An aircraft tire pressure monitoring data analysis system based on machine learning

Abstract

The present invention discloses an aircraft tire pressure monitoring data analysis system based on machine learning, and relates to the field of aircraft tire monitoring technology. The system includes an aircraft tire pressure sensor, a plurality of non-tire pressure parameter collectors, a processor and a display, wherein the processor is used to generate historical time series data of tire pressure parameters according to the collection results from the aircraft tire pressure sensor, and for each collector, generate corresponding historical time series data of non-tire pressure parameters according to the collection results from the corresponding collector, then calculate the correlation coefficients of each non-tire pressure parameter and tire pressure parameter based on these historical time series data, and calculate the final estimated value of tire pressure parameter in the future period in combination with machine learning algorithm, and finally transmit the estimated value to the display for output display, so that not only can the ability of early tire pressure prediction and early warning be provided, but also higher prediction accuracy and early warning timeliness are provided, which is conducive to improving flight safety.

Description

Technical Field

[0001] The invention belongs to the technical field of aircraft tire monitoring, and in particular relates to an aircraft tire pressure monitoring data analysis system based on machine learning. Background Art

[0002] Aircraft tire monitoring technology includes a technical solution for real-time monitoring of the internal air pressure of aircraft tires. That is, during the flight of the aircraft, the internal air pressure of the tires needs to be automatically and real-time monitored in order to alarm for abnormal conditions such as tire blowout, air leakage or low pressure to ensure aircraft safety.

[0003] At present, the existing aircraft tire pressure monitoring system generally measures the aircraft tire pressure value in real time through the air pressure sensor arranged inside the aircraft tire, and determines whether to trigger the alarm based on the comparison result between the pressure value and the preset threshold. However, due to the general lack of the ability to predict and warn tire pressure in advance, the alarm timing will be delayed, thereby affecting flight safety. Therefore, how to analyze aircraft tire pressure monitoring data based on tire pressure parameter historical data and non-tire pressure parameter historical data, combined with a machine learning algorithm with predictive capabilities, to achieve the purpose of early tire pressure prediction and warning, is a topic that technicians in this field urgently need to study. Summary of the invention

[0004] The purpose of the present invention is to provide an aircraft tire pressure monitoring data analysis system based on machine learning, so as to solve the problem that the existing aircraft tire pressure monitoring schemes generally lack the ability to predict and warn tire pressure in advance, which may cause the alarm timing to be delayed and affect flight safety.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] In a first aspect, a machine learning-based aircraft tire pressure monitoring data analysis system is provided, comprising an aircraft tire pressure sensor, a plurality of non-tire pressure parameter collectors, a processor, and a display, wherein the processor is communicatively connected to the aircraft tire pressure sensor, the plurality of non-tire pressure parameter collectors, and the display, respectively;

[0007] The aircraft tire pressure sensor is used to collect the air pressure value inside the aircraft tire in real time and transmit the collected result to the processor in real time;

[0008] The non-tire pressure parameter collector is used to collect the non-tire pressure parameter values ​​of the aircraft to which the aircraft tire belongs in real time, and transmit the collected results to the processor in real time;

[0009] The processor is used to generate historical time series data of tire pressure parameters according to the collection results from the aircraft tire pressure sensor, and for each collector among the multiple non-tire pressure parameter collectors, generate corresponding historical time series data of non-tire pressure parameters according to the collection results from the corresponding collector; then, for each non-tire pressure parameter corresponding to each collector, calculate the correlation coefficient between the corresponding parameter and the tire pressure parameter according to the corresponding historical time series data and the historical time series data of the tire pressure parameter; and then, based on the historical time series data of the tire pressure parameter, use a machine learning algorithm to predict in real time the tire pressure parameter in the future The first estimated value of each non-tire pressure parameter is predicted in real time by using the machine learning algorithm according to the corresponding historical time series data and the historical time series data of the tire pressure parameter, and the corresponding tire pressure parameter is predicted in the future. A second estimated value at a unit time; and then according to the tire pressure parameter in the future The first estimated value and all the second estimated values ​​at a unit time are combined with the correlation coefficients of the non-tire pressure parameters and the tire pressure parameter to obtain the tire pressure parameter in the future in real time. The final estimated value of the unit time; finally, the tire pressure parameter is The final estimated value of each unit moment is transmitted to the display in real time, wherein the historical time series data contains the time series data. The unit time and the corresponding parameters are described in The historical parameter values ​​of each unit moment in the unit moments, represents a positive integer greater than or equal to 3, Indicates less than A positive integer of ;

[0010] The display is used to output and display the data from the processor in real time.

[0011] Based on the above invention content, a new solution for predicting aircraft tire pressure based on tire pressure parameter historical data and non-tire pressure parameter historical data is provided, that is, it includes an aircraft tire pressure sensor, multiple non-tire pressure parameter collectors, a processor and a display, wherein the processor is used to generate historical time series data of tire pressure parameters according to the collection results from the aircraft tire pressure sensor, and for each collector, generate corresponding historical time series data of non-tire pressure parameters according to the collection results from the corresponding collector, then based on these historical time series data, calculate the correlation coefficient between each non-tire pressure parameter and tire pressure parameter, and calculate the final estimated value of tire pressure parameter in the future period in combination with machine learning algorithm, and finally transmit the estimated value to the display for output display, so that not only can the ability of early tire pressure prediction and warning be provided, but also because the prediction is made based on the comprehensive tire pressure parameter historical data and non-tire pressure parameter historical data, compared with the conventional prediction solution based on single-dimensional tire pressure parameter historical data, it has higher prediction accuracy and warning timeliness, which is beneficial to improve flight safety and facilitates practical application and promotion.

