Airborne meteorological radar echo tracking and predicting method
Through neural network models, predicting the position and intensity of future meteorological targets is solved, and the problem of difficulty in accurately predicting changes in meteorological targets in the existing technology is improved, and the perception ability of onboard meteorological radar is improved, ensuring flight safety and issuing alarms in advance.
Patent Information
- Application Number
- CN202411370475.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-29
- Publication Date
- 2025-01-10
AI Technical Summary
The existing airborne meteorological radar system is difficult to accurately predict the location and intensity changes of future meteorological targets, resulting in flight safety threatening and the pilot's workload increased, and the prediction results are uncertain.
Using neural network models, especially recurrent neural network models, predict future meteorological targets based on historical radar data. By collecting and preprocessing radar data, the model is trained to predict the development of meteorological targets and analyze them with the expected path of the aircraft. When the predicted meteorological target path is consistent with the aircraft path and the intensity reaches the threat threshold, an alarm and guidance message is issued.
Accurate prediction of changes in meteorological target positions and intensity are achieved, the perception and early warning capabilities of onboard meteorological radar are improved, flight safety is ensured, and alarms are issued in advance to provide time for route modification.
Smart Images
Figure CN119291635A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of airborne weather radar, and in particular to an airborne weather radar echo tracking and prediction method. Background Art
[0002] Airborne weather radar systems are used to detect and provide real-time information about the location and intensity of weather conditions ahead of the aircraft's current flight path. Typically, a radar antenna mounted on the nose of the aircraft scans in a prescribed manner to detect weather conditions within an arc centered on the aircraft's flight path, and displays information about the location and intensity of weather conditions at the current moment to the crew on a radar screen installed in the cockpit. For weather conditions ahead of the flight path at future moments, the use of traditional radar systems requires pilots to judge and analyze potential weather threats from the current radar display information based on a lot of training and experience.
[0003] In the actual atmosphere, there are many types of meteorological targets that affect flight safety with complex characteristics. They are usually characterized by strong suddenness, short life span, small scale, and high intensity, such as thunderstorms, turbulence, wind shear, lightning, hail, high-altitude ice crystals, etc. Therefore, for the development and changes of meteorological targets in the future, the pilot's empirical judgment based on the current radar display information will not only increase the pilot's workload, but the prediction results will also have strong uncertainty. It is impossible to accurately predict the changes in the position and intensity of meteorological targets and their impact on aircraft, making it difficult to ensure flight safety. Summary of the invention
[0004] In view of this, the present application provides an airborne weather radar echo tracking and prediction method, which solves the problems in the prior art, improves the airborne weather radar's perception and warning capabilities for atmospheric threats, and ensures flight safety.
[0005] The present application provides an airborne weather radar echo tracking and prediction method using the following technical solutions:
[0006] An airborne weather radar echo tracking and prediction method, comprising:
[0007] Step 1: Collect radar data of multiple historical hazardous weather targets and pre-process the historical radar data;
[0008] Step 2: Train the neural network model using historical weather radar echo data;
[0009] Step 3: Input the meteorological radar echo data in the current period into the trained neural network model to predict the development of meteorological targets in the future period;
[0010] Step 4: Analyze the expected flight path of the aircraft with the predicted future position and intensity of the hazardous meteorological target; when the predicted path of the hazardous meteorological target is consistent with the expected flight path of the aircraft and its intensity reaches the threshold that threatens flight safety, a warning and guidance information is issued.
[0011] Optionally, the step of preprocessing the historical radar data in step 1 includes: determining the three-dimensional geographic position information of each range gate of the echo according to the three-dimensional geographic position of the radar in the historical radar data, and cropping the historical radar echo data in the same continuous time to ensure that the radar echo data covers the same three-dimensional geographic position area in the continuous time.
[0012] Optionally, the radar data includes: the three-dimensional geographic position of the radar corresponding to each weather echo collection and radar echo data of multiple azimuths and multiple elevation angles of the radar position in a continuous time when the weather target is detected;
[0013] The azimuth angles corresponding to different data obtained at the same airborne position are evenly distributed between 0 and 360 degrees, and the pitch angles corresponding to different data obtained at the same airborne position are between -90 and +90 degrees.
[0014] Optionally, the radar echo data includes an echo image or a reflectivity.
