An aviation turning prediction method and device based on a logistic regression model

Abstract

The application provides an aviation turning prediction method and device based on a logistic regression model, and the method comprises the following steps: adopting a loss function and a bottom regression model to construct an aviation turning prediction initial model; based on known quaternion data and corresponding heading angles, the aviation turning prediction initial model is iteratively optimized by using the loss function to obtain an aviation turning prediction model based on logistic regression; quaternion data of an aircraft during flight operation is collected in real time, and binary quaternion data is obtained by binary processing of the quaternion data; the binary quaternion data is input into the aviation turning prediction model based on logistic regression, and the output result is used to predict the turning of the aircraft. The above method and device can solve the technical problems of high cost and excessive trajectory compensation in the prior art, which leads to abnormal heading angle values and affects the prediction accuracy.

Description

TECHNICAL FIELD

[0001] The application belongs to the field of aircraft data processing, and relates to an aviation turning prediction technology, in particular to an aviation turning prediction method and device based on a logistic regression model. BACKGROUND

[0002] In aircraft data processing, flight trajectory or direction may be predicted through flight data. Traditional prediction methods are to solve flight data through a gyroscope or an inverse trigonometric function. These two methods can solve problems such as static drift and uncontrollable factor interference in most heading angle operations.

[0003] The gyroscope method is to position the relative position through a GPS system, can solve the heading angle data in real time to predict the aviation turning, this method can achieve high precision, but has the problems of high cost and unstable signal; the inverse trigonometric function method is to design a heading trajectory identification method based on a magnetic field, to construct a model through a magnetic heading angle and a magnetic declination angle to calculate a heading angle value, and to obtain a heading position to repair a heading trajectory to predict the aviation turning, this method has the problems of high cost, needs additional computing power to process data, and has the problem of abnormal heading angle value caused by trajectory compensation when repairing the trajectory.

[0004] Obviously, the above two methods cannot meet the needs of all flight tasks, and a low-cost and high-accuracy aviation turning prediction method needs to be designed. SUMMARY

[0005] In order to solve the technical problems of high cost and excessive trajectory compensation leading to abnormal heading angle value when predicting aviation turning, and affecting the prediction accuracy in the prior art, the application discloses an aviation turning prediction method and device based on a logistic regression model.

[0006] The technical scheme for achieving the object of the application is as follows:

[0007] The embodiment of the application provides an aviation turning prediction method based on a logistic regression model, which comprises the following steps:

[0008] S1, using a loss function and a bottom regression model to construct an aviation turning prediction initial model;

[0009] S2, based on known quaternion data of different flight states of an aircraft and corresponding heading angles, using the loss function to iteratively optimize the aviation turning prediction initial model, to obtain an aviation turning prediction model based on a logistic regression;

[0010] S3, collecting quaternion data of the aircraft when performing flight operations in real time, and performing binary processing on the quaternion data to obtain binary quaternion data;

[0011] S4, inputting the binary quaternion data into the aviation steering prediction model based on the logistic regression, and predicting the aircraft steering according to the output result, the aircraft steering including any one of upward turning, left turning, downward turning, and right turning.

[0012] Further, in the step S1, the loss function includes any one of gradient descent method, steepest descent method, Adam algorithm, and AdaGrad algorithm.

[0013] Further, in the step S2, the known quaternion data and the corresponding heading angle of the aircraft in different flight states are used to iteratively optimize the aviation steering prediction initial model by using the loss function, to obtain the aviation steering prediction model based on the logistic regression, including:

[0014] S21, obtaining a set of quaternion data and a set of target data of the corresponding heading angle of the aircraft in different flight modes;

[0015] S22, performing binary processing on the set of quaternion data and the set of target data to obtain a set of binary quaternion data and a set of binary target data;

[0016] S23, dividing the set of binary quaternion data and the set of binary target data into a training set and a test set according to the same proportion;

[0017] S24, training and testing the training set and the test set, and iteratively optimizing the aviation steering prediction initial model by using the loss function.

