Flight guide generation method and device, electronic equipment and medium

Abstract

The embodiment of the invention provides a flight guidance generation method and device, electronic equipment and a medium, and relates to the field of flight simulation and the like. The method is executed by a flight guidance generation model, the model comprises a data routing module, a data prediction module, an operation evaluation module and a voice output module, and the method comprises the following steps: when the model executes flight guidance information generation, the data prediction module obtains historical flight data and actual flight data of a pilot sent by the data routing module; the data prediction module performs flight data prediction processing based on the historical flight data to obtain predicted flight data; and the data prediction module sends the predicted flight data and the actual flight data to the operation evaluation module and the voice output module, so that the operation evaluation module and the voice output module output flight guidance information based on the predicted flight data and the actual flight data. Based on the historical operation and the actual operation of the pilot, the flight guidance information is provided for the pilot, and the training efficiency of the pilot is improved.

Description

Technical Field

[0001] The present disclosure relates to the technical field of flight simulation evaluation, and in particular to a flight guidance generation method, device, electronic equipment and medium. Background Art

[0002] At present, in order to improve their flight operation skills, pilots generally conduct flight operation training through flight simulators.

[0003] In the prior art, when using flight simulators for flight training, the main method is to have real instructors on-site guide students. Real instructors directly observe students' flight operations and evaluate and guide students' operations during the flight based on their own flight experience. Due to manual guidance, instructors cannot guarantee to fully observe all the flight operation details of students in each evaluation and guidance. The flight evaluation of students will lack objectivity and incomplete guidance. In addition, human resources are precious and the cost is high.

[0004] Therefore, how to effectively and objectively guide pilots to conduct flight operation training has become an urgent problem to be solved. Summary of the invention

[0005] The embodiments of the present disclosure provide a flight guidance generation method, device, electronic device and medium. The embodiments of the present disclosure provide flight guidance information to the pilot based on the pilot's historical operations and actual operations, thereby effectively improving the pilot's training efficiency.

[0006] On the one hand, an embodiment of the present disclosure provides a flight guidance method, which is executed by a flight guidance generation model, wherein the flight guidance generation model includes a data routing module, a data prediction module, an operation evaluation module, and a voice output module, and the method includes: In the case where the flight guidance generation model generates flight guidance information, the data prediction module obtains the pilot's historical flight data and actual flight data sent by the data routing module; The data prediction module performs flight data prediction processing based on the acquired historical flight data to obtain predicted flight data; The data prediction module sends the predicted flight data and the actual flight data to the operation evaluation module and the voice output module, so that the operation evaluation module and the voice output module output flight guidance information based on the predicted flight data and the actual flight data, wherein the flight guidance information is used to guide the pilot to perform flight operations according to the flight guidance information.

[0007] In an optional embodiment, the flight data prediction processing is performed based on the historical flight data obtained to obtain predicted flight data, including: calling each flight data stored in a database, wherein the database contains at least two flight data; sampling the historical flight data to obtain sampled historical flight data, recorded as sampled flight data; preprocessing the each flight data and the sampled flight data, and performing prediction based on the preprocessed flight data and the preprocessed sampled flight data to obtain the predicted flight data.

[0008] In an optional embodiment, the above-mentioned flight data and the above-mentioned sampled flight data are preprocessed, and prediction is performed based on the preprocessed flight data and the preprocessed sampled flight data to obtain the above-mentioned predicted flight data, including: slicing operation processing on each of the above-mentioned flight data to obtain at least one group of data with the same dimension as the above-mentioned sampled flight data, recorded as same-dimensional data; determining similar flight data similar to the above-mentioned sampled flight data in the above-mentioned same-dimensional data; standardizing the above-mentioned similar flight data and the above-mentioned sampled flight data to obtain standardized similar flight data and standardized sampled flight data; linearly scaling and character encoding the above-mentioned standardized similar flight data and the above-mentioned standardized sampled flight data to obtain corresponding similar flight text data and sampled flight text data, wherein the above-mentioned similar flight text data is the preprocessed flight data, and the above-mentioned The sampled flight text data is the preprocessed sampled flight data; the similar flight text data are classified to obtain the corresponding horizontal flight data and vertical flight data; a fitting operation is performed based on the horizontal flight data and the vertical flight data to obtain the reference flight data; at least one flight operation parameter prompt word is created according to the similar flight text data, the reference flight data and the sampled flight text data, wherein each of the flight operation parameter prompt words is a standard example of a corresponding flight operation parameter, and each of the flight operation prompt words contains a prediction task of a corresponding flight operation parameter; at least one large model is called simultaneously by multiple processes, and for each called large model, flight data prediction is performed based on the flight operation parameter prompt word corresponding to the large model to obtain the predicted flight data, wherein the data prediction module includes at least one of the large models.

[0009] In an optional embodiment, the operation evaluation module and the voice output module output the flight guidance information based on the predicted flight data and the actual flight data according to the following steps: the operation evaluation module receives the predicted flight data sent by the data prediction module; the operation evaluation module creates a flight guidance prompt based on the predicted flight data and the actual flight data, wherein the flight guidance prompt includes relevant information of the flight operation and the flight evaluation standard; the operation evaluation module calls the big model, and performs an operation evaluation on the actual flight data based on the flight guidance prompt to obtain a flight guidance text, and sends the flight guidance text to the voice output module, wherein the big model is a module in the data prediction module; the voice output module receives the flight guidance text; the voice output module calls the multimodal big model to convert the flight guidance text into flight guidance audio information, wherein the multimodal big model is used to convert text information into audio information, and the flight guidance information includes the flight guidance audio information; the voice output module receives the flight guidance audio information returned by the multimodal, and outputs the flight guidance audio information.

[0010] In an optional embodiment, the above-mentioned similar flight data and the above-mentioned sampled flight data are standardized to obtain standardized similar flight data and standardized sampled flight data, including: determining the mean and standard deviation of each flight operation parameter contained in the above-mentioned flight data; based on the mean and standard deviation of each flight operation parameter, standardizing the above-mentioned similar flight data and the above-mentioned sampled flight data to obtain standardized similar flight data and standardized sampled flight data.

[0011] In an optional embodiment, the method further includes: displaying at least one of the historical flight data, the predicted flight data or the flight guidance information on a flight operation display interface.

[0012] In one aspect, an embodiment of the present disclosure provides a flight guidance generation device, the flight guidance generation device includes a flight guidance generation model, the flight guidance generation model includes a data routing module, a data prediction module, an operation evaluation module and a voice output module, the device includes: an acquisition module, used for acquiring the pilot's historical flight data and actual flight data sent by the data routing module of the flight guidance generation model through the data prediction module of the flight guidance generation model when the flight guidance generation model generates the flight guidance information; A first processing module, configured for the data prediction module to perform flight data prediction processing based on the acquired historical flight data to obtain predicted flight data; The second processing module is used for the data prediction module to send the predicted flight data and the actual flight data to the operation evaluation module and the voice output module, so that the operation evaluation module and the voice output module output flight guidance information based on the predicted flight data and the actual flight data, wherein the flight guidance information is used to guide the pilot to perform flight operations according to the flight guidance information.