[0015] Optional weather targets include thunderstorms, turbulence, wind shear, lightning, hail, and high-altitude ice crystals.
[0016] Optionally, in step 2, the neural network model is a recurrent neural network model.
[0017] Optionally, the threshold in step 4 is that the radar reflectivity is greater than 40 dBZ or the 1 km radial F factor is greater than 0.13.
[0018] In summary, this application includes the following beneficial technical effects:
[0019] This application combines historical radar echo data sets to extrapolate the temporal and spatial changes of the reflectivity factor, and trains a neural network based on a large amount of historical radar echo data. It can objectively and accurately obtain the changes in the position and intensity of meteorological targets in the short term in the future and their impact on aircraft.
[0020] The algorithm in this application has a relatively high computational efficiency, so warning and guidance information can be issued early enough before dangerous weather affects the safety of the aircraft, providing sufficient time for appropriate route modifications.
[0021] The present invention can be applied to the field of airborne meteorological radars to enhance the detection, perception and warning capabilities of airborne meteorological radars for meteorological targets, thereby further improving flight safety. BRIEF DESCRIPTION OF THE DRAWINGS
[0022] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0023] Figure 1 A schematic diagram of the flow chart of the airborne weather radar echo tracking and prediction method of the present application;
[0024] Figure 2 The meteorological target intensity and position information at each interval Δt in the future and the schematic diagram of the expected path of the aircraft calculated in the embodiment of the present application;
[0025] Figure 3 This is a schematic diagram of the process structure of an embodiment of the present application for warning and guiding according to the expected path of the aircraft and meteorological target prediction information. DETAILED DESCRIPTION
[0026] The embodiments of the present application are described in detail below with reference to the accompanying drawings.
[0027] The following describes the implementation methods of the present application through specific examples, and those skilled in the art can easily understand other advantages and effects of the present application from the contents disclosed in this specification. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. The present application can also be implemented or applied through other different specific implementation methods, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present application. It should be noted that, in the absence of conflict, the following embodiments and the features in the embodiments can be combined with each other. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without making creative work belong to the scope of protection of the present application.
[0028] It should be noted that various aspects of the embodiments within the scope of the appended claims are described below. It should be apparent that the aspects described herein can be embodied in a wide variety of forms, and any specific structure and / or function described herein is merely illustrative. Based on the present application, it should be understood by those skilled in the art that an aspect described herein can be implemented independently of any other aspect, and two or more of these aspects can be combined in various ways. For example, any number of aspects described herein can be used to implement the device and / or practice the method. In addition, other structures and / or functionalities other than one or more of the aspects described herein can be used to implement this device and / or practice this method.
[0029] It should also be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present application. The illustrations only show the components related to the present application rather than being drawn according to the number, shape and size of the components in actual implementation. In actual implementation, the type, quantity and proportion of each component may be changed at will, and the component layout may also be more complicated.
[0030] Additionally, in the following description, specific details are provided to facilitate a thorough understanding of the examples. However, it will be understood by those skilled in the art that the aspects described may be practiced without these specific details.
[0031] An embodiment of the present application provides an airborne weather radar echo tracking and prediction method.
[0032] like Figure 1 As shown, an airborne weather radar echo tracking and prediction method comprises:
[0033] Step 1: Collect radar data of multiple historical hazardous weather targets and preprocess the historical radar data.
[0034] Step 2: Train the neural network model using historical weather radar echo data.
[0035] Step 3: Input the weather radar echo data in the current period into the trained neural network model to predict the development of weather targets in the future period.
[0036] Step 4: Analyze the expected flight path of the aircraft with the predicted future position and intensity of the hazardous meteorological target; when the predicted path of the hazardous meteorological target is consistent with the expected flight path of the aircraft and its intensity reaches the threshold that threatens flight safety, issue appropriate warning and guidance information.
[0037] In step 1, radar data includes: radar three-dimensional geographic position corresponding to each meteorological echo collection and radar echo data of multiple azimuths and multiple elevation angles of the radar position in continuous time when the meteorological target is detected; radar echo data includes echo pictures or reflectivity, and the echo picture file can be the original radar echo image, or it can be an echo image that only displays three colors of red, yellow, and green, or it can be an echo image based on a specific recognition algorithm to display the position and intensity information of dangerous meteorological conditions such as thunderstorms, turbulence, wind shear, lightning, hail, and high-altitude ice crystals. The radar echo data includes the position of the aircraft, the relative position of the meteorological target and the aircraft, and the threat level of the meteorological target. Among them, the picture file of the present application can be in any format, and this embodiment does not impose specific restrictions on this.