[0012] In one possible design, the multiple non-tire pressure parameter collectors include a wheel temperature collector, a wheel speed collector, an aircraft speed collector and / or a flight altitude collector.

[0013] In one possible design, after generating the historical time series data of the tire pressure parameters and the non-tire pressure parameters, the processor is further configured to:

[0014] The unit moments in the historical time series data of the non-tire pressure parameters are shifted as a whole so that each shifted unit moment corresponds to each unit moment in the historical time series data of the tire pressure parameters, and for each shifted unit moment, the corresponding and updated historical time series data is calculated according to the following formula :

[0015]

[0016] In the formula, represents the total number of moments in the historical time series data of the tire pressure parameter, Indicates less than or equal to A positive integer, Indicates The non-tire pressure parameter value corresponding to the unit time after translation, Indicates the The time difference between the first moment in the historical time series data of the tire pressure parameter after the translation of the unit moment, , , and Respectively represent the historical time series data of the non-tire pressure parameters and One-to-one correspondence between adjacent units The coefficients of the local cubic polynomial for implementing cubic spline interpolation are obtained by using the least squares method. represents a positive integer greater than or equal to 4, The adjacent recording time refers to the distance from the The nearest previous unit moment after translation Unit moment.

[0017] In a possible design, for each non-tire pressure parameter corresponding to each collector one by one, a correlation coefficient between the corresponding parameter and the tire pressure parameter is calculated based on the corresponding historical time series data and the historical time series data of the tire pressure parameter, including:

[0018] For a non-tire pressure parameter, randomly extract a segment of time series data from the corresponding historical time series data to form the first sample , and also extracting the first sample from the historical time series data of the tire pressure parameter Another time series data of the same period constitutes the second sample ;

[0019] For the first sample Perform a normal distribution KS test to calculate the first test statistic value , and also for the second sample Perform the normal distribution KS check to calculate the second check statistic value ;

[0020] If the first check statistic value and the second check statistic value are greater than the preset threshold, the first sample is calculated according to the following formula With the sample The correlation coefficient :

[0021]

[0022] In the formula, represents a positive integer, In the first sample The Sample values, In the second sample The Sample values, Indicates the first sample The sample mean of The second sample The sample mean of ;

[0023] The correlation coefficient As the correlation coefficient between the certain non-tire pressure parameter and the tire pressure parameter.

[0024] In one possible design, based on the historical time series data of the tire pressure parameter, a machine learning algorithm is used to predict the tire pressure parameter in the future in real time. The first estimate of the unit time includes:

[0025] The first sample data is extracted from the historical time series data of the tire pressure parameter in the following manner: extracting the sequentially continuous sample data in time series from the historical time series data of the tire pressure parameter. The tire pressure historical parameter value of each unit moment is used as the first model input item, and the tire pressure parameter located at the unit moment is extracted from the historical time series data of the tire pressure parameter. After unit periods and sequentially in time The tire pressure historical parameter values ​​of the unit time period are used as the first model output item, and then the first model input item and the first model output item are used as a first sample data, wherein, Indicates greater than or equal to 2 and less than A positive integer, Indicates less than or equal to A positive integer of ;

[0026] Applying all of the first sample data to perform model training on an artificial intelligence model based on a machine learning algorithm to obtain a first tire pressure value prediction model;

[0027] Extract the most recent values ​​that are sequentially continuous in time series from the historical time series data of the tire pressure parameters. The tire pressure historical parameter values ​​of each unit moment are imported into the first tire pressure value prediction model, and the tire pressure parameter is output in the future. The first estimate of the unit time.

[0028] In one possible design, for each non-tire pressure parameter, the corresponding historical time series data and the historical time series data of the tire pressure parameter are predicted in real time using the machine learning algorithm and the tire pressure parameter is predicted in the future. The second estimated value of the unit time includes:

[0029] For each non-tire pressure parameter, the corresponding second sample data is extracted from the corresponding historical time series data in the following manner: extract the consecutive second sample data in time series from the corresponding historical time series data. The non-tire pressure historical parameter value at each unit time is used as the second model input item, and the non-tire pressure historical parameter value at each unit time is extracted from the historical time series data of the tire pressure parameter. After unit periods and sequentially in time The tire pressure historical parameter value of each unit period is used as the second model output item, and then the second model input item and the second model output item are used as a second sample data, wherein, Indicates greater than or equal to 2 and less than A positive integer, Indicates less than or equal to A positive integer of ;

[0030] For each of the non-tire pressure parameters, using all of the corresponding second sample data to perform model training on an artificial intelligence model based on the machine learning algorithm to obtain a second tire pressure value prediction model;

[0031] For each non-tire pressure parameter, extract the most recent parameters that are consecutive in time sequence from the corresponding historical time series data. The non-tire pressure historical parameter value of each unit moment is imported into the second tire pressure value prediction model, and the corresponding tire pressure parameter is output in the future. The second estimated value of the unit time.

[0032] In one possible design, the artificial intelligence model based on the machine learning algorithm adopts an LSTM model, a Bi-LSTM model, or an Attention-LSTM model.

