METHOD AND DEVICE FOR PREDICTING THE LIFESPAN OF AN AIRCRAFT HEAT WELL

Abstract

The present invention relates to a method and device for predicting the heat sink lifespan potential of an aircraft, in which: - characteristics of each airport are obtained, - environmental data for each airport are obtained, - a descriptive set (E500) is formed for each airport from the airport characteristics in the environmental data, - a list of airports visited by the aircraft since the last heat sink installation is determined (E510) for at least one known aircraft, - a training set (E520, E530, E540) is formed, - a model is trained and evaluated from the training set, - the aircraft's heat sink lifespan potential is predicted (E550) from the descriptive set and a list of airports visited by the aircraft. Fig. 5

Description

Title of the invention: METHOD and DEVICE FOR PREDICTING THE LIFESPAN OF AN AIRCRAFT HEAT WELL technical field

[0001] The present invention relates to a method and device for predicting the life potential of heat sinks of an aircraft. STATE OF PRIOR ART

[0002] In the aeronautical field, predicting equipment lifespan is a major challenge, and heat sinks are among the equipment for which this is particularly true. Indeed, there is a significant disparity in the lifespan of heat sinks from one aircraft to another. A heat sink is the stack of discs at the landing gear wheels which, when brought into contact with each other, allow the aircraft to brake. Predicting the potential lifespan of a heat sink is a complex task, and there is currently no solution capable of accurately making such predictions.

[0003] Indeed, although the use of data measured and / or stored by equipment included in an aircraft makes it possible in many cases to calculate a remaining life, the accessibility of this type of information is not guaranteed for all aircraft.

[0004] The absence of a suitable sensor or the difficulty of accessing data are indeed major obstacles.

[0005] Today, the majority of aircraft are not equipped with sensors measuring the wear and tear of a heat sink. Even if such sensors existed, accessing this data for predictive purposes for other aircraft is nevertheless difficult due to the amount of data to be processed and the requirement for authorization from the aircraft owner to process this data.

[0006] The present invention makes it possible to overcome the problem of accessing data, to make a personalized prediction of the number of cycles that a heat sink can perform and to provide recommendations to airlines that will allow them to optimize the lifespan of their heat sinks and better manage their fleet. Description of the invention

[0007] To this end, according to a first aspect, an embodiment proposes a method for predicting the heat sink lifespan potential of an aircraft, characterized in that the method comprises the steps of:

[0008] - obtaining characteristics of each airport, the characteristics of an airport being at least representative of the airport runways and their characteristics,

[0009] - obtaining data for different time periods within a year environmental conditions of each airport,

[0010] - formation, for each airport, of a descriptive set of the airport from airport characteristics and environmental data for different time periods,

[0011] - determination, for at least one known aircraft, of a list of airports visited by the aircraft since the last heat sink installation on the aircraft, the determination of airports being carried out from flight data of the aircraft since the last heat sink installation and the list of airports visited being obtained from flight tracking software,

[0012] - formation of a training game from the descriptive set and the list of airports visited by the aircraft,

[0013] - training and evaluation of a model from the training set, the model having a property of invariance to the order of airports visited,

[0014] - prediction of the potential lifespan of an aircraft heat sink from the descriptive information and a list of airports visited by the aircraft.

[0015] One embodiment also relates to a device for predicting the heat sink lifespan potential of an aircraft, characterized in that the device comprises:

[0016] - means of obtaining characteristics of each airport, the characteristics of an airport being at least representative of the airport's runways and their characteristics,

[0017] - means of obtaining, for different periods of time within a year, environmental data for each airport,

[0018] - means of training, for each airport, a descriptive set of the airport based on the airport's characteristics and environmental data for different time periods,

[0019] - means of determining, for at least one known aircraft, a list of airports visited by the aircraft since the last heat sink installation on the aircraft, the determination of the airports being carried out from flight data of the aircraft since the last heat sink installation and the list of airports visited being obtained from flight tracking software,

[0020] - training means, for each airport of a training game from the descriptive information and list of airports visited by the aircraft,

[0021] - means for training and evaluating a model from the game training model, having an invariance property with respect to the order of airports visited,

[0022] - means for predicting the potential lifespan of an aircraft heat sink starting from the descriptive set and a list of airports visited by the aircraft.