[0018] Further, the step S2 further includes:

[0019] The set of quaternion data and the set of target data of the heading angle in step S21 are preprocessed, and the preprocessing method includes standardization processing and normalization processing.

[0020] Preferably, the preprocessing method further includes abnormal data processing, and the abnormal data processing method is to remove abnormal data according to the weight of abnormal data.

[0021] Further, in the step S23, the proportion of the training set to the test set is 7:3.

[0022] Further, in the step S4, the aircraft steering predicted according to the output result includes:

[0023] When the output result is 1, the aircraft steering is predicted to be upward turning or left turning;

[0024] When the output result is -1, the aircraft steering is predicted to be downward turning or right turning.

[0025] The embodiment of the present application also provides an aviation turning prediction device based on a logistic regression model, comprising:

[0026] An aviation turning prediction initial model construction module is configured to adopt a loss function and an underlying regression model to construct an aviation turning prediction initial model.

[0027] A model optimization module is configured to adopt the loss function to iteratively optimize the aviation turning prediction initial model based on known quaternion data and corresponding heading angles of different flight states of an aircraft, and obtain an aviation turning prediction model based on logistic regression.

[0028] A collection and processing module is configured to collect quaternion data of the aircraft when performing flight operations in real time, and perform binary processing on the quaternion data to obtain binary quaternion data.

[0029] A turning prediction module is configured to input the binary quaternion data into the aviation turning prediction model based on logistic regression, and predict aircraft turning according to an output result.

[0030] The embodiment of the present application also provides a computer device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor implements any aviation turning prediction method based on a logistic regression model described above when executing the computer program, so as to solve the technical problem that the cost is high and the trajectory compensation is excessive when predicting aviation turning in the prior art, the heading angle value is abnormal, and the prediction accuracy is affected.

[0031] The embodiment of the present application also provides a computer readable storage medium, which stores a computer program for executing any aviation turning prediction method based on a logistic regression model described above, so as to solve the technical problem that the cost is high and the trajectory compensation is excessive when predicting aviation turning in the prior art, the heading angle value is abnormal, and the prediction accuracy is affected.

[0032] Compared with the prior art, the above at least one technical solution adopted by the embodiment of the present application can achieve at least the following beneficial effects:

[0033] The aviation turning prediction method based on a logistic regression model can realize heading prediction of the heading angle, can also predict subsequent attitude information, timely alarm abnormal flight operation, and better help the aircraft to perform normal operations according to the predetermined trajectory. This is very important for the normal operation of the aircraft.

[0034] Meanwhile, the known quaternion data of different flight states of the aircraft and the corresponding heading angles, and the loss function are used to iteratively optimize the initial model of the aircraft turning prediction, so that a decision boundary capable of accurately predicting the heading turning can be obtained, the stability of the prediction relationship between the quaternion data set and the target data set is ensured, the reliability of the generated model is ensured, and the accuracy of the aircraft turning prediction is improved. BRIEF DESCRIPTION OF DRAWINGS

[0035] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed in the embodiments will be briefly introduced as follows. Obviously, the drawings in the following description are only some embodiments of the present application, and all other drawings obtained by those of ordinary skill in the art without creative labor based on these drawings are within the scope of protection of the present application.

[0036] Figure 1 The flowchart of the aircraft turning prediction method based on the logistic regression model in the embodiments of the present application;

[0037] Figure 2 The flowchart of the aircraft turning prediction model construction based on the logistic regression in the embodiments of the present application;

[0038] Figure 3 The schematic diagram of the computer device in the embodiments of the present application;

[0039] Figure 4 The structural diagram of the aircraft turning prediction device based on the logistic regression model in the embodiments of the present application;

[0040] Wherein, 301, memory; 302, processor; 401, aircraft turning prediction initial model construction module; 402, model optimization module; 403, acquisition and processing module; 404, turning prediction module. DETAILED DESCRIPTION

[0041] The embodiments of the present application will be described in detail below with reference to the drawings.

