sums up the powers of the two powertrains.

[0065] In this embodiment, the main powertrain 1 includes a heat engine 10 which could be a gas turbine, this main powertrain being designed so as to supply most of the power necessary for the flight of the aircraft. The secondary powertrain 2 includes one or more electric motor(s) (in this case, one single electric motor 20), and allows supplying a complementary power which essentially allows continuing the flight, in a restricted domain, until landing in satisfactory safety conditions. The primary function of this secondary powertrain 2 is to be able to overcome the fault affecting the main chain 1, while minimising the on-board additional mass. The maximum power that it can deliver is lower than or equal to the maximum power of the main powertrain.

[0066] In FIG. 1, N.sub.R* is the rotational speed setpoint of the rotor 4 (also called propeller); N.sub.M1

[0067] (N.sub.M2) is the measurement of the speed of the engine of the first (second) powertrain; C.sub.M1 (C.sub.M2) is the measurement of the torque delivered by the engine of the first (second) powertrain; P.sub.M1* (P.sub.M2*) is the power control of the engine of the first (second) powertrain. The data N.sub.R*, N.sub.M1, N.sub.M2, C.sub.M1, C.sub.M2 are delivered to the regulation system 5. The data P.sub.M1* and P.sub.M2* are respectively delivered to the heat engine 10 of the first powertrain 1 and to the electric engine 20 of the second powertrain 2. Each of the engines is connected by a shaft 6 to a main transmission unit 3, which will transmit the power of the engine(s) to the rotor. Each shaft is provided with a measuring device 7 allowing measuring the speed and the torque delivered by the engine to which it is related.

[0068] FIG. 2 shows the current total power delivered over time to the rotor of the aircraft of FIG. 1 (curve 3), in the event of an actual fault affecting the engine of the first powertrain. The curves 1 and 2 respectively represent the instantaneous power delivered over time by the first and second powertrains.

[0069] When there is a fault affecting the engine of the first powertrain (a fault symbolised by a flash), the power delivered by the first powertrain drops rapidly until complete stop. To overcome this fault, the power of the second powertrain is increased up to its maximum power P.sub.M2max, which may be denoted P.sub.M2max_OEI to explicitly signal the actual fault situation (when not in the training mode).

[0070] FIG. 3 shows the case of a simulated fault affecting the engine 10. Like in FIG. 2, the curves 1, 2, 3 respectively represent the instantaneous power delivered by the first powertrain, by the second powertrain and the current total power delivered over time to the rotor.

[0071] We initially start from an operating point where the current total power is higher than the maximum power of the engine deemed to be functional during the simulation phase. In this example, this consists of the engine of the second powertrain and the maximum power is therefore P.sub.M2max_OEI. Indeed, the goal of the training mode is to simulate a flight situation where the available total power is limited by this maximum power of the second powertrain P.sub.M2max_OEI.

[0072] When the training mode is selected by the pilot and the fault affecting the engine 10 is triggered, the regulation system 5 reduces the power delivered by the first powertrain as rapidly as possible, without turning it off. Thus, this rapid power reduction simulates the loss of power available to the rotor when the first powertrain fails.

[0073] To do so, the regulation system makes the engine 10 of the first powertrain decelerate at its maximum achievable deceleration rate without shutting off the combustion chamber. Thus, instead of reducing the power of the engine 10 until complete stop, the regulation system reduces the power of the engine 10 down to an intermediate power level PM1Ecol and it then keeps it constant until the end of the fault training operation. In other words, there is a deceleration transient phase (drop in initial power) of the engine 10 at the time of triggering of the training mode, then the power level is then maintained at a stabilised power level P.sub.M1Ecol, the deceleration transient phase of the engine 10 being independent of the level P.sub.M1Ecol at which it will then be maintained.

[0074] The selection of this power level P.sub.M1Ecol of the engine 10 is the major advantage of this invention.

[0075] This level P.sub.M1Ecol is selected, on the one hand, high enough for the engine 10 to preserve a sufficient acceleration capacity in order to be able to rapidly return to its maximum power should the engine 20 fail during the training operation. Hence, this allows improving the level of safety during the training phases. In general, the higher the power level P.sub.M1Ecol, the quicker the engine 10 will be reactivated where necessary. Hence, the objective is to set the highest possible power level PM1Ecol while complying with the maximum power level of the engine 20 (P.sub.M2max_OEI).