[0033] In one possible design, the tire pressure parameter is used in the future The first estimated value and all the second estimated values ​​at a unit time are combined with the correlation coefficients of the non-tire pressure parameters and the tire pressure parameter to obtain the tire pressure parameter in the future in real time. The final estimate of the unit moment includes:

[0034] In the future At each unit moment in the unit moments, the final estimated value of the tire pressure parameter at the corresponding moment is calculated according to the following formula:

[0035]

[0036] In the formula, Indicates less than or equal to A positive integer, Indicates that the tire pressure parameter in the future Unit time The final estimate at the moment, represents the total number of parameters of the non-tire pressure parameters, Indicates less than or equal to A positive integer, Indicates The influence weight coefficient corresponding to the non-tire pressure parameters is: Indicates that The correlation coefficients corresponding to non-tire pressure parameters are: Indicates less than or equal to A positive integer, Indicates The correlation coefficients corresponding to non-tire pressure parameters are: Indicates that the tire pressure parameter is in the The first estimated value at the moment, Indicates that The tire pressure parameter corresponds to the tire pressure parameter in the first The second estimated value at the moment.

[0037] In a possible design, after calculating the correlation coefficients between each non-tire pressure parameter and the tire pressure parameter, the processor is further configured to:

[0038] At the end of each flight mission of the aircraft to which the aircraft tire belongs, for each non-tire pressure parameter corresponding to each collector, the correlation coefficient between the corresponding parameter and the tire pressure parameter is updated and calculated based on the corresponding historical time series data generated during the execution of the most recent flight mission and the tire pressure parameter and the historical time series data generated during the execution of the most recent flight mission.

[0039] In a possible design, the tire pressure parameter is calculated in real time in the future After obtaining the final estimated value of the unit time, the processor is further configured to:

[0040] According to the tire pressure parameter in the future The final estimated value of each unit moment is used to obtain the maximum tire pressure estimate or the average tire pressure estimate;

[0041] According to a plurality of tire pressure value ranges that are preset and correspond one to one with a plurality of tire pressure warning levels, determining a certain tire pressure value range that includes the maximum tire pressure estimate value or the average tire pressure estimate value from the plurality of tire pressure value ranges;

[0042] The preset alarm action of a certain tire pressure warning level corresponding to the certain tire pressure value range is triggered in real time.

[0043] Beneficial effects of the above scheme:

[0044] The present invention creatively provides a new solution for predicting aircraft tire pressure based on tire pressure parameter historical data and non-tire pressure parameter historical data, that is, it includes an aircraft tire pressure sensor, multiple non-tire pressure parameter collectors, a processor and a display, wherein the processor is used to generate historical time series data of tire pressure parameters according to the collection results from the aircraft tire pressure sensor, and for each collector, generate corresponding historical time series data of non-tire pressure parameters according to the collection results from the corresponding collector, then based on these historical time series data, calculate the correlation coefficient between each non-tire pressure parameter and the tire pressure parameter, and calculate the final estimated value of the tire pressure parameter in the future period in combination with the machine learning algorithm, and finally transmit the estimated value to the display for output display, so that not only can the ability of early tire pressure prediction and warning be provided, but also because the prediction is made based on the comprehensive tire pressure parameter historical data and the non-tire pressure parameter historical data, compared with the conventional prediction solution based on single-dimensional tire pressure parameter historical data, it has higher prediction accuracy and warning timeliness, which is beneficial to improving flight safety and convenient for practical application and promotion. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0046] Figure 1 A schematic diagram of the structure of an aircraft tire pressure monitoring data analysis system based on machine learning provided in an embodiment of the present application. DETAILED DESCRIPTION

[0047] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the present invention will be briefly introduced below in combination with the drawings and the description of the embodiments or the prior art. Obviously, the following description of the structures of the drawings is only some embodiments of the present invention. For ordinary technicians in this field, other embodiments can be obtained based on these embodiments without creative work. It should be noted that the description of these embodiments is used to help understand the present invention, but does not constitute a limitation of the present invention.

[0048] It should be understood that although the terms first and second, etc. may be used herein to describe various objects, these objects should not be limited by these terms. These terms are only used to distinguish one object from another object. For example, a first object can be referred to as a second object, and similarly, a second object can be referred to as a first object without departing from the scope of the exemplary embodiments of the present invention.

[0049] It should be understood that the term "and / or" that may appear in this document is merely a description of the association relationship between associated objects, indicating that there may be three relationships. For example, A and / or B can indicate three situations: A exists alone, B exists alone, or A and B exist at the same time. For another example, A, B and / or C can indicate the existence of any one of A, B and C or any combination of them. The term " / and" that may appear in this document describes another type of association object relationship, indicating that there may be two relationships. For example, A / and B can indicate two situations: A exists alone or A and B exist at the same time. In addition, the character " / " that may appear in this document generally indicates that the previous and next associated objects are in an "or" relationship.

[0050] Example

[0051] like Figure 1 As shown, the aircraft tire pressure monitoring data analysis system provided by this embodiment and based on machine learning includes but is not limited to an aircraft tire pressure sensor, multiple non-tire pressure parameter collectors, a processor and a display, etc., wherein the processor is respectively connected to the aircraft tire pressure sensor, the multiple non-tire pressure parameter collectors and the display in a communication manner. The specific manner in which the processor is connected to other devices in a communication manner can be conventionally implemented using an existing wired communication manner.

