A bird monitoring method and system based on radar data

Abstract

The present application relates to a method and system for bird monitoring based on radar data, and belongs to the field of data processing technology. The method includes: obtaining historical multidimensional data, including first time series information of radar and second time series information of flying birds; initializing and generating first node features according to the first time series information, and generating first hyperedges based on the first node features, and generating first hypergraphs based on the nodes corresponding to the first node features and the first hyperedges; initializing and generating second node features according to the second time series information, and generating second hyperedges based on the second node features, and generating second hypergraphs based on the nodes corresponding to the second node features and the second hyperedges; performing hypergraph convolution on the first hypergraph and the second hypergraph respectively to obtain first features and second features; inputting the first features and the second features into a bird assessment model, and determining the bird monitoring results according to the output values. The present application can improve the efficiency and accuracy of bird repelling.

Description

Technical Field

[0001] The present application relates to the field of data processing technology, and in particular to a bird monitoring method and system based on radar data. Background Art

[0002] Bird strike, also known as bird collision, refers to the incident in which birds collide with flying man-made aircraft, high-speed trains, cars, etc., causing injuries. Among them, the take-off and landing process of an aircraft is the stage where bird strikes are most likely to occur. According to statistics, more than 90% of bird strikes occur at airports and in the airspace near airports. Aircraft fly at high speeds and often cause great damage when they collide with flying birds. In severe cases, they can cause the crash of the aircraft. Therefore, bird strikes have become one of the important factors threatening aviation safety. In order to avoid bird strikes as much as possible and ensure aircraft safety, the bird repelling work at airports is highly valued. At present, airports generally use radar technology to detect birds and repel them when they are found. However, not all birds that appear will cause bird strikes. Therefore, this bird repelling method in the existing technology will undoubtedly cause a waste of bird repelling resources and lead to low bird repelling efficiency. Summary of the invention

[0003] Based on this, it is necessary to provide a bird monitoring method and system based on radar data that can effectively improve the efficiency of bird repelling in response to the above technical problems.

[0004] In a first aspect, a method for bird monitoring based on radar data is provided, the method comprising:

[0005] Acquire historical multidimensional data, where the historical multidimensional data includes at least first time series information of the radar and second time series information of the bird, wherein the time nodes of the first time series information and the second time series information correspond one to one, and the time series information includes spatial dimension information and time dimension information;

[0006] Initialize and generate a first node feature according to the first time series information, generate a first hyperedge based on the first node feature, and generate a first hypergraph based on the node corresponding to the first node feature and the first hyperedge;

[0007] Initialize and generate a second node feature according to the second time series information, generate a second hyperedge based on the second node feature, and generate a second hypergraph based on the nodes corresponding to the second node feature and the second hyperedge;

[0008] Performing hypergraph convolution on the first hypergraph and the second hypergraph respectively to obtain a first feature and a second feature;

[0009] The first feature and the second feature are input into the bird situation assessment model, and the bird situation monitoring result is determined according to the output value.

[0010] Optionally, after acquiring the historical multidimensional data, the method further includes:

[0011] The historical multi-dimensional data is cleansed, and based on the data cleaning result, the first time series information and the second time series information are classified and stored.

[0012] Optionally, the first time series information includes at least radar position information, and initializing a first node feature according to the first time series information, generating a first hyperedge based on the first node feature, and generating a first hypergraph based on the node corresponding to the first node feature and the first hyperedge includes:

[0013] The radar position information within the target time period is defined as a first node, and the target first node is initialized to obtain a first node feature;

[0014] Based on the objective function, a first likelihood value between the target first node feature and other node features is calculated, wherein the objective function includes:

[0015]

[0016] in, represents the likelihood value, represents the first node feature of the target, Represents the dimension, Indicates other node features;

[0017] Sort the multiple first likelihood values ​​in descending order, and select the node corresponding to the first likelihood value of the number of front targets to generate a first hyperedge;

[0018] A first hypergraph is generated based on the nodes corresponding to the target first node feature and the first hyperedge.

[0019] Optionally, the first time series information includes at least flying bird position information, initializing a second node feature according to the second time series information, generating a second hyperedge based on the second node feature, and generating a second hypergraph based on the node corresponding to the second node feature and the second hyperedge, comprising:

[0020] The flying bird position information within the target time period is defined as a second node, and the target second node is initialized to obtain the second node feature;

[0021] Based on the objective function, calculating a second likelihood value between the target second node feature and other node features;

[0022] Sort the multiple second likelihood numbers in descending order, and select the nodes corresponding to the second likelihood values ​​of the previous target number to generate a second hyperedge;

[0023] A second hypergraph is generated based on the nodes corresponding to the target second node feature and the second hyperedge.