[0013] In an optional embodiment, the above-mentioned first processing module is specifically used to: call each flight data stored in a database, wherein the above-mentioned database contains at least two flight data; perform sampling processing on the above-mentioned historical flight data to obtain sampled historical flight data, recorded as sampled flight data; pre-process the above-mentioned each flight data and the above-mentioned sampled flight data, and perform prediction based on the pre-processed flight data and the pre-processed sampled flight data to obtain the above-mentioned predicted flight data.

[0014] In an optional embodiment, the first processing module is specifically used to: perform slicing operations on the flight data to obtain at least one group of data with the same dimension as the sampled flight data, recorded as same-dimensional data; determine similar flight data similar to the sampled flight data in the same-dimensional data; standardize the similar flight data and the sampled flight data to obtain standardized similar flight data and standardized sampled flight data; perform linear scaling and character encoding on the standardized similar flight data and the standardized sampled flight data to obtain corresponding similar flight text data and sampled flight text data, wherein the similar flight text data is preprocessed flight data, and the sampled flight text data is preprocessed sampled flight data; The text data is classified to obtain corresponding lateral flight data and longitudinal flight data; a fitting operation is performed based on the lateral flight data and the longitudinal flight data to obtain reference flight data; at least one flight operation parameter prompt word is created according to the similar flight text data, the reference flight data and the sampled flight text data, wherein each of the flight operation parameter prompt words is a standard example of a corresponding flight operation parameter, and each of the flight operation prompt words contains a prediction task of a corresponding flight operation parameter; at least one large model is called simultaneously by multiple processes, and for each called large model, flight data prediction is performed based on the flight operation parameter prompt word corresponding to the large model to obtain the predicted flight data, wherein the data prediction module includes at least one of the large models.

[0015] In an optional embodiment, the second processing module is specifically used to perform the following steps: the operation evaluation module receives the predicted flight data sent by the data prediction module; the operation evaluation module creates a flight guidance prompt based on the predicted flight data and the actual flight data, wherein the flight guidance prompt includes relevant information of the flight operation and the flight evaluation standard; the operation evaluation module calls the big model, and performs an operation evaluation on the actual flight data based on the flight guidance prompt to obtain a flight guidance text, and sends the flight guidance text to the voice output module, wherein the big model is a module in the data prediction module; the voice output module receives the flight guidance text; the voice output module calls the multimodal big model to convert the flight guidance text into flight guidance audio information, wherein the multimodal big model is used to convert text information into audio information, and the flight guidance information includes the flight guidance audio information; the voice output module receives the flight guidance audio information returned by the multimodal, and outputs the flight guidance audio information.

[0016] In an optional embodiment, the first processing module is used to: determine the mean and standard deviation of each flight operation parameter contained in the flight data; based on the mean and standard deviation of each flight operation parameter, standardize the similar flight data and the sampled flight data to obtain standardized similar flight data and standardized sampled flight data.

[0017] In an optional embodiment, the above-mentioned device also includes: a display module, used to display at least one of the above-mentioned historical flight data, the above-mentioned predicted flight data or flight guidance information on the flight operation display interface.

[0018] On the one hand, an embodiment of the present disclosure provides an electronic device, which includes a processor and a memory, which are interconnected; the memory is used to store a computer program; the processor is configured to execute a method provided by any possible implementation of the above-mentioned flight guidance generation method when calling the above-mentioned computer program.

[0019] On the one hand, an embodiment of the present disclosure provides a computer-readable storage medium, which stores a computer program, and the computer program is executed by a processor to implement the method provided by any possible implementation of the above-mentioned flight guidance generation method.

[0020] On the one hand, an embodiment of the present disclosure provides a computer program product or a computer program, which includes computer instructions stored in a computer-readable storage medium. A processor of an electronic device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes the method provided by any possible implementation of the above-mentioned flight guidance generation method.

[0021] The technical solution provided by the present disclosure has the following beneficial effects: the flight guidance generation model is used for execution, and the model includes a data routing module, a data prediction module, an operation evaluation module and a voice output module, wherein, when the flight guidance generation model is executed to generate flight guidance information, the data prediction module obtains the pilot's historical flight data and actual flight data sent by the data routing module, and performs flight data prediction processing based on the acquired historical flight data to obtain predicted flight data, and the data prediction module sends the predicted flight data and actual flight data to the operation evaluation module and the voice output module, so that the operation evaluation module and the voice output module can output flight guidance information based on the predicted flight data and the actual flight data, and guide the pilot's next flight operation through the flight guidance information. Through the flight guidance generation model, the flight data generated during the pilot's flight, that is, the historical flight data and the actual flight data, are analyzed and processed, and objective flight guidance information is provided for the pilot's next flight operation, avoiding the existing technology of performing flight training by manual guidance, improving the objectivity of flight training, and the flight guidance model can provide flight guidance information in real time, so that the pilot can perform flight training according to the flight guidance information in a timely manner, and improving the efficiency of flight training. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solutions in the embodiments of the present disclosure, the drawings required for describing the embodiments of the present disclosure are briefly introduced below.

[0023] Figure 1 A flowchart of a flight guidance generation method provided by an embodiment of the present disclosure; Figure 2 A schematic diagram of the structure of a flight guidance generation model provided in an embodiment of the present disclosure; Figure 3 A schematic diagram of a data stream transmission principle provided by an embodiment of the present disclosure; Figure 4 A schematic diagram of a data preprocessing process provided by an embodiment of the present disclosure; Figure 5 Reference flight data and predicted flight data in a character encoding form provided by an embodiment of the present disclosure; Figure 6A schematic diagram of a visual data display interface provided by an embodiment of the present disclosure; Figure 7 A schematic diagram of the structure of a flight guidance generating device provided in an embodiment of the present disclosure; Figure 8 A schematic diagram of the structure of an electronic device for a flight guidance generating method provided in an embodiment of the present disclosure. DETAILED DESCRIPTION

[0024] The embodiments of the present disclosure are described below in conjunction with the drawings in the present disclosure. It should be understood that the implementation methods described below in conjunction with the drawings are exemplary descriptions for explaining the technical solutions of the embodiments of the present disclosure and do not constitute a limitation on the technical solutions of the embodiments of the present disclosure.

[0025] It will be understood by those skilled in the art that, unless specifically stated, the singular forms "one", "above", and "the" used herein may also include plural forms. It should be further understood that the terms "including" and "comprising" used in the embodiments of the present disclosure refer to that the corresponding features can be implemented as the presented features, information, data, steps, operations, elements and / or components, but do not exclude the implementation as other features, information, data, steps, operations, elements, components and / or combinations thereof supported by the technical field. It should be understood that when we refer to an element as being "connected" or "coupled" to another element, the one element may be directly connected or coupled to the other element, or may refer to the one element and the other element establishing a connection relationship through an intermediate element. In addition, the "connection" or "coupling" used herein may include wireless connection or wireless coupling. The term "and / or" used herein indicates at least one of the items defined by the term, for example, "A and / or B" or "A, B" indicates implementation as "A", or implementation as "B", or implementation as "A and B".