[0052] The aircraft tire pressure sensor is used to collect the air pressure value inside the aircraft tire in real time and transmit the collected result to the processor in real time. The aforementioned aircraft tire pressure sensor can be specifically implemented conventionally using an existing air pressure sensor arranged inside the aircraft tire.

[0053] The non-tire pressure parameter collector is used to collect the non-tire pressure parameter values ​​of the aircraft to which the aircraft tire belongs in real time, and transmit the collected results to the processor in real time. The aforementioned non-tire pressure parameter values ​​are specifically but not limited to wheel temperature values, wheel speed values, aircraft speed values ​​or flight altitude values, etc. Therefore, the non-tire pressure parameter collector can be conventionally implemented by using, but not limited to, existing wheel temperature collectors, wheel speed collectors, aircraft speed collectors or flight altitude collectors, that is, the multiple non-tire pressure parameter collectors include but are not limited to wheel temperature collectors, wheel speed collectors, aircraft speed collectors and / or flight altitude collectors, etc.

[0054] The processor, as the core component of the entire aircraft tire pressure monitoring data analysis system, is used to specifically but not limited to execute the following steps S1 to S5.

[0055] S1. Generate historical time series data of tire pressure parameters according to the collection results from the aircraft tire pressure sensor, and generate corresponding historical time series data of non-tire pressure parameters according to the collection results from the corresponding collector for each of the multiple non-tire pressure parameter collectors, wherein the historical time series data includes but is not limited to the time series of The unit time and the corresponding parameters are described in The historical parameter values ​​of each unit moment in each unit moment, etc. Represents a positive integer greater than or equal to 3.

[0056] In step S1, the non-tire pressure parameters are wheel temperature, wheel speed, aircraft speed or flight altitude, etc. Since the aircraft tire pressure sensor and the multiple non-tire pressure parameter collectors collect corresponding parameter values ​​in real time according to a certain period, the collection results can be recorded in sequence according to the collection time sequence, thereby generating the historical time series data of different parameters. The unit time can be, but is not limited to, every 10 seconds, every 30 seconds, or every 1 minute, etc., that is, the historical time series data of the tire pressure parameter includes, for example, but is not limited to, the tire pressure parameter in the most recent The historical parameter values ​​at the minute, The historical parameter values ​​at the last minute, ..., the historical parameter values ​​at the last second minute, and the historical parameter values ​​at the last minute, etc.; the historical time series data of the non-tire pressure parameters include, for example but not limited to, the non-tire pressure parameters (i.e., wheel temperature, wheel speed, aircraft speed or flight altitude, etc.) at the last minute. The historical parameter values ​​at the minute, The historical parameter values ​​at the last minute, ..., the historical parameter values ​​at the last second minute, and the historical parameter values ​​at the last minute, etc.

[0057] In the step S1, it is specifically considered that there is a misalignment in the collection time between the aircraft tire pressure sensor and the non-tire pressure parameter collector. Based on the characteristic that the adjacent collection times of the tire pressure parameter and the non-tire pressure parameter have equal time intervals (because the parameter values ​​are collected periodically, for example, they are collected every 1 minute), in order to better maintain the consistency of the historical time series data of the tire pressure parameter and the non-tire pressure parameter in the collection time, so as to avoid the aforementioned misalignment in the collection time affecting the accuracy of the subsequent data analysis results, preferably, after generating the historical time series data of the tire pressure parameter and the non-tire pressure parameter, the processor is also used to: perform an overall translation of the unit time in the historical time series data of the non-tire pressure parameter, so that each translated unit time corresponds to each unit time in the historical time series data of the tire pressure parameter one by one, and for each translated unit time, calculate the corresponding and updated historical time series data according to the following formula :

[0058]

[0059] In the formula, represents the total number of moments in the historical time series data of the tire pressure parameter, Indicates less than or equal to A positive integer, Indicates The non-tire pressure parameter value corresponding to the unit time after translation, Indicates the The time difference between the first moment in the historical time series data of the tire pressure parameter after the translation of the unit moment, , , and Respectively represent the historical time series data of the non-tire pressure parameters and One-to-one correspondence between adjacent units The coefficients of the local cubic polynomial for implementing cubic spline interpolation are obtained by using the least squares method. represents a positive integer greater than or equal to 4, The adjacent recording time refers to the distance from the The nearest previous unit moment after translation Unit moments. For example, if the unit moments in the historical time series data of the tire pressure parameters are 10:01, 10:02, ..., 10:14, and 10:15, etc., and the unit moments in the historical time series data of the non-tire pressure parameters are 10:00:21, 10:01:21, ..., 10:13:21, and 10:14:21, etc., then after these unit moments are shifted backward by 39 seconds as a whole, the shifted acquisition moments can be 10:01, 10:02, ..., 10:14, and 10:15, etc., and correspond one-to-one and are aligned with the unit moments in the historical time series data of the tire pressure parameters.

[0060] S2. For each non-tire pressure parameter corresponding to each collector, calculate the correlation coefficient between the corresponding parameter and the tire pressure parameter according to the corresponding historical time series data and the historical time series data of the tire pressure parameter.

[0061] In step S2, considering that the influence of different non-tire pressure parameters on tire pressure parameters must be different, it is necessary to calculate the correlation coefficient between each non-tire pressure parameter and the tire pressure parameter, so as to be used for subsequent comprehensive tire pressure prediction and ensure the accuracy of the comprehensive prediction result. Specifically, for each non-tire pressure parameter corresponding to each collector, the correlation coefficient between the corresponding parameter and the tire pressure parameter is calculated based on the corresponding historical time series data and the historical time series data of the tire pressure parameter, including but not limited to the following steps S21 to S24.

