Aircraft fuel oil measuring method, system and equipment based on deep learning and medium
The deep learning-based aircraft fuel measurement method enhances real-time accuracy and stability by preprocessing data, constructing time matrices, and using Transformer and recurrent networks with Bayesian algorithms to reduce errors.
Patent Information
- Application Number
- CN202411293827.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-14
- Publication Date
- 2025-01-03
AI Technical Summary
Existing aircraft fuel measurement methods based on neural networks do not fully consider the time dimension, leading to insufficient accuracy and lack of real-time measurement.
Aircraft fuel measurement method utilizing deep learning that preprocesses multi-dimensional data, constructs a time matrix, and employs a Transformer neural network and recurrent neural network to obtain fuel remaining values, with a Bayesian algorithm for error reduction.
Improves real-time performance and measurement accuracy by leveraging time series information and reducing measurement errors through correlation analysis and Bayesian fusion.
Smart Images
Figure CN119250123A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of aviation fuel measurement, and in particular to an aircraft fuel measurement method, system, equipment and medium based on deep learning. Background Art
[0002] The aviation electromechanical system is an important part of the aircraft, and its working performance and working status are directly related to the reliability and safety of the aircraft system. As one of the important components of the aviation electromechanical system, the fuel system has the characteristics of high precision and high reliability. It is necessary to accurately grasp the information on the amount of fuel in the fuel tank to ensure that it can supply fuel according to the engine demand at any altitude and any attitude. At present, aircraft fuel measurement technologies include mechanical, radiation, ultrasonic, and capacitance-based methods. Among them, the most widely used is capacitive fuel measurement technology, which uses the difference in dielectric constants between air and fuel for measurement. Fuel data includes the height of the fuel liquid level inside the fuel tank obtained by the above-mentioned capacitive sensor and other methods. The aircraft fuel system analyzes and calculates the fuel data to obtain the real-time remaining fuel.
[0003] The current existing methods for measuring aircraft fuel based on neural networks do not fully consider the time dimension, and the measurement is not real-time enough, which leads to insufficient accuracy. Summary of the invention
[0004] Aiming at the problem that the existing method for measuring aircraft fuel based on neural network does not fully consider the time dimension, the measurement is not real-time enough, and thus the accuracy is insufficient, the present invention proposes an aircraft fuel measurement method, system, equipment and medium based on deep learning; the method first pre-processes the acquired multi-dimensional aircraft fuel data; then constructs a multi-dimensional aircraft fuel time matrix according to the set time period and the pre-processed multi-dimensional aircraft fuel data; finally, the aircraft fuel time matrix is respectively input into the constructed Transformer neural network and the recurrent neural network to obtain the first aircraft fuel remaining oil value and the second aircraft fuel remaining oil value, and the Bayesian algorithm is called to obtain the true aircraft fuel remaining oil value; the measurement error of the aircraft fuel is reduced, and the real-time and stability of the aircraft fuel measurement are improved by using the time series information.
[0005] The specific implementation contents of the present invention are as follows:
[0006] A method for measuring aircraft fuel based on deep learning, comprising the following steps:
[0007] Step S1: preprocessing the acquired multi-dimensional aircraft fuel data;
[0008] Step S2: constructing a multidimensional aircraft fuel time matrix according to the set time period and the pre-processed multidimensional aircraft fuel data;
[0009] Step S3: Input the aircraft fuel time matrix into the constructed Transformer neural network and recurrent neural network respectively, obtain the first aircraft fuel remaining value and the second aircraft fuel remaining value, and call the Bayesian algorithm to fuse and obtain the actual aircraft fuel remaining value.