[0026] In order to make the objectives, technical solutions and advantages of the present disclosure more clear, the embodiments of the present disclosure will be further described in detail below with reference to the accompanying drawings.

[0027] See also Figure 1 , Figure 1 A flow chart of a flight guidance generation method provided by an embodiment of the present disclosure is shown. The flight guidance generation method provided by the present disclosure is executed by a flight guidance generation model. The flight guidance generation model includes a data routing module, a data prediction module, an operation evaluation module and a voice output module. Figure 1 As shown, the flight guidance generation method provided by the embodiment of the present disclosure includes the following steps: Step S101, when the flight guidance generation model generates flight guidance information, the data prediction module obtains the pilot's historical flight data and actual flight data sent by the data routing module; Step S102, the data prediction module performs flight data prediction processing based on the acquired historical flight data to obtain predicted flight data; Step S103, the data prediction module sends the predicted flight data and the actual flight data to the operation evaluation module and the voice output module, so that the operation evaluation module and the voice output module output flight guidance information based on the predicted flight data and the actual flight data, wherein the flight guidance information is used to guide the pilot to perform flight operations according to the flight guidance information.

[0028] Optionally, the above-mentioned flight guidance generation model is a software that can predict pilot operations, give voice prompts, and guide pilot operations during flight simulator training, and can monitor the pilot's operations during simulated flight in real time, identify the pilot's operation scenarios based on the reference flight process recorded in the database, objectively evaluate the pilot's operations and guide the pilot's next operation, and has strong multi-scenario generalization capabilities and high real-time performance.

[0029] The following is a detailed description of the process of generating flight guidance information by the flight guidance generation model.

[0030] See also Figure 2 , Figure 2 A schematic diagram of the structure of a flight guidance generation model provided by an embodiment of the present disclosure. Figure 2 As shown, the flight guidance generation model includes a data routing module 201, a data prediction module 202, an operation evaluation module 203 and a voice output module 204. The flight guidance generation method in the embodiment of the present disclosure is introduced in detail below in combination with each module.

[0031] Step 201, the data routing module 201 is used to send the pilot's historical flight data and actual flight data to the data prediction module 202 when the flight guidance generation model generates flight guidance information. The pilot's historical flight data and actual flight data are both short-term data.

[0032] The data routing module implements the functions of sending, receiving and storing data. After receiving requests from other modules and confirming that the requests are legal, the module sends the requested data. The communication protocol used by the data routing module is the Transmission Control Protocol / Internet Protocol (TCP / IP for short), and the TCP protocol can be used for communication between modules.

[0033] All sent and received data are packaged into .json files for transmission. The legal types of sent and received information are: return messages, data request messages, string messages, and floating point messages. The communication steps of the data routing module are as follows: Step 1, initialize the server, bind the address and port, select the listening port, and wait for other modules to connect.

[0034] Step 2: Other modules send data requests.

[0035] Step 3: The server receives the data request and determines whether the data request is legal. If it is legal, the server returns data according to the request type. If the request is illegal, the server does not send any data.

[0036] Step 4: The server continues to listen to the port and waits for the next data request.

[0037] Step 202 : the data prediction module 202 receives the pilot's historical flight data and actual flight data sent by the data routing module 201 .

[0038] Among them, the above-mentioned data prediction module can also be called a control data prediction module, which is used to receive the pilot control data (which may include elevator, aileron, rudder and throttle data), flight attitude data (which may include airspeed, ground speed, true airspeed, angle of attack, sideslip angle, pitch angle, roll angle, altitude and its acceleration) and flight environment data (which may include wind speed) sent by the data routing module, and acquire data according to the time information corresponding to the data acquisition request to obtain historical flight data and actual flight data. In other words, both the historical flight data and the actual flight data include pilot control data, flight attitude data and flight environment data, but the historical flight data and the actual flight data belong to different time periods.

[0039] Step 203, the data prediction module 202 is used to perform flight data prediction processing based on the acquired historical flight data, obtain predicted flight data, and send the predicted flight data and actual flight data to the operation evaluation module.

[0040] Among them, the above-mentioned data prediction module can pre-process the historical flight data to generate the best matching data (i.e. the sampled flight data in the text), and forge the flight data, that is, perform fitting operations on the data to obtain the reference flight data in the following text, package the reference flight data, sampled flight data and similar flight text data in the text, generate system prompt words (i.e. at least one flight operation parameter prompt word in the text), call the large model to predict the flight status, generate predicted flight control data (i.e. the predicted flight data in the text), and send the predicted flight control data to the operation evaluation module.

[0041] Among them, the large model is the Large Language Model (LLM), which refers to a deep learning model trained with a large amount of text data. It can generate natural language text or understand the meaning of language text.

[0042] Step 204, the operation evaluation module 203 is used to receive the predicted flight data and the actual flight data, determine the flight guidance text based on the predicted flight data and the actual flight data, and send the flight guidance text to the voice output module.

[0043] Among them, the operation evaluation receives the predicted flight control data (i.e., the predicted flight data in the text), the original pilot control data, the flight attitude data, and the flight environment data, which are the actual flight data, and packages these data to generate system prompt words (i.e., the flight guidance prompt words in the text), and based on the text knowledge such as the standard process (which is at least one knowledge document of the large model), calls the large model to generate text, which is the flight guidance text in the text, and sends the text to the voice output module. Among them, the content of the text can directly include the above-mentioned text knowledge such as the standard process, or the large model can directly write the text knowledge such as the standard process into the above-mentioned flight guidance prompt words, which is not limited in the embodiment of the present disclosure.

[0044] Step 205, the voice output module 204 is used to receive the flight guidance text, and output flight guidance information based on the flight guidance text, wherein the flight guidance information is used to guide the pilot to perform flight operations according to the flight guidance information.

[0045] Among them, the voice data module can convert the received flight guidance text into voice information output, that is, convert it into flight guidance audio information and output it, and guide the pilot's next operation through the flight guidance audio information.

[0046] Optional, in practical applications, such as Figure 2 As shown, the flight guidance text and the flight guidance audio information can be output through the voice output module.