[0062] S21. For a non-tire pressure parameter, randomly extract a segment of time series data from the corresponding historical time series data to form the first sample , and also extracting the first sample from the historical time series data of the tire pressure parameter Another time series data of the same period constitutes the second sample .

[0063] In step S21, for example, if the non-tire pressure parameter is the wheel temperature, the time series data from 00:00 on August 1 to 24:00 on August 31 can be extracted from the historical time series data of the wheel temperature to form the first sample , and also extracting the time series data from 00:00 on August 1 to 24:00 on August 31 from the historical time series data of the tire pressure parameter to form the second sample .

[0064] S22. For the first sample Perform a normal distribution KS test to calculate the first test statistic value , and also for the second sample Perform the normal distribution KS check to calculate the second check statistic value .

[0065] S23. If the first check statistic value and the second check statistic value are greater than the preset threshold, the first sample is calculated according to the following formula With the sample The correlation coefficient :

[0066]

[0067] In the formula, represents a positive integer, In the first sample The Sample values, In the second sample The Sample values, Indicates the first sample The sample mean of The second sample The sample mean of .

[0068] In the step S23, the preset threshold may be 0.05, for example.

[0069] S24. The correlation coefficient As the correlation coefficient between the certain non-tire pressure parameter and the tire pressure parameter.

[0070] After step S2, in order to dynamically update the correlation coefficient to ensure the accuracy of the subsequent comprehensive tire pressure prediction results, preferably, after calculating the correlation coefficients between each non-tire pressure parameter and the tire pressure parameter, the processor is also used to: at the end of each flight mission of the aircraft to which the aircraft tire belongs, for each non-tire pressure parameter corresponding to each collector, according to the corresponding historical time series data generated during the execution of the most recent flight mission and the tire pressure parameter and the historical time series data generated during the execution of the most recent flight mission, update the calculation to obtain the correlation coefficient between the corresponding parameter and the tire pressure parameter. The details of the update calculation of the aforementioned correlation coefficient can be found in the aforementioned steps S22 to S24, which will not be repeated here.

[0071] S3. Based on the historical time series data of the tire pressure parameters, a machine learning algorithm is used to predict the tire pressure parameters in the future in real time. The first estimate of unit time, where Indicates less than A positive integer.

[0072] In step S3, the machine learning algorithm is a core artificial intelligence algorithm that specifically studies how computers simulate or implement human learning behavior to acquire new knowledge or skills and reorganize existing knowledge structures to continuously improve their own performance. It is the fundamental way to make computers intelligent. Specifically, based on the historical time series data of the tire pressure parameters, the machine learning algorithm is used to predict the tire pressure parameters in the future in real time. The first estimated value of each unit time includes but is not limited to the following steps S31 to S33.

[0073] S31. Extracting the first sample data from the historical time series data of the tire pressure parameter in the following manner: extracting the sequentially consecutive sample data from the historical time series data of the tire pressure parameter. The tire pressure historical parameter value of each unit moment is used as the first model input item, and the tire pressure parameter located at the unit moment is extracted from the historical time series data of the tire pressure parameter. After unit periods and sequentially in time The tire pressure historical parameter values ​​of the unit time period are used as the first model output item, and then the first model input item and the first model output item are used as a first sample data, wherein, Indicates greater than or equal to 2 and less than A positive integer, Indicates less than or equal to A positive integer.

[0074] In the step S31, for example, when the unit time is every 1 minute, An example is 20. An example is 10. In this way, the historical parameter values ​​of tire pressure from 12:01 to 12:20 (i.e., a total of 20 unit moments) can be extracted from the historical time series data of the tire pressure parameters and used as the first model input item, and the historical parameter values ​​of tire pressure from 10:21 to 10:30 (i.e., a total of 10 unit moments) can be extracted as the first model output item, thereby obtaining a first sample data.

[0075] S32. Apply all of the first sample data to perform model training on an artificial intelligence model based on a machine learning algorithm to obtain a first tire pressure value prediction model.

[0076] In the step S32, the aforementioned artificial intelligence model based on the machine learning algorithm specifically but not limited to adopts LSTM model, Bi-LSTM model or Attention-LSTM model. The aforementioned LSTM (Long Short-Term Memory) model, Bi-LSTM model or Attention-LSTM model are all existing machine learning models, wherein the Attention-LSTM model is an improved model that adds a layer of attention mechanism to the existing LSTM model (which is often used for time series data prediction), and the added attention mechanism is used to allow the LSTM model to dynamically focus on part of the information during training, so as to capture more important information and improve the performance of the model. Therefore, this embodiment preferably adopts the Attention-LSTM model to train the first tire pressure value prediction model. The specific training process of the artificial intelligence model is a prior art means; specifically, the parameters of the artificial intelligence model can be, but not limited to, searched by using a grid search algorithm or a Bayesian search algorithm.

[0077] S33. Extract the most recent consecutive values ​​in time sequence from the historical time series data of the tire pressure parameter. The tire pressure historical parameter values ​​of each unit moment are imported into the first tire pressure value prediction model, and the tire pressure parameter is output in the future. The first estimate of the unit time.