[0024] Optionally, a hypergraph convolution is performed on the hypergraph to obtain corresponding features including:

[0025] Determine the degree of vertices in a hypergraph and the degree of the hyperedge ,in, Represents a vertex The degree, Represents a hyperedge The degree, represents the number of vertices, represents the number of hyperedges, represents the adjacency matrix, represents the weight coefficient;

[0026] According to the degree of the node and the degree of the hyperedge to which it belongs, a hypergraph convolution is performed on the hypergraph, and the formula of the hypergraph convolution is:

[0027]

[0028] in, represents the degree matrix of the hyperedge, represents the degree matrix of the vertex, represents the weight matrix, represents the adjacency matrix, Indicates The vertex feature relationship of the layer, represents the learnable parameters, Indicates Vertex feature relationships of layers;

[0029] Fuse multiple vertex feature relationships to obtain a fused representation of the hypergraph , and define the fusion representation as the features after hypergraph convolution.

[0030] Optionally, the bird situation assessment model includes:

[0031]

[0032] in, represents the evaluation value, represents the first eigenvalue, represents the second eigenvalue, Represents the fusion variable between spatial dimension information and temporal dimension information, represents the number of features, represents the aggregate feature value, represents a learnable parameter.

[0033] Optionally, inputting the first feature and the second feature into a bird situation assessment model, and determining the bird situation monitoring result according to the output value includes:

[0034] Obtaining a first feature and a second feature;

[0035] Inputting the first feature and the second feature into a bird situation assessment model to obtain an output value;

[0036] In response to detecting that the output value is less than a preset threshold, a bird situation warning message is sent to the user terminal.

[0037] In a second aspect, a bird monitoring system based on radar data is provided, the system comprising:

[0038] An information acquisition module, used to acquire historical multi-dimensional data, wherein the historical multi-dimensional data at least includes first time series information of radar and second time series information of flying birds, wherein the time nodes of the first time series information and the second time series information correspond one to one, and the time series information includes spatial dimension information and time dimension information;

[0039] A first hypergraph construction module, configured to initialize and generate a first node feature according to the first time series information, generate a first hyperedge based on the first node feature, and generate a first hypergraph based on the node corresponding to the first node feature and the first hyperedge;

[0040] A second hypergraph construction module, configured to initialize and generate a second node feature according to the second timing information, generate a second hyperedge based on the second node feature, and generate a second hypergraph based on the node corresponding to the second node feature and the second hyperedge;

[0041] A hypergraph convolution module, used to perform hypergraph convolution on the first hypergraph and the second hypergraph respectively to obtain a first feature and a second feature;

[0042] The bird situation assessment module is used to input the first feature and the second feature into the bird situation assessment model, and determine the bird situation monitoring result according to the output value.

[0043] Optionally, the bird situation assessment model includes:

[0044]

[0045] in, represents the evaluation value, represents the first eigenvalue, represents the second eigenvalue, Represents the fusion variable between spatial dimension information and temporal dimension information, represents the number of features, represents the aggregate feature value, represents a learnable parameter.

[0046] Optionally, inputting the first feature and the second feature into a bird situation assessment model, and determining the bird situation monitoring result according to the output value includes:

[0047] Obtaining a first feature and a second feature;

[0048] Inputting the first feature and the second feature into a bird situation assessment model to obtain an output value;

[0049] In response to detecting that the output value is less than a preset threshold, a bird situation warning message is sent to the user terminal.

[0050] In a third aspect, a computer device is provided, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the following steps are implemented:

[0051] Acquire historical multidimensional data, where the historical multidimensional data includes at least first time series information of the radar and second time series information of the bird, wherein the time nodes of the first time series information and the second time series information correspond one to one, and the time series information includes spatial dimension information and time dimension information;

[0052] Initialize and generate a first node feature according to the first time series information, generate a first hyperedge based on the first node feature, and generate a first hypergraph based on the node corresponding to the first node feature and the first hyperedge;

[0053] Initialize and generate a second node feature according to the second time series information, generate a second hyperedge based on the second node feature, and generate a second hypergraph based on the nodes corresponding to the second node feature and the second hyperedge;

[0054] Performing hypergraph convolution on the first hypergraph and the second hypergraph respectively to obtain a first feature and a second feature;

[0055] The first and second features are input into the bird situation assessment model, and the bird situation monitoring results are determined according to the output values.