[0010] In order to better implement the present invention, further, the step S1 specifically includes the following steps:
[0011] Step S11: Acquire aircraft fuel data from a ground fuel tank test bench; the aircraft fuel data includes multi-dimensional aircraft fuel data and original aircraft fuel remaining fuel data;
[0012] Step S12: Calculate the covariance and standard deviation of the multi-dimensional aircraft fuel data and the original aircraft fuel remaining fuel data;
[0013] Step S13: calculating the linear correlation between the aircraft fuel time data and the original aircraft fuel remaining amount data according to the covariance and standard deviation of the multi-dimensional aircraft fuel data and the original aircraft fuel remaining amount data;
[0014] Step S14: sorting the multidimensional aircraft fuel data according to the calculated linear correlation to obtain preprocessed multidimensional aircraft fuel data.
[0015] In order to better implement the present invention, further, the multi-dimensional aircraft fuel data includes time data, fuel tank oil level data, fuel supply tank oil level data, pitch angle data, and roll angle data.
[0016] In order to better implement the present invention, further, the step S2 specifically includes the following steps:
[0017] Step S21: converting the pre-processed multi-dimensional aircraft fuel data into multi-dimensional aircraft fuel time data according to a set cycle;
[0018] Step S22: constructing a multidimensional aircraft fuel time matrix according to the multidimensional aircraft fuel time data.
[0019] In order to better implement the present invention, further, the step S3 specifically includes the following steps:
[0020] Step S31: acquiring multidimensional aircraft fuel sequence data from the multidimensional aircraft fuel time matrix;
[0021] Step S32: constructing a Transformer neural network; the Transformer neural network includes an encoder and a decoder;
[0022] Step S33: inputting the acquired multi-dimensional aircraft fuel sequence data into the constructed Transformer neural network to obtain the first aircraft fuel remaining oil value;
[0023] Step S34: inputting the acquired multi-dimensional aircraft fuel sequence data into the recurrent neural network to obtain a second aircraft fuel remaining oil value;
[0024] Step S35: according to the remaining fuel amount value of the first aircraft and the remaining fuel amount value of the second aircraft, the Bayesian algorithm is used to calculate the absolute error value;
[0025] Step S36: Calculate the true value of the remaining fuel amount of the aircraft according to the absolute error value and the original remaining fuel amount data of the aircraft.
[0026] In order to better implement the present invention, further, the encoder of the Transformer neural network in step S32 includes a first multi-head attention module, a first residual network module, a first feedforward neural network module, and a second residual network module connected in sequence;
[0027] The input end of the first multi-head attention module inputs multi-dimensional aircraft fuel sequence data;
[0028] An output end of the second residual network module is connected to the encoder.
[0029] In order to better implement the present invention, further, the decoder of the Transformer neural network in step S32 includes a mask multi-head attention module, a third residual network module, a second multi-head attention module, a fourth residual network module, a second feedforward neural network module, and a fifth residual network module connected in sequence;
[0030] The input end of the masked multi-head attention module inputs multi-dimensional aircraft fuel sequence data;
[0031] An input end of the second multi-head attention module is connected to an output end of the second residual network module.
[0032] Based on the above-mentioned aircraft fuel measurement method based on deep learning, in order to better realize the present invention, further, a deep learning-based aircraft fuel measurement system is proposed, which is used to execute the above-mentioned deep learning-based aircraft fuel measurement method; comprising a preprocessing unit, a timing unit, and a measurement unit;
[0033] The preprocessing unit is used to preprocess the acquired multi-dimensional aircraft fuel data;
[0034] The timing unit is used to construct a multi-dimensional aircraft fuel time matrix according to a set time period and pre-processed multi-dimensional aircraft fuel data;
[0035] The measuring unit is used to input the aircraft fuel time matrix into the constructed Transformer neural network and the recurrent neural network respectively, obtain the first aircraft fuel remaining oil value and the second aircraft fuel remaining oil value, and call the Bayesian algorithm to fuse and obtain the actual aircraft fuel remaining oil value.