[0047] The following combination Figure 3 The data flow transmission principle of the flight guidance generation method is explained in Figure 3 , Figure 3 A schematic diagram of a data stream transmission principle provided by an embodiment of the present disclosure, wherein: Figure 3 The control data prediction 1 and control data prediction 2 belong to the data prediction module, the flight simulation program, the UDP server, the TCP server, and the interaction module between the UDP server and the TCP server all belong to the data routing module, and the control evaluation 1 and control evaluation 2 belong to the operation evaluation module. Figure 3 Only two are shown in the figure. In actual application, the required number can be determined according to the needs. This disclosure does not limit it. The specific transmission process is as follows: Pilots use flight simulators for flight training. The flight simulation program on the flight simulator generates corresponding flight data (such as historical flight data and actual flight data). Through the User Datagram Protocol (UDP), the flight data in the form of C++ structure byte stream (or other forms of byte stream, which is not limited in this disclosure) is sent to the UDP server, and the UDP server saves the flight data locally. The UDP server is also used to perform data parsing, channel splitting and data storage processing on the flight data according to predefined data packets. The following takes the manipulation of data prediction host 1, manipulation of data prediction host 2, operation evaluation host 1 and operation evaluation host 2 as examples for explanation. Assuming that the TCP server needs to perform prediction and evaluation on the manipulation of data prediction host 1, manipulation of data prediction host 2, manipulation evaluation host 1 and manipulation evaluation host 2, and needs 0~20s of flight data for prediction, the TCP server will generate a json byte stream according to the predefined message (Message) through logical processing, and send the generated message (Message) to the UDP server through task scheduling. The message (Message) is a legal request sent to the UDP server to request 0~20s of flight data. After receiving the legal request, the UDP server will return 0~ 20s flight data, and send 0~20s flight data to the control data prediction data 1 host and the control prediction data 2 host through task scheduling. After the control data prediction data 1 host and the control prediction data 2 host receive the 0~20s flight data, they make predictions based on the big model method to obtain predicted flight data, and then send the predicted flight data back to the UDP server. The UDP server receives the data and tags it: prediction of 0~20s data (that is, predicted flight data corresponding to 0~20s flight data). After that, the TCP server sends a request to the UDP server through logical processing. The control evaluation 1 host and the control evaluation 2 host need 0~20s flight data and predicted flight data. At this time, the UDP server sends 0~20s flight data and predicted flight data to the control evaluation 1 host.

[0048] Among them, logic processing is to process flight data according to requirements, such as data slicing, data preservation, etc. Task scheduling and logic processing are responsible for sending data transmission instructions to the TCP server according to time, such as setting data to be sent to the control data prediction host 1 every five seconds, and then sending data to the control evaluation host 1 after receiving the return data from the control data prediction host 1.

[0049] Through the disclosed embodiments, by generating flight guidance information, it is possible to predict pilot operations, give voice prompts, and guide the pilot to perform flight operations during flight simulator training. By recording the pilot's actual flight data in real time, it is possible to monitor the pilot's operations during simulated flight in real time, objectively evaluate the pilot's operations and guide the pilot's next operations. It has strong multi-scenario generalization capabilities and high real-time performance, which improves the pilot's flight training efficiency.

[0050] In an optional embodiment, the flight data prediction processing is performed based on the acquired historical flight data to obtain the predicted flight data, including: Retrieving each flight data stored in a database, wherein the database contains at least two flight data; Sampling the above historical flight data to obtain sampled historical flight data, which is recorded as sampled flight data; The above-mentioned flight data and the above-mentioned sampled flight data are preprocessed, and prediction is performed based on the preprocessed flight data and the preprocessed sampled flight data to obtain the above-mentioned predicted flight data.

[0051] Optionally, the following describes how the predicted flight data is obtained: The flight data stored in the database are called, wherein the database contains at least two complete flight data, which represent the correct operation of the pilot in various flight situations, have a high reference value, and can be used or trained as a sample set. In practical applications, when calling the flight data stored in the database, the required flight data can be called according to actual needs, such as setting the flight data within a preset time period to be called, or setting the flight data stored in the database to be called, which is not limited in the embodiments of the present disclosure.

[0052] The acquired historical flight data is then sampled and processed. Optionally, the original data can be downsampled by interpolation to improve data processing efficiency while ensuring data accuracy to obtain sampled historical flight data. The sampled historical flight data can be recorded as sampled flight data. Each flight data and the sampled flight data are then preprocessed to obtain preprocessed flight data and preprocessed sampled flight data. Prediction is then performed based on the preprocessed flight data and the preprocessed sampled flight data to obtain predicted flight data.

[0053] In an optional embodiment, the above-mentioned similar flight data and the above-mentioned sampled flight data are standardized to obtain the standardized similar flight data and the standardized sampled flight data, including: Determine the mean and standard deviation of each flight operation parameter included in each of the above flight data; Based on the mean and standard deviation of the above-mentioned flight operation parameters, the above-mentioned similar flight data and the above-mentioned sampled flight data are standardized to obtain standardized similar flight data and standardized sampled flight data.

[0054] In order to clearly illustrate the process of preprocessing each flight data and the sampled flight data, in an optional embodiment, the above-mentioned preprocessing each flight data and the sampled flight data, and predicting based on the preprocessed flight data and the preprocessed sampled flight data to obtain the above-mentioned predicted flight data, includes: Slicing the flight data to obtain at least one set of data with the same dimension as the sampled flight data, which is recorded as same-dimensional data; Determine similar flight data similar to the sampled flight data in the same-dimensional data; Standardizing the similar flight data and the sampled flight data to obtain standardized similar flight data and standardized sampled flight data; Performing linear scaling and character encoding on the standardized similar flight data and the standardized sampled flight data to obtain corresponding similar flight text data and sampled flight text data, wherein the similar flight text data is preprocessed flight data, and the sampled flight text data is preprocessed sampled flight data; Classify the similar flight text data to obtain corresponding horizontal flight data and vertical flight data; Perform a fitting operation based on the lateral flight data and the longitudinal flight data to obtain reference flight data; Creating at least one flight operation parameter prompt word according to the similar flight text data, the reference flight data and the sampled flight text data, wherein each of the flight operation parameter prompt words is a standard example of a corresponding flight operation parameter, and each of the flight operation prompt words contains a prediction task of a corresponding flight operation parameter; At least one large model is called simultaneously by using multiple processes. For each called large model, flight data prediction is performed based on the flight operation parameter prompt words corresponding to the large model to obtain the above-mentioned predicted flight data, wherein the above-mentioned data prediction module includes at least one of the above-mentioned large models.

[0055] Optionally, the following is a data preprocessing method provided by an embodiment of the present disclosure, wherein the data preprocessing is mainly performed by a data prediction module, wherein the data prediction module uses a large model for prediction, and the data prediction module includes two submodules: data preprocessing and large model data prediction. The data prediction module first sends a data acquisition request to the data routing module, and after the request is successful, receives the pilot's control data fragments (elevator, aileron, rudder and throttle data), flight attitude data (airspeed, ground speed, true airspeed, angle of attack, sideslip angle, pitch angle, roll angle, altitude and its acceleration) and flight environment data (wind speed). The above pilot control data fragments are collectively referred to as original data (i.e., historical flight data), and the original data is preprocessed according to the complete flight data stored in the database. Among them, the data generated by the pilot during the flight are all original data, including historical flight data and actual flight data, but the historical flight data and the actual flight data belong to different time periods. Among them, the historical flight data is used to predict the data of the next time period, and the actual flight data is used to compare and evaluate with the predicted data.

[0056] The specific data preprocessing steps are as follows: Step 301, calling the complete flight data in the database, calculating the mean and standard deviation of each parameter corresponding to the complete flight data, so as to standardize the data in subsequent steps.