[0056] In a fourth aspect, a computer-readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:

[0057] Acquire historical multidimensional data, where the historical multidimensional data includes at least first time series information of the radar and second time series information of the bird, wherein the time nodes of the first time series information and the second time series information correspond one to one, and the time series information includes spatial dimension information and time dimension information;

[0058] Initialize and generate a first node feature according to the first time series information, generate a first hyperedge based on the first node feature, and generate a first hypergraph based on the node corresponding to the first node feature and the first hyperedge;

[0059] Initialize and generate a second node feature according to the second time series information, generate a second hyperedge based on the second node feature, and generate a second hypergraph based on the nodes corresponding to the second node feature and the second hyperedge;

[0060] Performing hypergraph convolution on the first hypergraph and the second hypergraph respectively to obtain a first feature and a second feature;

[0061] The first and second features are input into the bird situation assessment model, and the bird situation monitoring results are determined according to the output values.

[0062] In a fifth aspect, a computer program product is provided, the computer program product comprising a computer program, and when the computer program is executed by a processor, the following steps are implemented:

[0063] Acquire historical multidimensional data, where the historical multidimensional data includes at least first time series information of the radar and second time series information of the bird, wherein the time nodes of the first time series information and the second time series information correspond one to one, and the time series information includes spatial dimension information and time dimension information;

[0064] Initialize and generate a first node feature according to the first time series information, generate a first hyperedge based on the first node feature, and generate a first hypergraph based on the node corresponding to the first node feature and the first hyperedge;

[0065] Initialize and generate a second node feature according to the second time series information, generate a second hyperedge based on the second node feature, and generate a second hypergraph based on the nodes corresponding to the second node feature and the second hyperedge;

[0066] Performing hypergraph convolution on the first hypergraph and the second hypergraph respectively to obtain a first feature and a second feature;

[0067] The first and second features are input into the bird situation assessment model, and the bird situation monitoring results are determined according to the output values.

[0068] The above-mentioned bird monitoring method and system based on radar data, the method includes: obtaining historical multidimensional data, the historical multidimensional data at least including first time series information of the radar, and second time series information of the flying bird, wherein the time nodes of the first time series information and the second time series information correspond one to one, and the time series information includes spatial dimension information and time dimension information; according to the first time series information, initializing the generation of a first node feature, and generating a first hyperedge based on the first node feature, and generating a first hypergraph based on the node corresponding to the first node feature and the first hyperedge; according to the second time series information, initializing the generation of a second node feature, and generating a second hyperedge based on the second node feature, and generating a second hypergraph based on the node corresponding to the second node feature and the second hyperedge; performing hypergraph convolution on the first hypergraph and the second hypergraph respectively to obtain a first feature and a second feature; inputting the first feature and the second feature into a bird situation assessment model, and determining the bird situation monitoring result according to the output value. The present application can accurately monitor and evaluate bird conditions or bird strike events, thereby improving the efficiency and accuracy of bird repelling while ensuring efficient use of resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] Figure 1 A schematic diagram of a process of a bird monitoring method based on radar data in one embodiment;

[0070] Figure 2 is a structural block diagram of a bird monitoring device based on radar data in one embodiment;

[0071] Figure 3 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION

[0072] In order to make the purpose, technical solutions and advantages of this application clearer, the technical solutions in the embodiments of this application will be clearly and completely described below in conjunction with the drawings in the embodiments of this application. Obviously, the described embodiments are only part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0073] It should be understood that in the description of the present application, unless the context clearly requires otherwise, words such as "include", "comprises", and the like throughout the specification should be interpreted as including rather than being exclusive or exhaustive; that is, as including but not limited to.

[0074] It should also be understood that the terms "first", "second", etc. are only used for descriptive purposes and cannot be understood as indicating or implying relative importance. In addition, in the description of this application, unless otherwise specified, "plurality" means two or more.

[0075] It should be noted that the terms "S1", "S2", etc. are only used for the purpose of describing the steps, and do not specifically refer to the order or sequence, nor are they used to limit the present application. They are only for the convenience of describing the method of the present application, and cannot be understood as indicating the order of the steps. In addition, the technical solutions between the various embodiments can be combined with each other, but they must be based on the ability of ordinary technicians in this field to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection required by the present application.

[0076] In one embodiment, Figure 1 As shown, a bird monitoring method based on radar data is provided, comprising the following steps:

[0077] S1: Acquire historical multi-dimensional data, which includes at least first time series information of radar and second time series information of flying birds, wherein the time nodes of the first time series information and the second time series information correspond one to one, and the time series information includes spatial dimension information and time dimension information.