[0036] Based on the above-mentioned aircraft fuel measurement method based on deep learning, in order to better realize the present invention, further, an electronic device is proposed, including a memory and a processor; a computer program is stored on the memory; when the computer program is executed on the processor, the above-mentioned aircraft fuel measurement method based on deep learning is implemented.
[0037] Based on the above-mentioned aircraft fuel measurement method based on deep learning, in order to better realize the present invention, further, a computer-readable storage medium is proposed, on which computer instructions are stored; when the computer instructions are executed on the above-mentioned electronic device, the above-mentioned aircraft fuel measurement method based on deep learning is implemented.
[0038] The present invention has the following beneficial effects:
[0039] (1) The present invention improves the real-time performance of aircraft fuel measurement by converting multi-dimensional aircraft fuel data into time series data and utilizing time series information. It also enhances the stability and reliability of measurement by calculating the correlation between the multi-dimensional aircraft fuel data and the original remaining fuel quantity data of the aircraft fuel.
[0040] (2) The present invention constructs a Transformer neural network and a recurrent neural network to process time series data respectively, and calls the Bayesian algorithm to further calculate the mean square error, thereby reducing the measurement error and improving the measurement accuracy. BRIEF DESCRIPTION OF THE DRAWINGS
[0041] Figure 1 This is a schematic block diagram of the overall process of the aircraft fuel measurement method based on deep learning provided by the present invention.
[0042] Figure 2 This is a schematic block diagram of the encoder structure provided by the present invention.
[0043] Figure 3 This is a schematic block diagram of the decoder structure provided by the present invention. DETAILED DESCRIPTION
[0044] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. It should be understood that the described embodiments are only part of the embodiments of the present invention, not all of the embodiments, and therefore should not be regarded as limiting the scope of protection. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technical personnel in this field without making creative work are within the scope of protection of the present invention.
[0045] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "disposed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be an indirect connection through an intermediate medium, or it can be the internal communication of two elements. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.
[0046] Embodiment 1:
[0047] This embodiment proposes an aircraft fuel measurement method based on deep learning, which specifically includes the following steps:
[0048] Step S1: pre-processing the acquired multi-dimensional aircraft fuel data.
[0049] The step S1 specifically includes the following steps:
[0050] Step S11: Acquire aircraft fuel data from a ground fuel tank test bench; the aircraft fuel data includes multi-dimensional aircraft fuel data and original aircraft fuel remaining fuel data;
[0051] Furthermore, the multi-dimensional aircraft fuel data includes time data, fuel transfer tank oil level data, fuel supply tank oil level data, pitch angle data, and roll angle data.
[0052] Step S12: Calculate the covariance and standard deviation of the multi-dimensional aircraft fuel data and the original aircraft fuel remaining fuel data;
[0053] Step S13: calculating the linear correlation between the aircraft fuel time data and the original aircraft fuel remaining amount data according to the covariance and standard deviation of the multi-dimensional aircraft fuel data and the original aircraft fuel remaining amount data;
[0054] Step S14: sorting the multidimensional aircraft fuel data according to the calculated linear correlation to obtain preprocessed multidimensional aircraft fuel data.
[0055] Step S2: construct a multidimensional aircraft fuel time matrix according to the set time period and the preprocessed multidimensional aircraft fuel data.
[0056] The step S2 specifically includes the following steps:
[0057] Step S21: converting the pre-processed multi-dimensional aircraft fuel data into multi-dimensional aircraft fuel time data according to a set cycle;
[0058] Step S22: constructing a multidimensional aircraft fuel time matrix according to the multidimensional aircraft fuel time data.
[0059] Step S3: Input the aircraft fuel time matrix into the constructed Transformer neural network and recurrent neural network respectively, obtain the first aircraft fuel remaining value and the second aircraft fuel remaining value, and call the Bayesian algorithm to fuse and obtain the actual aircraft fuel remaining value.