[0057] Among them, the complete flight data (i.e. each flight data) includes the pilot control data (elevator, aileron, rudder and throttle data), flight attitude data (airspeed, ground speed, true airspeed, angle of attack, sideslip angle, pitch angle, roll angle, altitude and its acceleration) and flight environment data (wind speed), and then calculate the mean and standard deviation of each parameter. Taking the elevator as an example, assuming that the elevator value at 200 moments is recorded in a flight, there are a total of 50 complete flight records. The mean and standard deviation of the elevator are the mean and standard deviation of the elevator values ​​at about 10,000 moments. Among them, the mean represents the average position, and the standard deviation can intuitively reflect the fluctuation. The larger the standard deviation, the greater the data fluctuation.

[0058] Step 302, downsample the original data (here, historical flight data) by interpolation method to improve data processing efficiency while ensuring data accuracy. Slice the data in the database, that is, the flight data (here, the flight data contains multiple complete flight data), and generate multiple groups of data with the same dimensions as the original data (that is, the same time interval as the original data, and the same total duration), which are recorded as same-dimensional data. Calculate the Euclidean distance between all the same-dimensional data and the original data, and select the 10 groups of data with the smallest Euclidean distance to the original data among all the same-dimensional data as similar data, which are the above-mentioned similar flight data.

[0059] Step 303, the similar data and the sampled original data (i.e., the sampled flight data) are standardized using the z-score method according to the mean and standard deviation of all the data in the database to obtain the standardized similar data (i.e., the standardized similar flight data) and the standardized original data (i.e., the standardized sampled flight data). The z-score, also called the standard score, is the process of dividing the difference between a number and the mean by the standard deviation. In statistics, the standard score is the symbol number of the standard deviation of the value of an observation or data point above the mean of the observed value or measurement.

[0060] Step 304, linear scaling and character encoding are performed on the standardized similar data and the standardized original data, and each parameter value is mapped to the interval of a~z, where a is the minimum value and z is the maximum value, and the corresponding similar flight text data and sampled flight text data are obtained, wherein the similar flight text data is the preprocessed flight data, and the sampled flight text data is the preprocessed sampled flight data. At this point, pure mathematical information is converted into text information, which is conducive to the large model to analyze the information.

[0061] Step 305, classify similar flight text data into two categories: horizontal similar data (i.e. horizontal flight data) and vertical similar data (i.e. vertical flight data). The horizontal data are parameters related to the aircraft's flight roll and yaw, including rudder, aileron, sideslip angle, roll angle, yaw angle and other data; the vertical data are data related to the aircraft's flight pitch, including elevator, angle of attack, throttle, horizontal stabilizer, pitch angle and other data. Classification and analysis of data can effectively prevent the large model from over-interpreting information, reduce the degree of information coupling, and improve the evaluation accuracy provided by the large model.

[0062] Step 306, the forged manipulation prediction data is fitted based on the lateral similarity data and the longitudinal similarity data to generate forged reference data (i.e., reference flight data), including the aircraft's elevator, aileron, rudder, and throttle data. The forged manipulation prediction data is used as reference data, waiting for the manipulation prediction module to call the reference flight data for analysis.

[0063] Through the above steps 301 to 306, the data preprocessing submodule processes the original data, classifies the original data and generates forged reference data (i.e., reference flight data), providing multiple reference standards for the large model data prediction submodule.

[0064] The large model data prediction submodule (also referred to as the large model manipulation prediction submodule) can create at least one system prompt word (i.e., flight operation parameter prompt word) based on the packaged data, i.e., the packaged similar flight text data, the reference flight data (i.e., forged reference data), and the sampled flight text data. Each system prompt word fully gives the prediction task content, application scenario, corresponding data meaning and unit, and complete example required for the large model. Then, the large model is called, and then, the large model performs manipulation prediction according to each system prompt word (i.e., each flight operation parameter prompt word), and generates lateral manipulation prediction data (including wing and rudder) and longitudinal manipulation prediction data (including left and right throttle and elevator) of a predefined time length. For example, the predefined time length can be set to 5 seconds, which is not limited here. The generated lateral manipulation prediction data and longitudinal manipulation prediction data are manipulation prediction data, corresponding to the predicted flight data in the text, wherein the predicted flight data contains prediction data corresponding to at least one flight operation parameter, and each flight operation parameter is predicted by a corresponding large model. Before calling the large model each time, clear the historical information to ensure that the large model prediction is not affected by the previous prediction result. Since calling a large model takes a certain amount of time, multiple large models are selected to be called in parallel, and multiple data predictions are performed at the same time, which greatly improves real-time performance.

[0065] Among them, the selected large models may include: GPT-4o, GPT-4o-mini, llama3.1-7B, llama3.1-70B, etc., which are not limited in the embodiments of the present disclosure. Among them, GPT-4 is the fourth-generation general pre-training model released by OpenAI, which has powerful generation and understanding capabilities and is suitable for a variety of complex scenarios. GPT-4O (GPT-4 Optimized) is an optimized version of GPT-4, focusing on efficiency and performance, and is suitable for applications that need to balance performance and resources. GPT-4O-mini is a lighter version, designed to provide efficient solutions for low-computing power devices and scenarios. llama3.1-7B and llama3.1-70B are the 7B and 70B versions included in the Llama3 series of large language models.

[0066] Optionally, the steps for the large model data prediction submodule to perform data prediction are as follows: Step 401, receiving the forged reference data (ie, reference flight data) and similar flight text data generated by the data preprocessing submodule, and packaging them with the sampled original data (ie, sampled flight text data) as the content in the system prompt words.

[0067] Step 402, generate multiple system prompts, including the prediction task content, application scenarios, meanings and units of each data, and complete examples required for the large model. Each system prompt corresponds to a single data that needs to be predicted: such as elevator, rudder, aileron, left and right throttle, etc. This facilitates parallel calling of large models.

[0068] Step 403, using multiple processes to call multiple large models at the same time, and predicting the control data based on multiple system prompt words, and returning the control prediction data (ie, predicted flight data).

[0069] The above steps 301 to 306, and steps 401 to 403, are different from the traditional method of using expert knowledge as the operation evaluation standard. In the embodiment of the present disclosure, a large language model (large model) is introduced as the core of flight operation evaluation. The large model has powerful natural language processing performance and can give a true, accurate and real-time flight operation evaluation based on the pilot's historical flight data. In addition, in the embodiment of the present disclosure, the flight data is downsampled and classified to reduce the data complexity, and then the large model is called to analyze the downsampled data for prediction and evaluation, thereby improving the accuracy of the large model evaluation data.

[0070] The following combination Figure 4 Let’s take an example to illustrate the process of data preprocessing. Figure 4 A schematic diagram of a data preprocessing process provided by an embodiment of the present disclosure is shown in FIG. Figure 4 As shown in the figure, the process of data preprocessing is as follows: At least one complete flight data in the database is called. Each complete flight data is stored in the database in the form of a comma-separated value (CSV) file, where the CSV is sometimes also called a character-separated value because the separator character may not be a comma. The CSV file stores tabular data (numbers and text) in plain text form. Then, these complete flight data are standardized and split to obtain multiple sets of data in the flight database [5s (10 data points), m parameters], where m is the number of flight parameters, n is the number of predicted parameters, 5s means predicting 5s, and 10 data points means predicting 10 data points (T=0.5).