[0076] On the other hand, this level P.sub.M1Ecol is also selected not too high so that the engine 10 does not influence the behaviour of the propulsion system felt by the pilot. As a reminder, the constraints may be summarised as follows: [0077] P.sub.M1Ecol+P.sub.M2min_Ecol=P.sub.tot_min_Ecol=P.sub.Rmin_Ecol, to comply with the minimum power; and P.sub.M1Ecol+P.sub.M2lim_Ecol=P.sub.tot_max_Ecol=P.sub.M2max_OEI, to comply with the maximum power.

[0078] In this manner, the power of the engine 10 can be maintained constant. Hence, the variations in power of the rotor can be carried out entirely by the engine 20. Hence, the piloting behaviour is faithful to what the pilot would feel with the power delivered entirely by the engine 20.

[0079] FIG. 4 shows a comparison of the power loss profiles on a simulated fault and on an actual fault, the curve 1 representing the profile of an actual fault of the engine 10, the curve 2 representing the profile of the deceleration on the so-called anti-extinction limit of the engine 10 and the curve 3 representing the profile of a simulated fault according to the variant 1 of the invention. As illustrated in this FIG. 4, the deceleration on the anti-extinction limit of the engine 10 (curve 2) could, depending on the performances of the engine 10, be slower than the power drop observed in some fault cases (for example, in the case of tightening of the engine by loss of lubrication or cut-off of a fuel supply valve).

Variant 1: Use of the engine 2 in the brake mode during the fault transient

[0080] In the variant 1, we act on the power drop initial transient phase at the time of entry into the training mode.

[0081] As we have just mentioned, a possible limitation of the engine fault simulation by controlling a controlled a deceleration of the regime of the engine 10 is that this maximum achievable deceleration can be significantly slower than a real power loss related to a real engine fault.

[0082] If the powertrain 2 of the engine 20 is reversible, i.e. the engine 20 can draw from the mechanical power of the BTP (whether by braking the BTP to recharge a battery or by dissipating the electric power), The engine 20 can be used to make the power delivered to the rotor drop quicker by drawing mechanical power from the engine 10.

[0083] As illustrated in FIG. 5, this variant 1 consists in transiently controlling a negative power on the engine 20 (part of the curve designated by the circle 4), in order to obtain a drop in the total power that is more representative of the power profile resulting from a real case of engine failure.

[0084] It should be noted that there is no mention here of a transient power hole simulation, as it might be the case in the documents describing a method for simulating a fault OEI in a two-engine situation (cf., for example, the document [2]). This transient power hole simulation might be not necessary in the case of an electric hybrid propulsion system, since the engine 20 (electric) offers a much superior responsiveness than a gas turbine. Hence, this responsiveness intrinsic to the electrical technology can allow compensating very quickly for the power loss of the engine 10 and therefore eliminating, or at least greatly attenuating, the transient power hole following the fault. Nonetheless, the present invention can also simulate this transient power hole, without limitation. This can be done in three ways, which could potentially be combined: [0085] the regulation system can transiently reduce the power of the engine 10 below PM1Ecol, before returning to this level; [0086] the regulation system can also delay and/or slow down the power of the engine 20, so that the sum of the powers of the two motors is transiently lower than the maximum power that the second powertrain can deliver (denoted P.sub.M2max) or to the power required by the aircraft P.sub.Rmin_Ecol; [0087] according to the variant 1 proposed hereinabove, a negative power level can transiently be controlled on the engine 20, so that it draws power on the BTP. By acting on the duration of power drawing, it is possible to simulate a more or less long transient power hole before returning to the maximum power.

[0088] Moreover, the trade-off between the responsiveness of the engine 10 and the required minimum power P.sub.Rmin_Ecol for the rest of the flight may be difficult to carry out. Two variants of the invention, described hereinbelow (hereinafter called variant 2 and variant 3), allow facilitating this trade-off by allowing selecting a level P.sub.M1Ecol above the minimum power required for the continuation of the flight P.sub.Rmin_Ecol.

[0089] The regulation system 5 maintains the engine 10 at the constant power PM1Ecol and continuously adapts the power of the engine 20 to the required level to maintain the rotational speed of the rotor at the desired speed.

[0090] P.sub.M2lim_Ecol=P.sub.M2max_OEI?P.sub.M1Ecol The regulation system 5 also limits the power of the engine 20 to the level P.sub.M2lim Ecol so that the total power delivered by the two engines does not exceed the maximum power P.sub.M2max_OEI of the engine 20. Hence, the limit P.sub.M2lim_Ecol is calculated as follows:

[00004]$P_{M2{\lim }_{-}\mathrm{Ecol}}=P_{M2\max \_\mathrm{OEI}}-P_{M1\mathrm{Ecol}}$

[0091] P.sub.M2lim_Ecol=P.sub.M2max_OEI?P.sub.M1Ecol Hence, the engine 20 thus operates at an average power level much lower than its maximum power, without this being perceptible by the pilot. This also has the advantage of consuming a much smaller amount of electrical energy, which could be interesting when the electrical energy is supplied by a battery the amount of available energy of which is necessarily limited.