[0060] The step S3 specifically comprises the following steps:
[0061] Step S31: acquiring multidimensional aircraft fuel sequence data from the multidimensional aircraft fuel time matrix;
[0062] Step S32: constructing a Transformer neural network; the Transformer neural network includes an encoder and a decoder;
[0063] Furthermore, the encoder of the Transformer neural network in step S32 includes a first multi-head attention module, a first residual network module, a first feedforward neural network module, and a second residual network module connected in sequence;
[0064] The input end of the first multi-head attention module inputs multi-dimensional aircraft fuel sequence data;
[0065] An output end of the second residual network module is connected to the encoder.
[0066] Furthermore, the decoder of the Transformer neural network in step S32 includes a masked multi-head attention module, a third residual network module, a second multi-head attention module, a fourth residual network module, a second feedforward neural network module, and a fifth residual network module connected in sequence;
[0067] The input end of the masked multi-head attention module inputs multi-dimensional aircraft fuel sequence data;
[0068] An input end of the second multi-head attention module is connected to an output end of the second residual network module.
[0069] Step S33: inputting the acquired multi-dimensional aircraft fuel sequence data into the constructed Transformer neural network to obtain the first aircraft fuel remaining oil value;
[0070] Step S34: inputting the acquired multi-dimensional aircraft fuel sequence data into the recurrent neural network to obtain a second aircraft fuel remaining oil value;
[0071] Step S35: according to the remaining fuel amount value of the first aircraft and the remaining fuel amount value of the second aircraft, the Bayesian algorithm is used to calculate the absolute error value;
[0072] Step S36: Calculate the true value of the remaining fuel amount of the aircraft according to the absolute error value and the original remaining fuel amount data of the aircraft.
[0073] Working principle: This embodiment first pre-processes the acquired multi-dimensional aircraft fuel data; then constructs a multi-dimensional aircraft fuel time matrix according to the set time period and the pre-processed multi-dimensional aircraft fuel data; finally, the aircraft fuel time matrix is respectively input into the constructed Transformer neural network and the recurrent neural network to obtain the first aircraft fuel remaining oil value and the second aircraft fuel remaining oil value, and calls the Bayesian algorithm to obtain the actual aircraft fuel remaining oil value; reduces the measurement error of aircraft fuel, and uses the time series information to improve the real-time and stability of aircraft fuel measurement.
[0074] Embodiment 2:
[0075] This embodiment is based on the above embodiment 1. Figure 1 , Figure 2 , Figure 3 As shown, a specific embodiment is described in detail.
[0076] Step S1: pre-processing the acquired multi-dimensional aircraft fuel data.
[0077] First, aircraft fuel data is obtained from a ground fuel tank test bench. The aircraft fuel data collected in this embodiment includes time data, boost pressure data, ambient pressure data, fuel supply pressure data, instantaneous fuel supply flow rate data, fuel transfer tank oil level data, fuel supply tank oil level data, pitch angle data, roll angle data and aircraft fuel original remaining fuel quantity data.
[0078] Then the aircraft fuel data is analyzed by calculating the correlation, in order to reduce the dimension and ensure the data accuracy.
[0079] According to the covariance and standard deviation of the multi-dimensional aircraft fuel data and the original aircraft fuel remaining fuel data, the linear correlation between the aircraft fuel time data and the original aircraft fuel remaining fuel data is calculated, and the calculated linear correlation takes an absolute value;
[0080] According to the calculated linear correlation, the data with lower correlation is removed and the data with higher linear correlation is retained, and the data are sorted to obtain the preprocessed multidimensional aircraft fuel data.
[0081] In this embodiment, the pre-processed multi-dimensional aircraft fuel data includes time data, fuel transfer tank oil level data, fuel supply tank oil level data, pitch angle data, and roll angle data.
[0082] Step S2: construct a multidimensional aircraft fuel time matrix according to the set time period and the preprocessed multidimensional aircraft fuel data.