[0071] Obtain historical data (i.e., historical flight data). For example, the historical data may be [5s (10 data points), m parameters]. Then, downsampling and standardization operations are performed through linear interpolation to obtain a set of historical data corresponding to standardization [5s (10 data points), m parameters] (corresponding to the sampled flight text data). Then, the similarity between the set of historical data [5s (10 data points), m parameters] and multiple sets of data [5s (10 data points), m parameters] in the flight database is calculated to obtain similar data (corresponding to similar flight data). Then, the set of historical data [5s (10 data points), m parameters] is packaged with the obtained similar data, and linear scaling and character encoding are performed to obtain actual reference operation strings and reference operation strings (corresponding to reference flight data). Based on the obtained strings, prompt words are generated (corresponding to flight operation parameter prompt words), and then prediction is performed through a large model to obtain a predicted operation string. Then, string inverse encoding, inverse linear scaling, and inverse standardization are performed based on the output to obtain a final prediction series [5s (10 data points), n parameters), corresponding to the predicted flight data.

[0072] Through the embodiments of the present disclosure, the flight data of the pilot during the flight can be received in real time for evaluation guidance, which has good real-time performance; in addition, the present disclosure makes use of a large amount of reference flight data stored in the database when providing operation evaluation guidance, so that the evaluation result is more objective and comprehensive; and the present disclosure uses a large model for predictive evaluation, which realizes automatic prediction, does not require manpower, and saves costs.

[0073] In an optional embodiment, the operation evaluation module and the voice output module output the flight guidance information based on the predicted flight data and the actual flight data in accordance with the following steps: The operation evaluation module receives the predicted flight data sent by the data prediction module; The operation evaluation module creates a flight guidance prompt based on the predicted flight data and the actual flight data, wherein the flight guidance prompt includes relevant information of the flight operation and a flight evaluation standard; The operation evaluation module calls the big model, and performs operation evaluation on the actual flight data based on the flight guidance prompt words to obtain a flight guidance text, and sends the flight guidance text to the voice output module, wherein the big model is a module in the data prediction module; The voice output module receives the flight guidance text; The voice output module calls the multimodal large model to convert the flight guidance text into flight guidance audio information, wherein the multimodal large model is used to convert text information into audio information, and the flight guidance information includes the flight guidance audio information; The voice output module receives the flight guidance audio information returned by the multi-modality, and outputs the flight guidance audio information.

[0074] Optionally, the foregoing describes a detailed process of how to obtain the predicted flight data. The following describes a detailed process of how the operation evaluation module and the voice output module output the flight guidance information based on the predicted flight data and the actual flight data. The specific steps are as follows: The operation evaluation module and the voice output module receive the operation prediction data (i.e., predicted flight data) generated by the large model operation data prediction module (i.e., data prediction module), package them together with the original data (here, the original data is the actual flight data), create system prompts (i.e., flight guidance prompts), and input aircraft flight control related knowledge, flight data meaning, and flight operation terms into the large model through the system prompts, providing a flight knowledge base for the large model. The system prompts also need to command the large model to analyze the pilot's operation and generate text-based flight guidance (i.e., flight guidance text).

[0075] The voice output module receives text-based flight guidance and calls an external multimodal large model to convert the text into audio output, which is the flight guidance audio information.

[0076] The specific steps for the operation evaluation module and the voice output module to output the flight guidance audio information are as follows: Step 501, accepting the maneuver prediction data (i.e., predicted flight data) generated by the large model maneuver prediction module and the data preprocessing module, and packaging it with the original data (here, the original data is the actual flight data) as the content in the system prompt words.

[0077] Step 502, generate system prompt words (i.e., flight guidance prompt words), which include an introduction to flight terms, a detailed explanation of flight parameters, and assessment task requirements. The flight terms can be in different languages, such as Chinese, English, etc., according to actual needs.

[0078] Step 503, calling the large model and performing operation evaluation guidance based on the generated system prompt words, and returning text-type flight guidance (ie, flight guidance text).

[0079] Step 504 : The voice output module receives text-based flight instructions.

[0080] Step 505, calling the multimodal large model to convert text to audio, and inputting text-type flight guidance into the multimodal large model.

[0081] Among them, the above-mentioned multimodal large model refers to a model that combines multimodal information such as text, images, videos, and audio for training, which requires Internet resources to be called up, while the large model used in the above-mentioned data prediction module and operation evaluation module is deployed locally and does not require Internet access.

[0082] Step 506: The multimodal large model returns the audio-type flight guidance (i.e., flight guidance audio information) to the voice output module as the final output.

[0083] For the above steps 505 and 506, in actual applications, they can also be implemented in the following manner: the voice output module receives the text-based flight guidance (i.e., the flight guidance text); then the voice output module converts the text-based flight guidance into flight guidance audio information, and outputs the flight guidance audio information.

[0084] Through the embodiments of the present disclosure, flight guidance audio information can be provided to the pilot, so that the pilot can simply and directly follow the guidance of the flight guidance audio information to perform the next flight operation, thereby improving the training efficiency of the pilot and improving the user experience.

[0085] In order to be able to more intuitively compare various data, such as historical flight data, actual flight data, predicted flight data, and flight guidance information, the disclosed embodiment can also display various data through a visual interface. In an optional embodiment, the above method also includes: At least one of the above historical flight data, the above predicted flight data or the flight guidance information is displayed on the flight operation display interface.

[0086] Optionally, through the above steps 301 to 306, and steps 401 to 403, forged reference data (i.e., reference flight data) and predicted flight data output by the large model prediction can be obtained. The forged reference data and the predicted flight data can be output in the form of character encoding. Optionally, in actual applications, these two types of data can be displayed in a visual interface as needed. See Figure 4 , Figure 5 A character encoding form of reference flight data and predicted flight data provided by an embodiment of the present disclosure, such as Figure 5 As shown, Figure 5 The fake historical questions and answers shown in are the character encoding form of the reference flight data. Figure 5 The actual questions and answers shown in are the character encoding form of the predicted flight data.

[0087] Optionally, in order to intuitively understand the complete flight data, the pilot's historical flight data, and the predicted flight data, the embodiment of the present disclosure provides a visualization interface diagram of the operation data, such as Figure 6 As shown, Figure 6 Taking the elevator, aileron, rudder, left throttle (traL) and right throttle (traR) as examples, the complete flight data and the predicted flight data output by the large model are displayed on the visualization interface. The curve corresponding to the dotted part is the complete flight data, the first solid curve is the original data (i.e., historical flight data), and the second solid curve is the predicted output data of the large model (i.e., predicted flight data).

[0088] Optionally, in order to understand all the data of the pilot from the beginning of the flight to the end of the flight, visualization can be performed through the user interface (User Interface, referred to as UI), wherein the UI interface implements data visualization based on Qt (Qt is a cross-platform C++ graphical user interface application framework). The interface has the function of selecting test data. After selecting the test data, the interface displays the complete curve of each parameter of the test data changing with time, wherein the test data can be all the data of the pilot from the beginning of the flight to the end of the flight, including elevator, aileron, rudder, throttle, airspeed, ground speed, true airspeed, angle of attack, sideslip angle, pitch angle, roll angle, altitude, angular acceleration, wind speed, etc., which can be moved over time, and the flight data can be displayed according to time, and the corresponding flight guidance information can be displayed according to time.