[0042] The embodiments of the present application will be described in detail below with reference to the drawings.

[0043] The embodiment of the present application provides an aviation turning prediction method based on a logistic regression model, referring to Figure 1 and Figure 2 The aviation turning prediction method comprises the following steps:

[0044] S1, adopting a loss function and a bottom regression model, an aviation turning prediction initial model is constructed.

[0045] The loss function comprises any one of a gradient descent method, a steepest descent method, an Adam algorithm and an AdaGrad algorithm.

[0046] In this step, the loss function is introduced into the selected bottom regression model to form an aviation turning prediction initial model, and the initial model constructed by assuming that target data is subject to a certain specific distribution can perform the highest possibility classification on specific parameters (i.e. quaternion data) through maximum likelihood estimation. The model is expected to fit a decision boundary (the boundary of the heading angle), establish the probability size of the boundary and classification, and finally realize the aviation turning prediction function. The loss function can iteratively optimize the model during the optimization of the aviation turning prediction initial model, and obtain the optimal decision boundary, so that the aviation turning can be accurately predicted through the optimal decision boundary.

[0047] S2, based on known quaternion data of different flight states of an aircraft and corresponding heading angles, the aviation turning prediction initial model is iteratively optimized by using the loss function, and an aviation turning prediction model based on logistic regression is obtained.

[0048] In this step S2, the specific method for obtaining the aviation turning prediction model based on logistic regression comprises the following steps:

[0049] S21, acquiring a quaternion data set of the aircraft in different flight modes and a target data set of the corresponding heading angles of each quaternion data;

[0050] S22, performing binary processing on the quaternion data set and the target data set to obtain a binary quaternion data set and a binary target data set;

[0051] S23, dividing the binary quaternion data set and the binary target data set into a training set and a test set according to the same proportion. Preferably, the proportion of the training set to the test set is 7:3.

[0052] S24, training by using the training set and testing by using the test set, and iteratively optimizing the aviation turning prediction initial model by using the loss function.

[0053] Since the amplitude of the quaternion data and the heading angle data are not the same, in order to ensure subsequent model fitting, the quaternion data set and the heading angle target data set of step S21 need to be preprocessed, for example, using standardization processing to convert the feature matrix of the overall sample distribution of the data to standard normal distribution; using normalization processing to convert the feature values of the sample to the same dimension and map the data to the interval [0, 1] or [-1, 1]. The two preprocessing methods can make the data characteristics more obvious and clear, reduce the interference of the data itself, and facilitate the optimization learning of the model.

[0054] Specifically, the quaternion data set and the target data set are standardized, and a unified formula x' = (x - μ) / σ is used for data standardization, x' is the standardized data, x is the data before standardization, μ is the mean, and σ is the standard deviation. In this embodiment, the data conforms to the standard normal distribution, that is, the mean μ is 0 and the standard deviation σ is 1.

[0055] The normalization processing method is to map the standardized quaternion data set and the target data set to (-1, 1) through the formula, and map the target data set to vector -1 or 1 through the 0-axis criterion. At the same time, in order to make the data characteristics more obvious and clear, and further improve the learning efficiency of the model, the quaternion data is binarized, the quaternion in the quaternion data set is mapped to (-1, 1); the heading angle in the target data set is mapped to (-1, 1), and the data in the range of (-1, 0) is defined as -1, or the data in [0, 1) is defined as 1, which is convenient for model learning.

[0056] Preferably, the above preprocessing method further includes an abnormal data processing method, which is an abnormal data weight elimination method.

[0057] S3, real-time acquisition of quaternion data when the aircraft performs flight operation, and binarization processing of the quaternion data to obtain binarized quaternion data.

[0058] S4, inputting the binarized quaternion data into the aviation turning prediction model based on logistic regression, and predicting the turning of the aircraft according to the output result, wherein the turning of the aircraft includes any one of upward turning, left turning, downward turning and right turning. For example, when the output result of the model is 1, it is predicted that the aircraft turns upward or left; when the output result of the model is -1, it is predicted that the aircraft turns downward or right.