[0078] In some specific embodiments, historical multidimensional data includes first time series information of the radar and second time series information of the bird. The first time series information may include the position information of the radar, such as the coordinate information of the radar, etc., and may also include the time information when the radar is at a certain position, and the angle at which the radar sends and receives signals, etc. The radar position may be variable or fixed. For example, if the radar is mounted on a drone, the position information within a time period may be variable. If the radar is fixed on the ground, the radar position information is fixed; the second time series information may include the position information of the bird, such as the coordinate information of the bird, etc., and may also include the time information when the bird is at a certain position, and information such as the speed of the bird within a time period.

[0079] In some specific implementations, after acquiring the historical multidimensional data, the method further includes:

[0080] The historical multidimensional data is cleaned, and based on the data cleaning result, the first time series information and the second time series information are classified and stored, wherein data cleaning is a common technique and will not be repeated here, and the classified storage is to store the first time series information and the second time series information in different databases to facilitate the subsequent extraction of different time series information to construct a hypergraph.

[0081] S2: Initialize and generate a first node feature according to the first timing information, generate a first hyperedge based on the first node feature, and generate a first hypergraph based on the nodes corresponding to the first node feature and the first hyperedge.

[0082] S3: Initialize and generate a second node feature according to the second timing information, generate a second hyperedge based on the second node feature, and generate a second hypergraph based on the nodes corresponding to the second node feature and the second hyperedge.

[0083] It should be noted that a hypergraph may include multiple nodes and hyperedges. A hypergraph may represent high-order dependencies between nodes. Each node may carry additional information, such as timing information, to form node features.

[0084] In some specific implementations, initializing a first node feature according to the first time series information, generating a first hyperedge based on the first node feature, and generating a first hypergraph based on the node corresponding to the first node feature and the first hyperedge includes:

[0085] The radar position information within the target time period is defined as a first node, and the target first node is initialized to obtain a first node feature, wherein the initialization processing method is to randomly initialize the vector feature corresponding to the node;

[0086] Based on the objective function, a first likelihood value between the target first node feature and other node features is calculated, wherein the objective function includes:

[0087]

[0088] in, represents the likelihood value, represents the first node feature of the target, Represents the dimension, Indicates other node features;

[0089] Sort multiple first likelihood values ​​in descending order, and select nodes corresponding to the first likelihood values ​​of the first target number to generate a first hyperedge, wherein the target number can be set according to actual needs, such as 4, that is, select the first 4 nodes with the highest sorting and the nodes corresponding to the target first node feature to construct the first hyperedge;

[0090] Based on the nodes corresponding to the target first node feature and the first hyperedge, a first hypergraph is generated, wherein the nodes corresponding to the target first node feature include the nodes corresponding to themselves and the nodes corresponding to the first likelihood values ​​of the target number selected above.

[0091] In some specific implementations, initializing a second node feature according to the second timing information, generating a second hyperedge based on the second node feature, and generating a second hypergraph based on the node corresponding to the second node feature and the second hyperedge includes:

[0092] The flying bird position information within the target time period is defined as a second node, and the target second node is initialized to obtain the second node feature, wherein the initialization method is to randomly initialize the vector feature corresponding to the node;

[0093] Based on the objective function, calculating a second likelihood value between the target second node feature and other node features;

[0094] Sort the multiple second likelihood numbers in descending order, and select the nodes corresponding to the second likelihood values ​​of the first target number to generate a second hyperedge, wherein the target number can be set according to actual needs, such as 4, that is, select the first 4 nodes and the nodes corresponding to the target second node features to construct the second hyperedge;

[0095] Based on the nodes corresponding to the target second node feature and the second hyperedge, a second hypergraph is generated, wherein the nodes corresponding to the target second node feature include the nodes corresponding to themselves and the nodes corresponding to the second likelihood values ​​of the target number selected above.

[0096] S4: Perform hypergraph convolution on the first hypergraph and the second hypergraph respectively to obtain a first feature and a second feature.

[0097] It should be noted that a hypergraph convolutional neural network is generally used to perform convolution processing on the hypergraph.

[0098] In some specific implementations, the hypergraph is convolved to obtain corresponding features including:

[0099] Determine the degree of vertices in a hypergraph and the degree of the hyperedge ,in, Represents a vertex The degree, Represents a hyperedge The degree, represents the number of vertices, represents the number of hyperedges, represents the adjacency matrix, represents the weight coefficient;

[0100] According to the degree of the node and the degree of the hyperedge to which it belongs, a hypergraph convolution is performed on the hypergraph, and the formula of the hypergraph convolution is:

[0101]

[0102] in, represents the degree matrix of the hyperedge, represents the degree matrix of the vertex, represents the weight matrix, represents the adjacency matrix, Indicates The vertex feature relationship of the layer, represents the learnable parameters, Indicates Vertex feature relationships of layers;

[0103] Fuse multiple vertex feature relationships to obtain a fused representation of the hypergraph , and define the fusion representation as the features after hypergraph convolution.