[0078] In the step S33, continuing to refer to the example in the above step S31, if the current time is 14:00, the historical parameter values ​​of the tire pressure from 13:41 to 14:00 (i.e., a total of 20 unit moments) can be extracted from the historical time series data of the tire pressure parameters and imported into the first tire pressure value prediction model to obtain the first estimated value of the tire pressure parameter from 14:01 to 14:10 (i.e., a total of 10 unit moments).

[0079] S4. For each non-tire pressure parameter, according to the corresponding historical time series data and the historical time series data of the tire pressure parameter, the machine learning algorithm is used to predict in real time the corresponding tire pressure parameter in the future The second estimated value of the unit time.

[0080] In step S4, specifically, for each non-tire pressure parameter, according to the corresponding historical time series data and the historical time series data of the tire pressure parameter, the machine learning algorithm is used to predict in real time the corresponding tire pressure parameter in the future. The second estimated value of each unit time includes but is not limited to the following steps S41 to S43.

[0081] S41. For each non-tire pressure parameter, extract the corresponding second sample data from the corresponding historical time series data in the following manner: extract the consecutive second sample data from the corresponding historical time series data. The non-tire pressure historical parameter value at each unit time is used as the second model input item, and the non-tire pressure historical parameter value at each unit time is extracted from the historical time series data of the tire pressure parameter. After unit periods and sequentially in time The tire pressure historical parameter value of each unit period is used as the second model output item, and then the second model input item and the second model output item are used as a second sample data, wherein, Indicates greater than or equal to 2 and less than A positive integer, Indicates less than or equal to A positive integer.

[0082] S42. For each of the non-tire pressure parameters, use all of the corresponding second sample data to perform model training on the artificial intelligence model based on the machine learning algorithm to obtain a second tire pressure value prediction model.

[0083] S43. For each non-tire pressure parameter, extract the most recent parameters that are consecutive in time sequence from the corresponding historical time series data. The non-tire pressure historical parameter value of each unit moment is imported into the second tire pressure value prediction model, and the corresponding tire pressure parameter is output in the future. The second estimated value of the unit time.

[0084] The technical details of the aforementioned steps S41 to S43 can be conventionally derived by referring to the aforementioned steps S31 to S33, and will not be repeated here.

[0085] S5. Based on the tire pressure parameter in the future The first estimated value and all the second estimated values ​​at a unit time are combined with the correlation coefficients of the non-tire pressure parameters and the tire pressure parameter to obtain the tire pressure parameter in the future in real time. The final estimate at unit time.

[0086] In step S5, since the correlation coefficient may be positively correlated with the weight coefficient of the influence of the non-tire pressure parameter on the tire pressure parameter (i.e., the larger the correlation coefficient, the larger the influence weight coefficient), the correlation coefficients of the non-tire pressure parameters and the tire pressure parameter may be combined to achieve the purpose of comprehensive tire pressure prediction. The first estimated value and all the second estimated values ​​at a unit time are combined with the correlation coefficients of the non-tire pressure parameters and the tire pressure parameter to obtain the tire pressure parameter in the future in real time. The final estimated value of the unit moment includes but is not limited to: At each unit moment in the unit moments, the final estimated value of the tire pressure parameter at the corresponding moment is calculated according to the following formula:

[0087]

[0088] In the formula, Indicates less than or equal to A positive integer, Indicates that the tire pressure parameter in the future Unit time The final estimate at the moment, represents the total number of parameters of the non-tire pressure parameters, Indicates less than or equal to A positive integer, Indicates The influence weight coefficient corresponding to the non-tire pressure parameters is: Indicates that The correlation coefficients corresponding to non-tire pressure parameters are: Indicates less than or equal to A positive integer, Indicates The correlation coefficients corresponding to non-tire pressure parameters are: Indicates that the tire pressure parameter is in the The first estimated value at the moment, Indicates that The tire pressure parameter corresponds to the tire pressure parameter in the first The second estimated value at the moment.

[0089] S6. The tire pressure parameter is set in the future The final estimated value at each unit time is transmitted to the display in real time.

[0090] The aforementioned processor may specifically be, but is not limited to, an onboard computer of the aircraft to which the aircraft tire belongs.

[0091] The display is used to output and display the data from the processor in real time. The display may be, but is not limited to, a cabin instrument display of the aircraft to which the aircraft tire belongs. The data displayed may be, but is not limited to, a display of the aircraft in the future according to the tire pressure parameter. The maximum tire pressure estimate or the average tire pressure estimate is obtained by calculating the final estimate value of each unit moment. In addition, in order to achieve the purpose of early tire pressure warning, preferably, the tire pressure parameter is calculated in real time in the future. After the final estimated value of the unit time, the processor is also used for but not limited to performing the following steps: first, according to the tire pressure parameter in the future The final estimated value of each unit moment is used to obtain the maximum tire pressure estimate or the average tire pressure estimate; then, according to a plurality of tire pressure value ranges that are preset and correspond to a plurality of tire pressure warning levels, a certain tire pressure value range that includes the maximum tire pressure estimate or the average tire pressure estimate is determined from the plurality of tire pressure value ranges; finally, the preset alarm action of a certain tire pressure warning level corresponding to the certain tire pressure value range is triggered in real time. It is necessary to ensure that there is no intersection in any of the aforementioned plurality of tire pressure value ranges. For example, there are four plurality of tire pressure warning levels, namely, low pressure warning level, air leakage warning level, no warning required level, and tire blowout warning level (their corresponding tire pressure value ranges are increasing in sequence). The aforementioned preset alarm action may be, but is not limited to, a conventional sound and light alarm action.