[0059] In this embodiment, a computer device is provided, as shown in Figure 3As shown, the computer device includes a memory 301, a processor 302, and a computer program stored in the memory 301 and executable on the processor 302, and the processor implements the above-mentioned any aviation turning prediction method based on a logistic regression model when executing the computer program.

[0060] In particular, the computer device can be a computer terminal, a server, or similar computing device.

[0061] In the embodiment, a computer readable storage medium is provided, which stores a computer program for executing the above-mentioned any aviation turning prediction method based on a logistic regression model.

[0062] In particular, the computer readable storage medium includes permanent and non-permanent, removable and non-removable media, which can be implemented by any method or technology to store information. The information can be computer readable instructions, data structures, program modules or other data. Examples of computer readable storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disc read-only memory (CD-ROM), digital versatile disc (DVD) or other optical storage, magnetic cassette, magnetic tape disk storage or other magnetic storage device, or any other non-transmission medium that can be used to store information that can be accessed by a computing device. According to the definition herein, the computer readable storage medium does not include transitory computer readable media such as modulated data signals and carriers.

[0063] Based on the same inventive concept, the embodiment of the present application also provides an aviation turning prediction device based on a logistic regression model, as described in the following embodiments. Since the principle of solving problems of the aviation turning prediction device based on a logistic regression model is similar to that of the aviation turning prediction method based on a logistic regression model, the implementation of the aviation turning prediction device can refer to the implementation of the aviation turning prediction method, and the repeated parts will not be described. The term "unit" or "module" used below can be a combination of software and / or hardware that implements a predetermined function. Although the device described in the following embodiments is preferably implemented in software, hardware, or a combination of software and hardware is also possible and is contemplated.

[0064] Figure 4 is a structural block diagram of the aviation turning prediction device based on a logistic regression model of the embodiment of the present application, as Figure 4As shown, the aerial turning prediction device comprises an aerial turning prediction initial model construction module 401, a model optimization module 402, a collection and processing module 403, and a turning prediction module 404, and the structure is described as follows.

[0065] The aerial turning prediction initial model construction module 401 is configured to construct an aerial turning prediction initial model by using a loss function and a bottom regression model.

[0066] The model optimization module 402 is configured to iteratively optimize the aerial turning prediction initial model by using the loss function based on known quaternion data and corresponding heading angles of different flight states of the aircraft, to obtain a logic regression-based aerial turning prediction model.

[0067] The collection and processing module 403 is configured to collect quaternion data in real time when the aircraft performs flight operations, and to obtain binary quaternion data by binary processing the quaternion data.

[0068] The turning prediction module 404 is configured to input the binary quaternion data into the logic regression-based aerial turning prediction model, and to predict the turning of the aircraft according to the output result.

[0069] The embodiment of the present application achieves the following technical effects:

[0070] The aerial turning prediction method based on the logic regression model can realize the heading prediction of the heading angle, and can also predict the subsequent attitude information, timely alarm abnormal flight operation, and better help the aircraft to perform normal operations according to the predetermined trajectory. This is very important for the normal operation of the aircraft.

[0071] Meanwhile, when the known quaternion data and corresponding heading angles of different flight states of the aircraft and the loss function are used to iteratively optimize the aerial turning prediction initial model, a decision boundary capable of accurately predicting the heading turning can be obtained, the stability of the prediction relationship between the quaternion data set and the target data set is ensured, the reliability of the generated model is ensured, and the accuracy of the aircraft turning prediction is improved.

[0072] Obviously, those skilled in the art should understand that each module or each step of the above-mentioned embodiments of the present application can be realized by a general computing device, which can be centralized on a single computing device or distributed on a network composed of multiple computing devices, and optionally, each module or each step can be realized by program codes executable by a computing device, so that each module or each step can be stored in a storage device and executed by a computing device, and in some cases, the steps shown or described can be executed in different orders, or each module can be manufactured as an individual integrated circuit module, or multiple modules or steps can be manufactured as a single integrated circuit module. Therefore, the embodiments of the present application are not limited to any specific combination of hardware and software.