[0092] Throughout the duration of the training, the engine parameters returned by the control system for the pilot display are manipulated so that they are representative of what would be displayed during a real fault situation. Thus, the speed, the torque or the power of the engine 10 is indicated at zero to represent its simulated fault status, while this same engine actually delivers a significant power level. Conversely, the equivalent parameters of the engine 20 are indicated at the levels where they would be if this engine was the only one to supply power to the rotor.

[0093] Still throughout the duration of the training, the regulation system continuously monitors the operation of the two engines. Thus, in the event of an actual fault detected on either engine, the regulation system immediately interrupts the failure training and simulation procedure and instantaneously reactivates the engine that is not affected by a fault, so that it delivers all the power necessary for continuing the flight.

Variant 2: Use of the engine 2 in the brake mode in the rest of the training flight

[0094] In the variant 2, we act on the average power level delivered by the engine 1 in the rest of the training mode.

[0095] As mentioned hereinabove, the trade-off between the power required to maintain a good responsiveness of the engine 1 and the minimum power level necessary for the continuation of the training flight may be very difficult to meet.

[0096] To facilitate this trade-off, a variant of the invention consists in using the engine 2 in a reversible manner so as to be able to increase the power PM1Ecol of the engine 1. This solution can be carried out only if the engine 2 can draw mechanical power on the BTP and that the powertrain of the engine 2 is reversible, either by recharging a storage member (for example, a battery), or by instantaneously dissipating this power (for example through electrical power resistors).

[0097] According to this variant 2, the regulation system controls a power level P.sub.M1Ecol higher than what would be controlled according to the basic invention. The power P.sub.M1Ecol delivered by the engine 10 while a fault thereof is simulated, in this variant 2, higher than the minimum power of the flight P.sub.Rmin Ecol. In order to maintain the rotational speed of the rotor at the desired level when P.sub.Rmin_Ecol(t)<P.sub.M1Ecol, the regulation system controls a negative power on the engine 20. Thus, the sum of the powers of the two engines is maintained at the level of the rotor need.

[0098] In this variant 2, the selection of the constant P.sub.M1Ecol power is always subject to two constraints: [0099] it should always be as high as possible to improve the responsiveness of the engine 10 in the event of a fault affecting the engine 20 during the training;

[0100] P.sub.M1Ecol?P.sub.Rmin_Ecol?P.sub.M2minit cannot exceed the minimum power of the engine 20, P.sub.M2min:

[0101] P.sub.M1Ecol?P.sub.Rmin_Ecol?P.sub.M2min

[0102] P.sub.M1Ecol?P.sub.Rmin_Ecol?P.sub.M2minThis minimum power PM2min is here negative and corresponds (in absolute value) to the maximum power that the engine 20 can draw from the BTP. This minimum power P.sub.M2min is not necessarily equal to the maximum power P.sub.M2max_OEI, and may depend on the capacity of the powertrain of the engine 20 to absorb the power regenerated by this engine. In the case of an electrical powertrain, it may be the maximum recharging power of the battery, or the maximum power dissipated by the braking resistors. In the case where only one battery allows absorbing the power drawn by the engine 20, the minimum power P.sub.M2min may also be constrained by energy considerations. Indeed, at any instant of the training flight, the energy regenerated by the engine 20 should not exceed the maximum capacity of the battery.

Variant 3: Real-time adaptation of P.SUB.M1Ecol .to eliminate the safety/representativeness trade-off for the selection of the constant P.SUB.M1Ecol

[0103] In the variant 3, we act on the average power level delivered by the engine 1 in the rest of the training mode.

[0104] It has been explained hereinabove that, during a training, PM1Ecol should be as high as possible in order to: [0105] guarantee the safety of the flight in the event of an actual fault affecting the engine 20 while bearing in mind that:

[0106] the power loss due to the actual fault affecting the engine 20 will be all the lower as the engine 20 operates at low power (and therefore as the engine 10 operates at high power); the responsiveness of the engine 1 will be as more rapid as the engine 10 operates at high power (case of gas turbines only); [0107] enable the engine 20 to operate at low power and thus preserve its energy source (particularly interesting if it is a battery, to enable sessions of training at coping with a fault affecting the engine 10 repeated without battery recharging).