[0083] Firstly, the pre-processed multi-dimensional aircraft fuel data is converted into multi-dimensional aircraft fuel time data according to a set period;
[0084] Multi-dimensional aircraft fuel time data = (Ti, Hi, hi, Pi, Ri);
[0085] Among them, Ti represents the set time period, Hi represents the oil level height of the fuel delivery tank, hi represents the oil level height of the fuel supply tank, Pi represents the pitch angle, and Ri represents the roll angle.
[0086] According to the multi-dimensional aircraft fuel time data, a multi-dimensional aircraft fuel time matrix is constructed.
[0087] Step S3: Input the aircraft fuel time matrix into the constructed Transformer neural network and recurrent neural network respectively, obtain the first aircraft fuel remaining value and the second aircraft fuel remaining value, and call the Bayesian algorithm to fuse and obtain the actual aircraft fuel remaining value.
[0088] The step S3 specifically comprises the following steps:
[0089] Step S31: acquiring multidimensional aircraft fuel sequence data from the multidimensional aircraft fuel time matrix;
[0090] Step S32: constructing a Transformer neural network; the Transformer neural network includes an encoder and a decoder;
[0091] like Figure 2 As shown, the encoder includes a first multi-head attention module, a first residual network module, a first feedforward neural network module, and a second residual network module connected in sequence;
[0092] The input end of the first multi-head attention module inputs multi-dimensional aircraft fuel sequence data;
[0093] An output end of the second residual network module is connected to the encoder.
[0094] like Figure 3 As shown, the decoder includes a mask multi-head attention module, a third residual network module, a second multi-head attention module, a fourth residual network module, a second feedforward neural network module, and a fifth residual network module connected in sequence;
[0095] The input end of the masked multi-head attention module inputs multi-dimensional aircraft fuel sequence data;
[0096] An input end of the second multi-head attention module is connected to an output end of the second residual network module.
[0097] Step S33: inputting the acquired multi-dimensional aircraft fuel sequence data into the constructed Transformer neural network to obtain the first aircraft fuel remaining oil value;
[0098] Step S34: inputting the acquired multi-dimensional aircraft fuel sequence data into the recurrent neural network to obtain a second aircraft fuel remaining oil value;
[0099] Step S35: according to the remaining fuel amount value of the first aircraft and the remaining fuel amount value of the second aircraft, the Bayesian algorithm is used to calculate the absolute error value;
[0100] Step S36: Calculate the true value of the remaining fuel amount of the aircraft according to the absolute error value and the original remaining fuel amount data of the aircraft.
[0101] Working principle: This embodiment improves the real-time performance of aircraft fuel measurement by converting multi-dimensional aircraft fuel data into time series data and utilizing time series information. It also enhances the stability and reliability of measurement by calculating the correlation between multi-dimensional aircraft fuel data and the original remaining fuel volume data of aircraft fuel. It also constructs a Transformer neural network and a recurrent neural network to process time series data respectively, and calls the Bayesian algorithm to further calculate the mean square error, thereby reducing measurement errors and improving measurement accuracy.
[0102] The other parts of this embodiment are the same as those of the above-mentioned embodiment 1, and thus will not be described in detail.
[0103] Embodiment 3:
[0104] This embodiment, based on any one of the above-mentioned embodiments 1-2, proposes an aircraft fuel measurement system based on deep learning, which is used to execute the above-mentioned aircraft fuel measurement method based on deep learning; it includes a preprocessing unit, a timing unit, and a measurement unit;
[0105] The preprocessing unit is used to preprocess the acquired multi-dimensional aircraft fuel data;
[0106] The timing unit is used to construct a multi-dimensional aircraft fuel time matrix according to a set time period and pre-processed multi-dimensional aircraft fuel data;
[0107] The measuring unit is used to input the aircraft fuel time matrix into the constructed Transformer neural network and the recurrent neural network respectively, obtain the first aircraft fuel remaining oil value and the second aircraft fuel remaining oil value, and call the Bayesian algorithm to fuse and obtain the actual aircraft fuel remaining oil value.