[0089] Optionally, the flight guidance information can also be displayed to the pilot through a visual interface, so that the pilot can intuitively see the flight guidance information, can more clearly understand the next flight operation, and adjust the flight operation in time according to the flight guidance information, greatly improving the pilot's training efficiency.

[0090] Through the embodiments of the present disclosure, changes in various data and differences in changes between various data can be intuitively understood through a visual interface, which facilitates intuitive analysis of the pilot's flight operations and improves the efficiency of data analysis.

[0091] The present disclosure provides a flight guidance generating device 10, wherein the flight guidance generating device 10 comprises a flight guidance generating model, wherein the flight guidance generating model comprises a data routing module, a data prediction module, an operation evaluation module and a voice output module, such as Figure 7 As shown, the above device includes an acquisition module 101, a first processing module 102 and a second processing module 103, wherein: An acquisition module 101 is used to acquire the pilot's historical flight data and actual flight data sent by the data routing module of the flight guidance generation model through the data prediction module of the flight guidance generation model when the flight guidance generation model generates flight guidance information; A first processing module 102, configured for the data prediction module to perform flight data prediction processing based on the acquired historical flight data to obtain predicted flight data; The second processing module 103 is used for the data prediction module to send the predicted flight data and the actual flight data to the operation evaluation module and the voice output module, so that the operation evaluation module and the voice output module output flight guidance information based on the predicted flight data and the actual flight data, wherein the flight guidance information is used to guide the pilot to perform flight operations according to the flight guidance information.

[0092] In an optional embodiment, the first processing module 102 is specifically used to: call each flight data stored in a database, wherein the database contains at least two flight data; perform sampling processing on the historical flight data to obtain sampled historical flight data, recorded as sampled flight data; pre-process each flight data and the sampled flight data, and perform prediction based on the pre-processed flight data and the pre-processed sampled flight data to obtain the predicted flight data.

[0093] In an optional embodiment, the first processing module 102 is specifically used for: performing a slicing operation on each of the flight data to obtain at least one group of data with the same dimension as the sampled flight data, recorded as same-dimensional data; determining similar flight data similar to the sampled flight data in the same-dimensional data; standardizing the similar flight data and the sampled flight data to obtain standardized similar flight data and standardized sampled flight data; performing linear scaling and character encoding on the standardized similar flight data and the standardized sampled flight data to obtain corresponding similar flight text data and sampled flight text data, wherein the similar flight text data is preprocessed flight data, and the sampled flight text data is preprocessed sampled flight data; The flight text data is classified to obtain corresponding lateral flight data and longitudinal flight data; a fitting operation is performed based on the lateral flight data and the longitudinal flight data to obtain reference flight data; at least one flight operation parameter prompt word is created according to the similar flight text data, the reference flight data and the sampled flight text data, wherein each of the flight operation parameter prompt words is a standard example of a corresponding flight operation parameter, and each of the flight operation prompt words contains a prediction task of a corresponding flight operation parameter; at least one large model is called simultaneously by multiple processes, and for each called large model, flight data prediction is performed based on the flight operation parameter prompt word corresponding to the large model to obtain the predicted flight data, wherein the data prediction module includes at least one of the large models.

[0094] In an optional embodiment, the second processing module 103 is specifically used to perform the following steps: the operation evaluation module receives the predicted flight data sent by the data prediction module; the operation evaluation module creates a flight guidance prompt based on the predicted flight data and the actual flight data, wherein the flight guidance prompt includes relevant information of the flight operation and the flight evaluation standard; the operation evaluation module calls the big model, and performs an operation evaluation on the actual flight data based on the flight guidance prompt to obtain a flight guidance text, and sends the flight guidance text to the voice output module, wherein the big model is a module in the data prediction module; the voice output module receives the flight guidance text; the voice output module calls the multimodal big model to convert the flight guidance text into flight guidance audio information, wherein the multimodal big model is used to convert text information into audio information, and the flight guidance information includes the flight guidance audio information; the voice output module receives the flight guidance audio information returned by the multimodal, and outputs the flight guidance audio information.

[0095] In an optional embodiment, the first processing module 102 is used to: determine the mean and standard deviation of each flight operation parameter contained in the flight data; based on the mean and standard deviation of each flight operation parameter, standardize the similar flight data and the sampled flight data to obtain standardized similar flight data and standardized sampled flight data.

[0096] In an optional embodiment, the above-mentioned device also includes: a display module, used to display at least one of the above-mentioned historical flight data, the above-mentioned predicted flight data or flight guidance information on the flight operation display interface.

[0097] Through the disclosed embodiments, by generating flight guidance information, it is possible to predict pilot operations, give voice prompts, and guide the pilot to perform flight operations during flight simulator training. By recording the pilot's actual flight data in real time, it is possible to monitor the pilot's operations during simulated flight in real time, objectively evaluate the pilot's operations and guide the pilot's next operations. It has strong multi-scenario generalization capabilities and high real-time performance, which improves the pilot's flight training efficiency.

[0098] The device of the embodiment of the present disclosure can execute the method provided by the embodiment of the present disclosure, and its implementation principle is similar and has corresponding technical effects. The actions performed by each module in the device of each embodiment of the present disclosure correspond to the steps in the method of each embodiment of the present disclosure. For the detailed functional description of each module of the device, please refer to the description in the corresponding method shown in the previous text, which will not be repeated here.

[0099] An embodiment of the present disclosure provides an electronic device (computer device / equipment / system), including a memory, a processor, and a computer program stored in the memory, and the processor executes the above computer program to implement the steps of the method provided in any optional embodiment of the present disclosure.

[0100] In an alternative embodiment, an electronic device is provided, such as Figure 8 As shown, Figure 8 The electronic device 1000 shown includes: a processor 1001 and a memory 1003. The processor 1001 and the memory 1003 are connected, such as through a bus 1002. Optionally, the electronic device 1000 may also include a transceiver 1004, which may be used for data interaction between the electronic device and other electronic devices, such as data transmission and / or data reception. It should be noted that in actual applications, the transceiver 1004 is not limited to one, and the structure of the electronic device 1000 does not constitute a limitation on the embodiments of the present disclosure.

[0101] Processor 1001 may be a CPU (Central Processing Unit), a general-purpose processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array) or other programmable logic devices, transistor logic devices, hardware components or any combination thereof. It may implement or execute various exemplary logic blocks, modules and circuits described in conjunction with the disclosure of the present invention. Processor 1001 may also be a combination that implements computing functions, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, etc.

[0102] The bus 1002 may include a path to transmit information between the above components. The bus 1002 may be a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus. The bus 1002 may be divided into an address bus, a data bus, a control bus, etc. For ease of representation, Figure 8 Only one thick line is used in the diagram, but this does not mean that there is only one bus or only one type of bus.