[0038] The azimuth angles corresponding to different data obtained at the same airborne position are evenly distributed between 0 and 360 degrees, and the pitch angles corresponding to different data obtained at the same airborne position are between -90 and +90 degrees. The full-range echo extrapolation prediction can be performed based on the current radar data to expand the prediction range.
[0039] The steps of preprocessing the historical radar data include: determining the three-dimensional geographic position information of each range gate of the echo according to the three-dimensional geographic position of the radar in the historical radar data, clipping the historical radar echo data in the same continuous time to ensure that the radar echo data covers the same three-dimensional geographic position area in the continuous time; aligning the airborne radar echo data at different times according to the position information such as the longitude, latitude and altitude of the aircraft, so that the longitude, latitude and altitude indicated by the corresponding pixels in the picture are completely consistent. Among them, the areas that do not overlap in the original echo image need to be cropped, and it is necessary to ensure that when dangerous meteorological targets appear, the corresponding areas of the echo data at different times include the location and intensity information of the meteorological targets. Among them, meteorological targets include thunderstorms, turbulence, wind shear, lightning, hail and high-altitude ice crystals.
[0040] In step 2, the neural network model is a recurrent neural network model, abbreviated as RNN. The recurrent neural network can effectively reflect the time-varying characteristics of the data stream and make predictions on the time series. The input data information at time t is linearly combined and nonlinearly transformed through the "gating structure" in the long short-term memory unit. The long short-term memory neural unit is referred to as the LSTM neural unit below. The obtained information is converted into Ct and ht in the form of tensors, and is transmitted to the LSTM neural unit at time t+1, and participates in the calculation of the input data information at time t+1. The training model structure adopts an Encoder-Decoder sequence structure that is transmitted by upsampling, downsampling and the same-layer hidden state, and strengthens the transmission of the high network layer information at time t to the initial network layer information at time t+1 in the Decoder part, and designs the connection structure between layers. The grid data at the initial moment is downsampled by convolution calculation, the spatial features are extracted, and the output tensor is input into the LSTM neural unit. Finally, after 3 identical calculation processes, the corresponding size data tensor is obtained. At the same time, the input data at subsequent moments is calculated in the same way as at the initial moment, and the transmission of the output ht at the previous moment is received when it is input to the LSTM neural unit node, and finally the transmission of the data tensor of the entire Encoder part is completed, and the combination of the encoded information is transmitted to the Decoder part. In the Decoder part, the calculation process is similar to that of the Encoder, which is a bottom-up decoding process. The encoded information transmitted by the Encoder is transmitted to the LSTM neural unit in time series, and then upsampled through inverse convolution. The LSTM neural unit of each layer receives the ht information of the ConvLSTM neural unit at the last moment of the Encoder layer, and re-convolves the prediction result output at the current moment as the input at the subsequent moment of the Decoder bottom layer. Finally, the data grid output is returned at each time point, and the value in the tensor is read to obtain the prediction result at each moment.
[0041] Specifically, step 3 inputs the meteorological radar echo data in the current time period into the trained recurrent neural network model, and predicts the development of meteorological targets in the future based on the airborne meteorological radar results in the current time period to obtain the intensity and position movement of meteorological targets. Meteorological targets may include dangerous weather such as thunderstorms, turbulence, wind shear, lightning, hail, and high-altitude ice crystals.
[0042] Analyzing the expected flight path of the aircraft with the predicted future position and intensity of the dangerous meteorological target in step 4 specifically includes: comparing the expected path of the aircraft with the predicted information of the meteorological target. In the embodiment of the present application, the airborne meteorological radar system communicates with the airborne flight management system, accesses the expected flight path of the aircraft stored therein, and compares it with the future position and intensity information of the meteorological target predicted by the recurrent neural network, such as Figure 2 As shown, Figure 2 The objects indicated by numbers 40, 50, 60, and 70 are the predicted future positions and intensity information of the meteorological targets, the object indicated by number 80 is the aircraft position diagram, and the object indicated by 90 is the expected flight path of the aircraft. When the expected flight path overlaps with the predicted future position of the meteorological target, it is considered that the predicted dangerous meteorological path is consistent with the expected flight path of the aircraft. The predicted future position and intensity information of the meteorological target is not limited to the above four moments. When the time difference between adjacent moments is less than 30 minutes, the prediction results obtained by the method of the present application have a high degree of credibility.