[0023] Thus, the present invention makes it possible to predict the life potential of heat sinks without requiring the use of sensors.

[0024] The present invention makes it possible to overcome the problem of accessing the intrinsic data of an aircraft.

[0025] The present invention also makes it possible to make a precise and personalized prediction of the number of cycles that a heat well will be able to perform depending on the airports on which it will operate.

[0026] The present invention also makes it possible to take into account the different factors influencing the wear of heat sinks and to provide recommendations to airlines to optimize the lifespan of their heat sinks and to better manage their fleet.

[0027] The invariance property of the model makes it possible to evaluate a heat sink life potential and to train the model only once on the training set in order to then predict the number of cycles that a heat sink will be able to perform.

[0028] According to a particular mode, the method further comprises a step of determining a severity score for each airport from the trained model, the descriptive set of the airport and a list containing only the airport.

[0029] The invariance property of the model allows extrapolation of a severity score for individual airports and training of the model only once on the training set.

[0030] Thus, the model has two capabilities, that of predicting the number of cycles that a heat well can achieve given its frequencies of visits to the different airports.

[0031] Thus the present invention makes it possible to identify the airports which are the source of significant wear and tear on heat sinks.

[0032] According to a particular mode, the characteristics of an airport are the number of runways of each airport, the number of runways having a surface, an indicator of the condition of the surface of each runway, the length of the runways of each airport, the mechanical characteristics of the runway surface, the distances of the airport taxiways laid out for the movement of aircraft between parking points and runways, the cleanliness of the runways, the altitude of the airport.

[0033] Thus the present invention makes it possible to identify the airports which are the source of significant wear and tear on heat sinks.

[0034] According to a particular mode, the environmental data include temperature, humidity, wind, visibility, pollution during the time period of the year.

[0035] Thus the present invention makes it possible to make a precise and personalized prediction of the number of cycles that a heat well will be able to perform depending on the airports on which it will operate and depending on the seasons.

[0036] According to a particular mode, the method further comprises a step of determining the frequency of visits to each airport.

[0037] The model thus takes as input a list of airports and an associated visit frequency list, and allows, by training the model only once, to obtain both a prediction of the number of cycles and a prediction of the severity of the airport.

[0038] According to a particular method, the determination of a severity score and the prediction are carried out using machine learning software.

[0039] According to a particular mode, the machine learning software determines the severity score and the prediction from two levels of dense neural networks and between the two levels of dense neural networks the process includes a weighting step by frequency of visits to each airport.

[0040] A particular mode also relates to a computer program product. It includes instructions for implementing, by equipment, the process according to one of the preceding embodiments, when said program is executed by a processor of the device.

[0041] A particular mode also relates to a storage medium. It stores a computer program comprising instructions to implement, by a node device, the process according to one of the preceding embodiments, when said program is executed by a processor of the device. Brief description of the drawings

[0042] The features of the invention mentioned above, as well as others, will become clearer upon reading the following description of an exemplary embodiment, said description being made in relation to the accompanying drawings, among which:

[0043] [Fig.1] schematically illustrates an example of an algorithm for determining descriptive data of an airport for different periods of a year;

[0044] [Fig.2] schematically illustrates an example of an algorithm for determining, for each aircraft, the airports visited by the aircraft since the last heat sink installation on the aircraft;

[0045] [Fig.3] schematically illustrates an example of an algorithm for training and evaluating a model;

[0046] [Fig.4] schematically illustrates an example of an algorithm for predicting the life potential of a heat sink;

[0047] [Fig.5] schematically illustrates an example of an algorithm for predicting the severity score of an airport;

[0048] [Fig.6] schematically illustrates an example of the architecture of a device for predicting the life potential of heat sinks of an aircraft.