[0108] This embodiment also proposes an electronic device, including a memory and a processor; a computer program is stored on the memory; when the computer program is executed on the processor, the above-mentioned aircraft fuel measurement method based on deep learning is implemented.
[0109] This embodiment also proposes a computer-readable storage medium, on which computer instructions are stored; when the computer instructions are executed on the above-mentioned electronic device, the above-mentioned aircraft fuel measurement method based on deep learning is implemented.
[0110] The other parts of this embodiment are the same as any one of the above-mentioned embodiments 1-2, so they will not be repeated here.
[0111] The processor involved in the embodiments of the present application may be a chip. For example, it may be a field programmable gate array (FPGA), an application specific integrated circuit (ASIC), a system on chip (SoC), a central processor unit (CPU), a network processor (NP), a digital signal processor (DSP), a microcontroller unit (MCU), a programmable logic device (PLD), or other integrated chips.
[0112] The memory involved in the embodiments of the present application may be a volatile memory or a non-volatile memory, or may include both volatile and non-volatile memories. Among them, the non-volatile memory may be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. The volatile memory may be a random access memory (RAM), which is used as an external cache. By way of example and not limitation, many forms of RAM are available, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDR SDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), and direct RAM (DR RAM). It should be noted that the memory of the systems and methods described herein is intended to include, but is not limited to, these and any other suitable types of memory.
[0113] It should be understood that in the various embodiments of the present application, the size of the serial numbers of the above-mentioned processes does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.
[0114] Those of ordinary skill in the art will appreciate that the modules and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.
[0115] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and modules described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.
[0116] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the modules is only a logical function division. There may be other division methods in actual implementation, such as multiple modules or components can be combined or integrated into another device, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or modules, which can be electrical, mechanical or other forms.
[0117] The modules described as separate components may or may not be physically separated, and the components shown as modules may or may not be physical modules, that is, they may be located in one device or distributed on multiple devices. Some or all of the modules may be selected according to actual needs to achieve the purpose of the present embodiment.
[0118] In addition, each functional module in each embodiment of the present application may be integrated into one device, or each module may exist physically separately, or two or more modules may be integrated into one device.
[0119] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented using a software program, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When loading and executing computer program instructions on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions may be stored in a computer-readable storage medium, or transmitted from one computer-readable storage medium to another computer-readable storage medium, for example, the computer instructions may be transmitted from a website site, computer, server or data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (Digital Subscriber Line, DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) mode to another website site, computer, server or data center. The computer-readable storage medium may be any available medium that a computer can access or may contain one or more servers, data centers and other data storage devices that can be integrated with a medium. The available medium may be a magnetic medium (eg, a floppy disk, a hard disk, a magnetic tape), an optical medium (eg, a DVD), or a semiconductor medium (eg, a solid state disk (SSD)).
[0120] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.
Claims
1. A method for measuring aircraft fuel based on deep learning, characterized in that: The specific steps include: Step S1: preprocessing the acquired multi-dimensional aircraft fuel data; Step S2: constructing a multidimensional aircraft fuel time matrix according to the set time period and the pre-processed multidimensional aircraft fuel data; Step S3: Input the aircraft fuel time matrix into the constructed Transformer neural network and recurrent neural network respectively, obtain the first aircraft fuel remaining value and the second aircraft fuel remaining value, and call the Bayesian algorithm to fuse and obtain the actual aircraft fuel remaining value.
2. The aircraft fuel measurement method based on deep learning according to claim 1, characterized in that: The step S1 specifically includes the following steps: Step S11: Acquire aircraft fuel data from a ground fuel tank test bench; the aircraft fuel data includes multi-dimensional aircraft fuel data and original aircraft fuel remaining fuel data; Step S12: Calculate the covariance and standard deviation of the multi-dimensional aircraft fuel data and the original aircraft fuel remaining fuel data; Step S13: calculating the linear correlation between the aircraft fuel time data and the original aircraft fuel remaining amount data according to the covariance and standard deviation of the multi-dimensional aircraft fuel data and the original aircraft fuel remaining amount data; Step S14: sorting the multidimensional aircraft fuel data according to the calculated linear correlation to obtain preprocessed multidimensional aircraft fuel data.