[0043] like Figure 3 As shown, firstly, the predicted meteorological target position information is compared and analyzed with the expected flight path. If the two paths do not overlap, then continue to fly according to the expected flight path. If the two paths overlap, then detect the intensity information of the meteorological target in the overlapped prediction. If the intensity does not exceed the danger threshold, then still fly according to the expected flight path. If its intensity exceeds the danger threshold, then according to the predicted meteorological target information, alarm and guidance are performed, and a warning is issued before the predicted path overlaps, and the flight route is modified to avoid dangerous meteorological targets. Among them, the danger threshold can be radar reflectivity, and an alarm is issued when the radar reflectivity is greater than 40dBZ. The danger threshold can also be the 1km radial average F factor, and an alarm is issued when the 1km radial F factor is greater than 0.13. The danger threshold can also be other physical quantities representing dangerous weather.
[0044] An embodiment of the present application also discloses an airborne radar echo tracking and prediction system based on a recurrent neural network, including a device for storage, calculation and warning guidance, wherein the device is configured to: store and obtain radar echo data of historical dangerous meteorological targets; store and obtain radar echo data at the current moment; train a neural network model based on a large amount of historical echo data; calculate and obtain a data file of the intensity and position information of meteorological targets predicted by the neural network model in a short time in the future from the current moment; communicate with an airborne flight management system to access the expected flight path of the aircraft stored therein; calculate and obtain information on the predicted future position and intensity of dangerous meteorological targets relative to the expected flight path of the aircraft; and issue appropriate warning and guidance information when the predicted dangerous meteorological path is consistent with the expected flight path of the aircraft and its intensity reaches a threshold that threatens flight safety.
[0045] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed in the present application should be included in the protection scope of the present application. Therefore, the protection scope of the present application shall be based on the protection scope of the claims.
Claims
1. A method for tracking and predicting airborne weather radar echoes, characterized in that: include: Step 1: Collect radar data of multiple historical hazardous weather targets and pre-process the historical radar data; Step 2: Train the neural network model using historical weather radar echo data; Step 3: Input the meteorological radar echo data in the current period into the trained neural network model to predict the development of meteorological targets in the future period; Step 4: Analyze the expected flight path of the aircraft with the predicted future position and intensity of the hazardous meteorological target; when the predicted path of the hazardous meteorological target is consistent with the expected flight path of the aircraft and its intensity reaches the threshold that threatens flight safety, a warning and guidance information is issued.
2. The airborne weather radar echo tracking and prediction method according to claim 1, characterized in that: The step of preprocessing the historical radar data in step 1 includes: determining the three-dimensional geographic position information of each range gate of the echo according to the three-dimensional geographic position of the radar in the historical radar data, and cutting the historical radar echo data in the same continuous time to ensure that the radar echo data covers the same three-dimensional geographic position area in the continuous time.
3. The airborne weather radar echo tracking and prediction method according to claim 2, characterized in that: Radar data includes: the three-dimensional geographic position of the radar corresponding to each weather echo collection, and radar echo data of multiple azimuths and multiple elevation angles of the radar position in continuous time when a weather target is detected; The azimuth angles corresponding to different data obtained at the same airborne position are evenly distributed between 0 and 360 degrees, and the pitch angles corresponding to different data obtained at the same airborne position are between -90 and +90 degrees.
4. The airborne weather radar echo tracking and prediction method according to claim 3, characterized in that: Radar echo data includes echo images or reflectivity.
5. The airborne weather radar echo tracking and prediction method according to claim 1, characterized in that: Meteorological targets include thunderstorms, turbulence, wind shear, lightning, hail, and high-altitude ice crystals.
6. The airborne weather radar echo tracking and prediction method according to claim 1, characterized in that: In step 2, the neural network model is a recurrent neural network model.
7. The airborne weather radar echo tracking and prediction method according to claim 1, characterized in that: The threshold in step 4 is that the radar reflectivity is greater than 40 dBZ or the 1 km radial F factor is greater than 0.13.