[0108] On the other hand, it has also been explained that P.sub.M1Ecol should be low enough for the engine 20 to be able to compensate for the power drops required by the rotor while complying with P.sub.M2inst>P.sub.M2min (with P.sub.M2min=0 if the powertrain of the engine 20 is not reversible and P.sub.M2min<0 if the powertrain of the engine 20 is reversible) in order to ensure good representativeness of the dynamic behaviour of the powertrain.

[0109] In practice, the trade-off aiming to define the constant P.sub.M1Ecol hereinabove might be difficult to find (and even impossible).

[0110] In the variant 3, it is suggested to adapt in real-time the regime PM1Ecol over time, in order to make the engine 20 operate around a power that is just necessary (piloting margin) PM2_ref(t), in order to ensure a good representativeness of the dynamic behaviour of the powertrain. An example of real-time adaptation of P.sub.M1Ecol is given in the following figures

[0111] (FIGS. 7 and 8).

[0112] In FIG. 7, a particular embodiment of the variant 3 has been detailed in the regulation system.

[0113] The slow nature of the real-time adaptation of the power PM1Ecol enables the engine 20 (faster) to perfectly compensate for the additional power supplied to the rotor and thus to make the variations in the power of the engine 10 transparent to the pilot.

[0114] It goes without saying that in the case of the variant 3, it is also necessary to adapt in real-time the power limit of the engine P.sub.M2lim_Ecol so that the total power supplied by the two engines never exceeds the maximum power of the engine 20:

[00005]$P_{M2{\lim }_{-}\mathrm{Ecol}}\left(t\right)=P_{M2\max \_\mathrm{OEI}}-P_{M1\mathrm{Ecol}}\left(t\right)$

[0115] In order to adapt the regime PM1Ecol of the gas turbine (turboshaft engine) in real-time, this adaptation may be carried out based, for example, on one or more of the elements listed hereinbelow: [0116] control of the collective pitch of the aircraft; [0117] the power anticipation information originating from the aircraft; [0118] the power delivered by the engine 20, averaged over a given time period; [0119] any other information allowing estimating the average level of power requirement of the rotor.

[0120] For example, it is possible to have:

[00006]$P_{M1\mathrm{Ecol}}\left(t\right)=\mathrm{Low}-\mathrm{Pass}\mathrm{Filter}\left(P_{\mathrm{Helico}}\left(t\right)-\mathrm{PM}2\mathrm{ref}\right)$

where PM2ref is a constant value and where the real-time power of the helicopter (P.sub.Helico(t)) is, for example, equal to: [0121] P.sub.M1inst(t)+P.sub.M2inst(t), where P.sub.Minst(t) is the power delivered in real-time by the first powertrain and P.sub.M2inst(t) is the power delivered in real-time by the second powertrain; or [0122] an estimated power of collective pitch type; or [0123] a power estimated by the avionics; etc.

[0124] Hence, we will have on average:

[00007]$P_{M1\mathrm{Ecol}}\left(t\right)=P_{M1\mathrm{inst}}\left(t\right)$ [0125] P.sub.M2inst(t)=P.sub.M2ref, P.sub.M2ref being the desired piloting margin.

[0126] It is specified that the dynamics of the low-pass filter should be slower than the possible dynamics of the first powertrain.

[0127] The above-described illustrative examples are based on a two-engine hybrid propulsion system. Nonetheless, the invention may cover any multi-engine application where training of the pilot consists in simulating the fault affecting an engine amongst several ones. Mention may be made, for example, of an architecture with three engines in parallel, one or two of these three engines being electric.

[0128] In addition, the above-described illustrative examples depict a situation of training a pilot to cope with a fault affecting a motor that no longer delivers power (total fault affecting an engine), since it is in general the most demanding situation in terms of piloting and the most restrictive in terms of simulation. Nonetheless, the invention may cover all situations of partial fault affecting an engine where the latter continues to deliver a given power level with more or less degraded performances. For example, mention may be made of the case of a total fault affecting the power regulation, so-called freeze, where the simulated faulty engine is frozen at a constant power level on a point of the flight domain.

CITED REFERENCES

[0129] [1] U.S. Pat. No. 6,917,908 B2 [0130] [2] U.S. Pat. No. 8,025,503 B2 [0131] [3] EP 2 724 939 B1 [0132] [4] EP 2 886 456 A1