[0049] DETAILED DESCRIPTION OF IMPROVEMENTS

[0050] Fig. 1 schematically illustrates an example of an algorithm for determining descriptive data of an airport for different periods of a year.

[0051] The device 60 for predicting the life potential of an aircraft heat sink obtains the characteristics of each airport at step E100.

[0052] The characteristics of an airport are, for example, the number of runways at each airport, the number of runways with a surface, an indicator of the condition of the surface of each runway, the length of the runways at each airport, the mechanical characteristics of the runway surface, the distances of the airport taxiways arranged for the movement of aircraft between parking points and runways, the cleanliness of the runways and the altitude of the airport.

[0053] By design, any characteristic that can be measured across all airports worldwide can be integrated. For example, satellite data can be used to associate each runway with the length of its taxiways or its proximity to residential areas. Indeed, the length of a runway affects whether or not an aircraft pilot needs to reverse thrust during a short landing. Similarly, the proximity of residential areas to one of the airport's runways can influence its use or help limit noise impact.

[0054] The device 60 for predicting the life potential of an aircraft heat sink obtains environmental data from each airport at step El 10.

[0055] Environmental data include, for example, temperature, humidity, wind, visibility, pollution, etc...

[0056] In step E120, the device 60 for predicting the heat sink life potential of an aircraft forms predetermined airport / time period pairs.

[0057] Since environmental conditions can vary significantly throughout the year, environmental data are grouped into several predetermined periods of the year, such as the seasons. Representing data by season provides a more accurate picture of airport conditions at each time of year.

[0058] In step E130, the device 60 for predicting the heat sink lifespan potential of an aircraft aggregates all the information obtained in step E120 in order to provide a statistical representation of the airport's environmental characteristics. Data from the same airport are aggregated using the following statistical tools: median, quantiles (Q10 and Q90), and mean deviation from the median.

[0059] The value below which 10% of the lowest values ​​are found is called the 10th quantile (Q10), and the value above which 10% of the highest values ​​are called the 90th quantile (Q90).

[0060] At step E140, the device 60 for predicting the life potential of an aircraft heat sink performs a preprocessing of the data obtained at steps El00 and El30.

[0061] Each pair (season, airport) is thus represented by a vector whose values ​​represent the airport's characteristics. To make these vectors usable by a neural network, we perform a 1-of-n encoding, which consists of encoding an n-state variable using n bits for categorical variables and normalizing the data for numerical variables. Outliers are identified and removed. Missing values ​​are replaced by the mean in the numerical case or by the most frequent value in the categorical case.

[0062] The pre-processed data are thus available at step E150 and form a descriptive set of airport season pairs.

[0063] Figure 2 schematically illustrates an example of an algorithm for determining, for each aircraft, the airports visited by the aircraft since the last heat sink installation on the aircraft.

[0064] At step E200, the device 60 for predicting the life potential of an aircraft heat sink obtains an identifier (tail number) for each aircraft.

[0065] At step E210, the device 60 for predicting the life potential of an aircraft heat sink obtains, for each identified aircraft, the last date of shipment and receipt of the heat sink.

[0066] At step E220, the device 60 for predicting the life potential of an aircraft heat sink obtains, for each identified aircraft, the list of all airports visited by the aircraft since it has been equipped with the received heat sinks.

[0067] According to the invention, the list of airports visited is obtained from flight tracking software such as "FlightRadar24" and "FlightAware," available on the internet. This software indicates the position of aircraft in flight. From the positions of each identified aircraft, it is possible to deduce the list of airports visited by each identified aircraft, starting from the date it is equipped with the heat sinks. This solution avoids having to question the operator of the identified aircraft.

[0068] In step E230, the device 60 for predicting the heat sink life potential of an aircraft determines for each identified aircraft the frequency of visits, per season, of the identified aircraft to each airport.

[0069] At step E240, the device 60 for predicting the life potential of an aircraft heat sink determines a training dataset at least from the frequency of visits and the descriptive set of airport season pairs.

[0070] The data obtained at steps E230 and E240 form a training dataset in the form of vectors of the representations of the airports visited and the frequencies of visits at step E250.