3. The aircraft fuel measurement method based on deep learning according to claim 2, characterized in that: The multi-dimensional aircraft fuel data includes time data, fuel tank oil level data, fuel supply tank oil level data, pitch angle data, and roll angle data.
4. The aircraft fuel measurement method based on deep learning according to claim 3 is characterized in that: The step S2 specifically includes the following steps: Step S21: converting the pre-processed multi-dimensional aircraft fuel data into multi-dimensional aircraft fuel time data according to a set cycle; Step S22: constructing a multidimensional aircraft fuel time matrix according to the multidimensional aircraft fuel time data.
5. The aircraft fuel measurement method based on deep learning according to claim 4, characterized in that: The step S3 specifically comprises the following steps: Step S31: acquiring multidimensional aircraft fuel sequence data from the multidimensional aircraft fuel time matrix; Step S32: constructing a Transformer neural network; the Transformer neural network includes an encoder and a decoder; Step S33: inputting the acquired multi-dimensional aircraft fuel sequence data into the constructed Transformer neural network to obtain the first aircraft fuel remaining oil value; Step S34: inputting the acquired multi-dimensional aircraft fuel sequence data into the recurrent neural network to obtain a second aircraft fuel remaining oil value; Step S35: according to the remaining fuel amount value of the first aircraft and the remaining fuel amount value of the second aircraft, the Bayesian algorithm is used to calculate the absolute error value; Step S36: Calculate the true value of the remaining fuel amount of the aircraft according to the absolute error value and the original remaining fuel amount data of the aircraft.
6. The aircraft fuel measurement method based on deep learning according to claim 5, characterized in that: The encoder of the Transformer neural network in step S32 includes a first multi-head attention module, a first residual network module, a first feedforward neural network module, and a second residual network module connected in sequence; The input end of the first multi-head attention module inputs multi-dimensional aircraft fuel sequence data; An output end of the second residual network module is connected to the encoder.
7. The aircraft fuel measurement method based on deep learning according to claim 6, characterized in that: The decoder of the Transformer neural network in step S32 includes a masked multi-head attention module, a third residual network module, a second multi-head attention module, a fourth residual network module, a second feedforward neural network module, and a fifth residual network module connected in sequence; The input end of the masked multi-head attention module inputs multi-dimensional aircraft fuel sequence data; An input end of the second multi-head attention module is connected to an output end of the second residual network module.
8. An aircraft fuel measurement system based on deep learning, used to execute the aircraft fuel measurement method based on deep learning as claimed in claim 1; characterized in that: It includes a pre-processing unit, a timing unit, and a measuring unit; The preprocessing unit is used to preprocess the acquired multi-dimensional aircraft fuel data; The timing unit is used to construct a multi-dimensional aircraft fuel time matrix according to a set time period and pre-processed multi-dimensional aircraft fuel data; The measuring unit is used to input the aircraft fuel time matrix into the constructed Transformer neural network and the recurrent neural network respectively, obtain the first aircraft fuel remaining oil value and the second aircraft fuel remaining oil value, and call the Bayesian algorithm to fuse and obtain the actual aircraft fuel remaining oil value.
9. An electronic device, characterized in that: It comprises a memory and a processor; a computer program is stored in the memory; when the computer program is executed on the processor, the aircraft fuel measurement method based on deep learning as described in any one of claims 1 to 7 is implemented.
10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores computer instructions; when the computer instructions are executed on the electronic device as described in claim 9, the aircraft fuel measurement method based on deep learning as described in any one of claims 1-7 is implemented.