[0092] In summary, the aircraft tire pressure monitoring data analysis system provided by this embodiment has the following technical effects:

[0093] The present embodiment provides a new solution for predicting aircraft tire pressure based on tire pressure parameter historical data and non-tire pressure parameter historical data, that is, it includes an aircraft tire pressure sensor, multiple non-tire pressure parameter collectors, a processor and a display, wherein the processor is used to generate historical time series data of tire pressure parameters according to the collection results from the aircraft tire pressure sensor, and for each collector, generate corresponding historical time series data of non-tire pressure parameters according to the collection results from the corresponding collector, then calculate the correlation coefficient between each non-tire pressure parameter and tire pressure parameter based on these historical time series data, and calculate the final estimated value of the tire pressure parameter in the future period in combination with the machine learning algorithm, and finally transmit the estimated value to the display for output display, so that not only can the ability of early tire pressure prediction and warning be provided, but also because the prediction is made based on the comprehensive tire pressure parameter historical data and the non-tire pressure parameter historical data, compared with the conventional prediction solution based on single-dimensional tire pressure parameter historical data, it has higher prediction accuracy and warning timeliness, which is conducive to improving flight safety and is convenient for practical application and promotion.

[0094] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the protection scope of the present invention. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. An aircraft tire pressure monitoring data analysis system based on machine learning, characterized in that: It includes an aircraft tire pressure sensor, a plurality of non-tire pressure parameter collectors, a processor and a display, wherein the processor is respectively communicatively connected to the aircraft tire pressure sensor, the plurality of non-tire pressure parameter collectors and the display; The aircraft tire pressure sensor is used to collect the air pressure value inside the aircraft tire in real time and transmit the collected result to the processor in real time; The non-tire pressure parameter collector is used to collect the non-tire pressure parameter values ​​of the aircraft to which the aircraft tire belongs in real time, and transmit the collected results to the processor in real time; The processor is used to generate historical time series data of tire pressure parameters according to the collection results from the aircraft tire pressure sensor, and for each collector among the multiple non-tire pressure parameter collectors, generate corresponding historical time series data of non-tire pressure parameters according to the collection results from the corresponding collector; then, for each non-tire pressure parameter corresponding to each collector, calculate the correlation coefficient between the corresponding parameter and the tire pressure parameter according to the corresponding historical time series data and the historical time series data of the tire pressure parameter; and then, based on the historical time series data of the tire pressure parameter, use a machine learning algorithm to predict in real time the tire pressure parameter in the future The first estimated value of each non-tire pressure parameter is predicted in real time by using the machine learning algorithm according to the corresponding historical time series data and the historical time series data of the tire pressure parameter, and the corresponding tire pressure parameter is predicted in the future. A second estimated value at a unit time; and then according to the tire pressure parameter in the future The first estimated value and all the second estimated values ​​at a unit time are combined with the correlation coefficients of the non-tire pressure parameters and the tire pressure parameter to obtain the tire pressure parameter in the future in real time. The final estimated value of the unit time; finally, the tire pressure parameter is The final estimated value of each unit moment is transmitted to the display in real time, wherein the historical time series data contains the time series data. The unit time and the corresponding parameters are described in The historical parameter values ​​of each unit moment in the unit moments, represents a positive integer greater than or equal to 3, Indicates less than A positive integer of ; According to the tire pressure parameter in the future The first estimated value and all the second estimated values ​​at a unit time are combined with the correlation coefficients of the non-tire pressure parameters and the tire pressure parameter to obtain the tire pressure parameter in the future in real time. The final estimate of the unit moment includes: At each unit moment in the unit moments, the final estimated value of the tire pressure parameter at the corresponding moment is calculated according to the following formula: In the formula, Indicates less than or equal to A positive integer, Indicates that the tire pressure parameter in the future Unit time The final estimate at the moment, represents the total number of parameters of the non-tire pressure parameters, Indicates less than or equal to A positive integer, Indicates The influence weight coefficient corresponding to the non-tire pressure parameters is: Indicates that The correlation coefficients corresponding to non-tire pressure parameters are: Indicates less than or equal to A positive integer, Indicates The correlation coefficients corresponding to non-tire pressure parameters are: Indicates that the tire pressure parameter is in the The first estimated value at the moment, Indicates that The tire pressure parameter corresponds to the tire pressure parameter in the first The second estimated value at the time instant; The display is used to output and display the data from the processor in real time.

2. The aircraft tire pressure monitoring data analysis system according to claim 1, characterized in that: The multiple non-tire pressure parameter collectors include a wheel temperature collector, a wheel speed collector, an aircraft speed collector and / or a flight altitude collector.

3. The aircraft tire pressure monitoring data analysis system according to claim 1, characterized in that: After generating the historical time series data of the tire pressure parameters and the non-tire pressure parameters, the processor is further configured to: The unit moments in the historical time series data of the non-tire pressure parameters are shifted as a whole so that each shifted unit moment corresponds to each unit moment in the historical time series data of the tire pressure parameters, and for each shifted unit moment, the corresponding and updated historical time series data is calculated according to the following formula : In the formula, represents the total number of moments in the historical time series data of the tire pressure parameter, Indicates less than or equal to A positive integer, Indicates The non-tire pressure parameter value corresponding to the unit time after translation, Indicates the The time difference between the first moment in the historical time series data of the tire pressure parameter after the translation of the unit moment, , , and Respectively represent the historical time series data of the non-tire pressure parameters and One-to-one correspondence between adjacent units The coefficients of the local cubic polynomial for implementing cubic spline interpolation are obtained by using the least squares method. represents a positive integer greater than or equal to 4, The adjacent recording time is the distance from the The nearest previous unit moment after translation Unit moment.