[0071] Fig. 3 schematically illustrates an example of a model training and evaluation algorithm.

[0072] According to the invention, a learning model as described in the publication entitled “Deep Sets” whose authors are Manzil Zaheer, Satwik Kottur, and Siamak Ravanbhakhsh is used in step E320.

[0073] The model complies with the conditions specified in the DeepSet paper in order to ensure an invariance property to the order of airports visited.

[0074] The model takes as input E300 the vectors of the representations of the visited airports and the visit frequencies obtained in step E250. A first succession of dense neural network layers creates a representation of these vectors in a latent space to perform a prediction E330 of the number of cycles completed and subsequently performs an error penalty E340. Training is performed in a supervised manner using gradient descent E350 and weight adjustment E360.

[0075] Several combinations of E370 hyperparameters are tested using a technique such as GridSearch to parameterize the model in the best possible way (number of neurons, layers, etc.) in order to obtain a trained E310 model which after several iterations is evaluated as E380.

[0076] Fig. 4 schematically illustrates an example of an algorithm for predicting the severity score of an airport.

[0077] This algorithm is performed for each airport.

[0078] To generate the severity score of an airport, simply input E410 The trained model is given a list containing, for each season, only the airport whose severity is to be predicted and associated with a visit frequency of 1. The prediction thus obtained corresponds to the total number of cycles theoretically expected if only this airport were visited. This value can therefore be interpreted as the airport's severity for the model.

[0079] Since the model is used to predict the severity of each airport visited individually, it is preferable for the model to be independent of the order in which the airports are visited. Indeed, the present invention aims to obtain absolute severity scores, independent of the conditions encountered at other airports.

[0080] In the context of the present invention, the order of airport visits has no impact on wear. This property is expressed in the following form: . ,¾}) = / ({^jr(l); ■ • •

[0081] With TT an arbitrary permutation, x^y the representation of the airport ir(i), and f our model. Such a function is said to be invariant. According to the theorem stated in the Deepset paper, it can also be decomposed as follows: P

[0082] In our case q> is the function which associates to an airport its representation by the model, weighted by its frequency of visit, and p is the function which from its weighted representations predicts the wear.

[0083] The prediction is made from the descriptive vectors of the airport season pairs E400 obtained in step E150 and the data from step E410.

[0084] A first succession of dense neuronal layers E420 creates a representation of these vectors in a latent space to obtain a representation model of the airport E430. A weighted sum of these representations is performed using the airport visit frequencies E410 as weighting. The result of this weighted sum is given as input to a second succession of neuronal layers E440, which predicts at output E450 the total number of cycles that the heat sink can perform if only the airport is visited, corresponding to a severity score for the visited airport.

[0085] Figure 5 schematically illustrates an example of a prediction algorithm for the life potential of a heat well.

[0086] The user provides a list of visited airports and their frequency. Our solution retrieves the corresponding vector representations for each of these airports and provides them as input to the model along with the list of associated visit frequencies. The model then predicts the total number of cycles expected for the heat sink.

[0087] The prediction of the life potential of a heat well is carried out from the descriptive vectors of the airport season pairs E500 obtained in step El50 and the anticipated visit frequencies for each airport / season pair.

[0088] A first succession of dense neuronal layers E520 creates a representation of these vectors in a latent space to obtain a model representing the airport models E530. A weighted sum of these representations is performed using the frequencies E510 of anticipated visits for each airport / season pair as weighting. The result of this weighted sum is given as input to a second succession of neuronal layers E540, which predicts as output the total number of cycles that the heat sink E550 can perform.

[0089] Figure 6 schematically illustrates an example of the architecture of a device prediction of the heat sink lifespan potential of an aircraft.

[0090] According to the hardware architecture example shown in [Fig.6], the device 60 for predicting the life potential of an aircraft heat sink comprises, connected by a communication bus 600: a processor or CPU (Central Processing Unit) 601; a RAM (Random Access Memory) 602; a ROM (Read Only Memory) 603; a storage unit such as a hard disk drive (or a storage media reader, such as an SD card reader) 604; at least one communication interface 605 allowing the device 60 for predicting the life potential of an aircraft heat sink to access different databases or different flight tracking software.

