System and method for monitoring an aircraft engine
A system for monitoring an aircraft engine, including an acquisition module for acquiring current measurements of physical quantities, referred to as input and output physical sizes, relative to the aircraft engine, a module for simulating the physical behavior of the aircraft engine, in order to simulate output physical quantities as a function of the current measurements of input physical sizes, a processor for calculating physical margins, referred to as actual physical margins, between the simulated values of output physical quantities and the corresponding current measurements of output physical quantities, a learning module for predicting margins, the margins being predicted using current measurements of input physical quantities and wherein the processor is configured to calculate surveillance residuals giving an indication of the state of the aircraft engine.
1 . A system for monitoring an aircraft engine, including:
an acquisition module configured to acquire, during a flight time of the aircraft, current measurements of physical quantities, referred to as physical input quantities and physical output quantities, relating to said aircraft engine and its environment,
a module for simulating the physical behavior of said aircraft engine, configured to simulate values of physical output quantities as a function of said current measurements of physical input quantities observed in real time,
a processor configured to calculate, during a preliminary learning phase in an earliest time period, at each of a plurality of instants of real time, physical margins, referred to as actual margins, between said simulated values of physical output quantities and said corresponding current measurements of physical output quantities observed in real time,
a learning module configured to predict, during an operational phase in remaining time periods following the learning phase for a plurality of the instants of real time, margins, referred to as predicted margins, the predicted margins being expected actual margins between the simulated values and the current measurements of physical input quantities based on the actual margins that were already calculated in past time periods that have elapsed, and
wherein said processor is further configured to calculate monitoring residuals as a difference, at each instant of time, between (i) said actual margins that are calculated by said processor during the operational phase between said simulated values of physical output quantities and said corresponding current measurements of physical output quantities observed in real time and (ii) said predicted margins that are predicted from the learning module, said monitoring residuals giving an indication of the state of the aircraft engine, such that the monitoring residuals are only calculated during the operational phase and the actual margins are calculated during both the learning phase and the operational phase,
wherein the system further comprises a display interface for viewing graphical representations of said monitoring residuals, and the processor is configured to control the display interface to display the graphical representations of said monitoring residuals as a scatterplot graph that includes a plurality of time periods which form (i) the preliminary learning phrase in the earliest time period, where only the actual margins are displayed and (ii) the operational phase in the remaining time periods where the monitoring residuals are displayed in addition to the actual margins.
2 . The system according to claim 1 , wherein said current measurements of physical input quantities and physical output quantities are acquired during stable and transient phases of said flight of the aircraft.
3 . The system according to claim 1 , wherein the learning module is based on a learning model previously constructed by using a reference aircraft engine during a predetermined number of learning flights, the measurements of physical input quantities relating to the reference engine as well as the actual margins generated by the simulation module being injected during each learning flight into the learning module enabling the latter to construct the learning model.
4 . The system according to claim 3 , wherein said number of learning flights is selected to provide compromise between accuracy and stability of the learning model and in that only first elements in the series of flights are taken into account.
5 . The system according to claim 3 , wherein said learning model is constructed according to a statistical technique of linear regression or random forests.
6 . The system according to claim 1 , wherein the physical input quantities include at least one input parameter relating to the aircraft engine and/or to the flight conditions of the aircraft, comprising at least one parameter selected from the speed of rotation of the engine, external temperature, external pressure, fuel flow rate, air flow rate taken from the engine, electrical energy drawn from the engine, position of the vanes, flight altitude, absence or presence of filters, and in that the physical output quantities include at least one output parameter representative of the operating state of the aircraft engine, comprising at least one parameter selected from the internal temperature of the engine and the torque of a shaft of the engine.
7 . The system according to claim 1 , wherein the monitoring residuals are aggregated as averages or modes for synthetic representation.
8 . The system according to claim 1 , wherein the aircraft engine is a helicopter turbine engine.
9 . A method, implemented by a system, for monitoring an aircraft engine, including the following steps of:
acquiring, during a flight time of the aircraft, current measurements of physical quantities, referred to as physical input quantities and physical output quantities, relating to said aircraft engine and its environment,
simulating values of physical output quantities as a function of said current measurements of physical input quantities observed in real time,
calculating, during a preliminary learning phase in an earliest time period, at each of a plurality of instants of real time, physical margins, referred to as actual margins, between said simulated values of physical output quantities and said corresponding current measurements of physical output quantities observed in real time,
predicting, during an operational phase in remaining time periods following the learning phase for a plurality of the instants of real time, margins, referred to as predicted margins, the predicted margins being expected actual margins between the simulated values and the current measurements of physical input quantities based on the actual margins that were already calculated in past time periods that have elapsed, and
calculating monitoring residuals as a difference, at each instant of time, between (i) said actual margins that are calculated by said processor during the operational phase between said simulated values of physical output quantities and said corresponding current measurements of physical output quantities observed in real time and (ii) said predicted margins that are predicted from the learning module, said monitoring residuals giving an indication of the state of the aircraft engine, such that the monitoring residuals are only calculated during the operational phase and the actual margins are calculated during both the learning phase and the operational phase,
wherein the system further comprises a display interface for viewing graphical representations of said monitoring residuals, and the method further includes controlling the display interface to display the graphical representations of said monitoring residuals as a scatterplot graph that includes a plurality of time periods which form (i) the preliminary learning phrase in the earliest time period, where only the actual margins are displayed and (ii) the operational phase in the remaining time periods where the monitoring residuals are displayed in addition to the actual margins.