4. The aircraft tire pressure monitoring data analysis system according to claim 1, characterized in that: For each non-tire pressure parameter corresponding to each collector, a correlation coefficient between the corresponding parameter and the tire pressure parameter is calculated according to the corresponding historical time series data and the historical time series data of the tire pressure parameter, including: For a non-tire pressure parameter, randomly extract a time series data segment from the corresponding historical time series data to form the first sample , and also extracting the first sample from the historical time series data of the tire pressure parameter Another time series data of the same period constitutes the second sample ; For the first sample Perform a normal distribution KS test to calculate the first test statistic value , and also for the second sample Perform the normal distribution KS check to calculate the second check statistic value ; If the first check statistic value and the second check statistic value are greater than the preset threshold, the first sample is calculated according to the following formula With the sample The correlation coefficient : In the formula, represents a positive integer, In the first sample The Sample values, In the second sample The Sample values, Indicates the first sample The sample mean of The second sample The sample mean of ; The correlation coefficient As the correlation coefficient between the certain non-tire pressure parameter and the tire pressure parameter.

5. The aircraft tire pressure monitoring data analysis system according to claim 1, characterized in that: According to the historical time series data of the tire pressure parameters, a machine learning algorithm is used to predict the tire pressure parameters in the future in real time. The first estimate of the unit time includes: The first sample data is extracted from the historical time series data of the tire pressure parameter in the following manner: extracting the time series data of the tire pressure parameter that are sequentially continuous in time series. The tire pressure historical parameter value of each unit moment is used as the first model input item, and the tire pressure parameter located at the unit moment is extracted from the historical time series data of the tire pressure parameter. After unit periods and sequentially in time The tire pressure historical parameter values ​​of the unit time period are used as the first model output item, and then the first model input item and the first model output item are used as a first sample data, wherein, Indicates greater than or equal to 2 and less than A positive integer, Indicates less than or equal to A positive integer of ; Applying all of the first sample data to perform model training on an artificial intelligence model based on a machine learning algorithm to obtain a first tire pressure value prediction model; Extract the most recent values ​​that are sequentially continuous in time series from the historical time series data of the tire pressure parameters. The tire pressure historical parameter values ​​of each unit moment are imported into the first tire pressure value prediction model, and the tire pressure parameter is output in the future. The first estimate of the unit time.

6. The aircraft tire pressure monitoring data analysis system according to claim 1, characterized in that: For each non-tire pressure parameter, according to the corresponding historical time series data and the historical time series data of the tire pressure parameter, the machine learning algorithm is used to predict in real time the corresponding tire pressure parameter in the future. The second estimated value of the unit time includes: For each non-tire pressure parameter, the corresponding second sample data is extracted from the corresponding historical time series data in the following manner: extract the consecutive second sample data in time series from the corresponding historical time series data. The non-tire pressure historical parameter value at each unit time is used as the second model input item, and the non-tire pressure historical parameter value at each unit time is extracted from the historical time series data of the tire pressure parameter. After unit periods and sequentially in time The tire pressure historical parameter value of each unit period is used as the second model output item, and then the second model input item and the second model output item are used as a second sample data, wherein, Indicates greater than or equal to 2 and less than A positive integer, Indicates less than or equal to A positive integer of ; For each of the non-tire pressure parameters, using all of the corresponding second sample data to perform model training on an artificial intelligence model based on the machine learning algorithm to obtain a second tire pressure value prediction model; For each non-tire pressure parameter, extract the most recent parameters that are consecutive in time sequence from the corresponding historical time series data. The non-tire pressure historical parameter value of each unit moment is imported into the second tire pressure value prediction model, and the corresponding tire pressure parameter is output in the future. The second estimated value of the unit time.

7. The aircraft tire pressure monitoring data analysis system according to claim 5 or 6, characterized in that: The artificial intelligence model based on machine learning algorithm adopts LSTM model, Bi-LSTM model or Attention-LSTM model.

8. The aircraft tire pressure monitoring data analysis system according to claim 1, characterized in that: After calculating the correlation coefficients between each non-tire pressure parameter and the tire pressure parameter, the processor is further configured to: At the end of each flight mission of the aircraft to which the aircraft tire belongs, for each non-tire pressure parameter corresponding to each collector, the correlation coefficient between the corresponding parameter and the tire pressure parameter is updated and calculated based on the corresponding historical time series data generated during the execution of the most recent flight mission and the tire pressure parameter and the historical time series data generated during the execution of the most recent flight mission.

9. The aircraft tire pressure monitoring data analysis system according to claim 1, characterized in that: The tire pressure parameters are calculated in real time in the future After obtaining the final estimated value of the unit time, the processor is further configured to: According to the tire pressure parameter in the future The final estimated value of each unit moment is used to obtain the maximum tire pressure estimate or the average tire pressure estimate; According to a plurality of tire pressure value ranges that are preset and correspond one to one with a plurality of tire pressure warning levels, determining a certain tire pressure value range that includes the maximum tire pressure estimate value or the average tire pressure estimate value from the plurality of tire pressure value ranges; The preset alarm action of a certain tire pressure warning level corresponding to the certain tire pressure value range is triggered in real time.

Images