[0103] The memory 1003 may be a ROM (Read Only Memory) or other types of static storage devices that can store static information and instructions, a RAM (Random Access Memory) or other types of dynamic storage devices that can store information and instructions, or an EEPROM (Electrically Erasable Programmable Read Only Memory), a CD-ROM (Compact Disc Read Only Memory) or other optical disk storage, optical disk storage (including compressed optical disk, laser disk, optical disk, digital versatile disk, Blu-ray disk, etc.), magnetic disk storage medium, other magnetic storage devices, or any other medium that can be used to carry or store computer programs and can be read by a computer, without limitation herein.

[0104] The memory 1003 is used to store the computer program for executing the embodiment of the present disclosure, and the execution is controlled by the processor 1001. The processor 1001 is used to execute the computer program stored in the memory 1003 to implement the steps shown in the above method embodiment.

[0105] An embodiment of the present disclosure provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps and corresponding contents of the aforementioned method embodiment can be implemented.

[0106] The embodiments of the present disclosure also provide a computer program product, including a computer program, which can implement the steps and corresponding contents of the aforementioned method embodiments when executed by a processor.

[0107] It should be understood that, although the flowchart of the embodiment of the present disclosure indicates each operation step by arrows, the implementation order of these steps is not limited to the order indicated by the arrows. Unless clearly stated herein, in some implementation scenarios of the embodiment of the present disclosure, the implementation steps in each flowchart can be executed in other orders according to demand. In addition, some or all of the steps in each flowchart may include multiple sub-steps or multiple stages based on the actual implementation scenario. Some or all of these sub-steps or stages can be executed at the same time, and each sub-step or stage in these sub-steps or stages can also be executed at different times. In scenarios with different execution times, the execution order of these sub-steps or stages can be flexibly configured according to demand, and the embodiment of the present disclosure does not limit this.

[0108] The above are only optional implementation methods for some implementation scenarios of the present disclosure. It should be pointed out that for ordinary technicians in this technical field, without departing from the technical concept of the scheme of the present disclosure, other similar implementation methods based on the technical ideas of the present disclosure are also within the protection scope of the embodiments of the present disclosure.

Claims

1. A flight guidance generation method, characterized in that: The method is performed by a flight guidance generation model, the flight guidance generation model includes a data routing module, a data prediction module, an operation evaluation module and a voice output module, and the method includes: In the case where the flight guidance generation model executes to generate flight guidance information, the data prediction module obtains the pilot's historical flight data and actual flight data sent by the data routing module; The data prediction module performs flight data prediction processing based on the acquired historical flight data to obtain predicted flight data; The data prediction module sends the predicted flight data and the actual flight data to the operation evaluation module and the voice output module, so that the operation evaluation module and the voice output module output flight guidance information based on the predicted flight data and the actual flight data, wherein the flight guidance information is used to guide the pilot to perform flight operations according to the flight guidance information.

2. The method according to claim 1, characterized in that The performing flight data prediction processing based on the acquired historical flight data to obtain predicted flight data includes: Retrieving each flight data stored in a database, wherein the database contains at least two flight data; Sampling the historical flight data to obtain sampled historical flight data, which is recorded as sampled flight data; The flight data and the sampled flight data are preprocessed, and prediction is performed based on the preprocessed flight data and the preprocessed sampled flight data to obtain the predicted flight data.

3. The method according to claim 2, characterized in that The preprocessing of each flight data and the sampled flight data, and prediction based on the preprocessed flight data and the preprocessed sampled flight data to obtain the predicted flight data, includes: Performing slicing operation on each of the flight data to obtain at least one set of data with the same dimension as the sampled flight data, recorded as same-dimensional data; Determining similar flight data in the same-dimensional data that is similar to the sampled flight data; Standardizing the similar flight data and the sampled flight data to obtain standardized similar flight data and standardized sampled flight data; Performing linear scaling and character encoding on the standardized similar flight data and the standardized sampled flight data to obtain corresponding similar flight text data and sampled flight text data, wherein the similar flight text data is preprocessed flight data, and the sampled flight text data is preprocessed sampled flight data; Classifying the similar flight text data to obtain corresponding horizontal flight data and vertical flight data; Perform a fitting operation based on the lateral flight data and the longitudinal flight data to obtain reference flight data; Creating at least one flight operation parameter prompt word according to the similar flight text data, the reference flight data and the sampled flight text data, wherein each of the flight operation parameter prompt words is a standard example of a corresponding flight operation parameter, and each of the flight operation prompt words contains a prediction task of a corresponding flight operation parameter; At least one large model is called simultaneously by using multiple processes. For each called large model, flight data prediction is performed based on the flight operation parameter prompt words corresponding to the large model to obtain the predicted flight data, wherein the data prediction module includes at least one of the large models.

4. The method according to claim 3, characterized in that The operation evaluation module and the voice output module output the flight guidance information based on the predicted flight data and the actual flight data according to the following steps: The operation evaluation module receives the predicted flight data sent by the data prediction module; The operation evaluation module creates a flight guidance prompt based on the predicted flight data and the actual flight data, wherein the flight guidance prompt includes relevant information of the flight operation and a flight evaluation standard; The operation evaluation module calls the large model, and performs operation evaluation on the actual flight data based on the flight guidance prompt words to obtain a flight guidance text, and sends the flight guidance text to the voice output module, wherein the large model is a module in the data prediction module; The voice output module receives the flight guidance text; The speech output module calls the multimodal large model to convert the flight guidance text into flight guidance audio information, wherein the multimodal large model is used to convert text information into audio information, and the flight guidance information includes the flight guidance audio information; The voice output module receives the flight guidance audio information returned by the multi-modality, and outputs the flight guidance audio information.

5. The method according to claim 3, characterized in that: The step of normalizing the similar flight data and the sampled flight data to obtain the normalized similar flight data and the normalized sampled flight data includes: Determining the mean and standard deviation of each flight operation parameter included in each flight data; Based on the mean and standard deviation of each flight operation parameter, the similar flight data and the sampled flight data are standardized to obtain standardized similar flight data and standardized sampled flight data.

6. The method according to claim 1, characterized in that The method further comprises: At least one of the historical flight data, the predicted flight data or the flight guidance information is displayed on the flight operation display interface.

7. A flight guidance generating device, characterized in that: The flight guidance generation device comprises a flight guidance generation model, the flight guidance generation model comprises a data routing module, a data prediction module, an operation evaluation module and a voice output module, and the device comprises: an acquisition module, configured to acquire, through a data prediction module of the flight guidance generation model, historical flight data and actual flight data of the pilot sent by a data routing module of the flight guidance generation model when the flight guidance generation model generates flight guidance information; A first processing module, configured for the data prediction module to perform flight data prediction processing based on the acquired historical flight data to obtain predicted flight data; A second processing module is used for the data prediction module to send the predicted flight data and the actual flight data to the operation evaluation module and the voice output module, so that the operation evaluation module and the voice output module output flight guidance information based on the predicted flight data and the actual flight data, wherein the flight guidance information is used to guide the pilot to perform flight operations according to the flight guidance information.

8. An electronic device comprising a memory, a processor and a computer program stored in the memory, characterized in that: The processor executes the computer program to implement the steps of the method according to any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images