[0073] The above only describes the preferred embodiments of the present application and is not intended to limit the present application. The embodiments of the present application can be variously changed and modified by those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. An aviation diversion prediction method based on a logistic regression model, characterized in that, The method comprises the steps of: An initial model for predicting the aviation steering is constructed by using a loss function and a bottom regression model; Based on the known quaternion data and the corresponding heading angles of the aircraft in different flight states, the initial model for predicting the aviation steering is iteratively optimized by using the loss function to obtain a model for predicting the aviation steering based on the logistic regression, including: obtaining a set of quaternion data and a set of target data of the heading angles corresponding to each quaternion data when the aircraft flies in different flight modes; performing binary processing on the set of quaternion data and the set of target data to obtain a set of binary quaternion data and a set of binary target data; dividing the set of binary quaternion data and the set of binary target data into a training set and a test set according to the same proportion; training and testing by using the training set and the test set, iteratively optimizing the initial model for predicting the aviation steering by using the loss function until a model for predicting the aviation steering based on the logistic regression is obtained; Real-time acquisition of quaternion data when the aircraft performs flight operations, binary processing of the quaternion data to obtain binary quaternion data; Inputting the binary quaternion data into the model for predicting the aviation steering based on the logistic regression, and predicting the aircraft steering according to the output result, wherein the aircraft steering includes any one of upward tilt, left turn, downward tilt and right turn. 2.The method of claim 1, wherein, The loss function includes any one of gradient descent method, steepest descent method, Adam algorithm and AdaGrad algorithm. 3.The method of claim 1, wherein, Further comprising: Preprocessing the set of quaternion data and the set of target data of the heading angles, including standardization processing and normalization processing. 4.The method of claim 3, wherein, The preprocessing further includes abnormal data processing, which is to exclude abnormal data according to the weight of the abnormal data. 5.The method of claim 1, wherein, The proportion of the training set to the test set is 7:

3. 6.The method of claim 1, wherein, The output result includes: When the output result is 1, the aircraft steering is predicted to be upward tilt or left turn; When the output result is -1, the aircraft steering is predicted to be downward tilt or right turn.

7. An aircraft diversion prediction apparatus based on a logistic regression model, characterized in that, The method comprises the steps of: An initial model for predicting the aviation steering is constructed by using a loss function and a bottom regression model; Based on the known quaternion data and the corresponding heading angles of the aircraft in different flight states, the initial model for predicting the aviation steering is iteratively optimized by using the loss function to obtain a model for predicting the aviation steering based on the logistic regression, including: obtaining a set of quaternion data and a set of target data of the heading angles corresponding to each quaternion data when the aircraft flies in different flight modes; performing binary processing on the set of quaternion data and the set of target data to obtain a set of binary quaternion data and a set of binary target data; dividing the set of binary quaternion data and the set of binary target data into a training set and a test set according to the same proportion; training and testing by using the training set and the test set, iteratively optimizing the initial model for predicting the aviation steering by using the loss function until a model for predicting the aviation steering based on the logistic regression is obtained; The collection and processing module is configured to collect quaternion data in real time when the aircraft performs a flight task, and perform binary processing on the quaternion data to obtain binary quaternion data. The turning prediction module is configured to input the binary quaternion data into the aviation turning prediction model based on the logistic regression, and predict the turning of the aircraft according to an output result.

8. A computer device comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, The processor implements the aviation turning prediction method based on the logistic regression model in any one of claims 1 to 6 when executing the computer program.

9. A computer-readable storage medium, characterized in that, The computer readable storage medium stores the computer program for implementing the aviation turning prediction method based on the logistic regression model in any one of claims 1 to 6.

Images