[0091] The processor 601 is capable of executing instructions loaded into RAM 602 from ROM 603, external memory (not shown), a storage medium (such as an SD card), or a communication network. When the Cont controller device is powered on, the processor 601 is capable of reading instructions from RAM 602 and executing them. These instructions form a computer program causing the processor 601 to implement all or part of the process described in relation to Figs. 1 to 5.

[0092] The method described below in relation to Figs. 1 to 5 can be implemented in software form by executing a set of instructions by a programmable machine, for example a DSP (Digital Signal Processor) or a microcontroller, or can be implemented in hardware form by a dedicated machine or component, for example a FPGA (Field-Programmable Gate Array) or an ASIC (Application-Specific Integrated Circuit). In general, the device 60 for predicting the heat sink lifetime potential of an aircraft comprises electronic circuitry configured to implement the methods described in relation to Figs. 1 to 5.

Claims

Demands

1. A method for predicting the heat sink lifespan potential of an aircraft, characterized in that the method comprises the steps of: - obtaining characteristics of each airport, the characteristics of an airport being at least representative of the airport's runways and their characteristics, - obtaining, for different time periods of a year, environmental data for each airport, - forming (E500), for each airport, a descriptive set of the airport from the airport characteristics of the environmental data for the different time periods, - determining (E510), for at least one known aircraft, a list of the airports visited by the aircraft since the last heat sink installation on the aircraft,Airport determination is carried out using flight data from the aircraft since the last heat sink installation, and the list of visited airports is obtained from flight tracking software; - training (E520, E530, E540) of a training set from the descriptive set and the list of airports visited by the aircraft; - training and evaluation of a model from the training set, the model having an invariance property with respect to the order of the airports visited; - prediction (E550) of the aircraft's heat sink lifespan potential from the descriptive set and a list of airports visited by the aircraft.

2. A method according to claim 1, characterized in that the method further comprises a step of determining a severity score for each airport from the trained model, the descriptive set of the airport and a list containing only the airport.

3. A method according to claim 1, characterized in that the characteristics of an airport are the number of runways at each airport, the number of runways with a surface, an indicator of the condition of the surface of each runway, the length of the runways at each airport, the mechanical characteristics of the runway surface, and the distances of the airport taxiways designed for the movement of aircraft between points parking and runways, runway cleanliness, airport altitude.

4. A method according to any one of claims 1 to 3, characterized in that the environmental data comprise one or more quantities among temperature, humidity, wind, visibility, pollution.

5. A method according to any one of claims 1 to 4, characterized in that the method further comprises a step of determining the frequency of visits to each airport.

6. Method according to claim 2, characterized in that the determination of a severity score and the prediction are carried out from a machine learning software.

7. A method according to claim 6, characterized in that the machine learning software determines the severity score and prediction from two levels of dense neural networks and in that, between the two levels of dense neural networks, the method includes a weighting step by frequency of visits to each airport.

8. A device for predicting the lifespan potential of an aircraft heat sink, characterized in that the device comprises: - means for obtaining characteristics of each airport, the characteristics of an airport being at least representative of the airport's runways and their characteristics, - means for obtaining, for different time periods of a year, environmental data for each airport, - means for forming, for each airport, a descriptive set of the airport from the airport characteristics and environmental data for the different time periods, - means for determining, for at least one known aircraft, a list of the airports visited by the aircraft since the last heat sink installation on the aircraft.the determination of airports being carried out from flight data of the aircraft since the last installation of the heat sink and the list of airports visited being obtained from flight tracking software, - means of training a training set from the descriptive set and the list of airports visited by the aircraft, - means of training and evaluating a model from the training set, the model having a property of invariance to the order of airports visited, - means of predicting the potential lifespan of an aircraft heat sink from the descriptive set and a list of airports visited by the aircraft.

Images