[0104] Specifically, if the first hypergraph is convolved, the hypergraph in the above steps refers to the first hypergraph, and the fusion representation obtained is , if the second hypergraph is convolved, the hypergraph in the above steps refers to the second hypergraph, and the resulting fusion is expressed as .

[0105] S5: Input the first feature and the second feature into a bird situation assessment model, and determine the bird situation monitoring result according to the output value.

[0106] In some embodiments, the bird condition assessment model includes:

[0107]

[0108] in, represents the evaluation value, represents the first eigenvalue, represents the second eigenvalue, Represents the fusion variable between spatial dimension information and temporal dimension information, represents the number of features, represents the aggregate feature value, represents a learnable parameter.

[0109] Specifically, the bird situation assessment model can be trained and verified through historical data sets. When its assessment accuracy reaches the preset standard, the final bird situation assessment model can be output. The preset standard can be set according to actual needs, such as the assessment accuracy reaches 90%, etc. The specific training and verification process is a common method and will not be repeated here.

[0110] In some specific implementations, the first feature and the second feature are input into a bird situation assessment model, and according to the output value, determining the bird situation monitoring result includes:

[0111] Obtaining a first feature and a second feature;

[0112] Inputting the first feature and the second feature into a bird situation assessment model to obtain an output value, wherein the data value refers to the above-mentioned assessment value;

[0113] In response to detecting that the output value is less than a preset threshold, a bird situation warning message is sent to the user end, wherein the preset threshold can be set according to actual needs. Generally, the smaller the output value, the greater the possibility of a bird strike incident. At this time, it is necessary to send a warning message to the workers so that the staff can take bird-repelling measures in time to improve the efficiency of bird-repelling.

[0114] In the above-mentioned bird monitoring method based on radar data, the method includes: obtaining historical multidimensional data, the historical multidimensional data at least including first time series information of the radar and second time series information of the flying bird, wherein the time nodes of the first time series information and the second time series information correspond one to one, and the time series information includes spatial dimension information and time dimension information; according to the first time series information, initializing the generation of a first node feature, and generating a first hyperedge based on the first node feature, and generating a first hypergraph based on the node corresponding to the first node feature and the first hyperedge; according to the second time series information, initializing the generation of a second node feature, and generating a second hyperedge based on the second node feature, and generating a second hypergraph based on the node corresponding to the second node feature and the second hyperedge; performing hypergraph convolution on the first hypergraph and the second hypergraph respectively to obtain a first feature and a second feature; inputting the first feature and the second feature into a bird assessment model, and determining the bird monitoring result according to the output value. The present application can accurately monitor and assess bird conditions or bird strike events, thereby improving the efficiency and accuracy of bird repelling while ensuring efficient use of resources.

[0115] It should be understood that although Figure 1 The steps in the flowchart are shown in sequence as indicated by the arrows, but these steps are not necessarily executed in the order indicated by the arrows. Unless otherwise specified in this document, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. Moreover, Figure 1 At least part of the steps may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed in turn or alternately with other steps or at least part of the sub-steps or stages of other steps.

[0116] In one embodiment, Figure 2As shown, a bird monitoring system based on radar data is provided, including: an information acquisition module, a first hypergraph construction module, a second hypergraph construction module, a hypergraph convolution module and a bird assessment module, wherein:

[0117] An information acquisition module, used to acquire historical multi-dimensional data, wherein the historical multi-dimensional data at least includes first time series information of radar and second time series information of flying birds, wherein the time nodes of the first time series information and the second time series information correspond one to one, and the time series information includes spatial dimension information and time dimension information;

[0118] A first hypergraph construction module, configured to initialize and generate a first node feature according to the first time series information, generate a first hyperedge based on the first node feature, and generate a first hypergraph based on the node corresponding to the first node feature and the first hyperedge;

[0119] A second hypergraph construction module, configured to initialize and generate a second node feature according to the second timing information, generate a second hyperedge based on the second node feature, and generate a second hypergraph based on the node corresponding to the second node feature and the second hyperedge;

[0120] A hypergraph convolution module, used to perform hypergraph convolution on the first hypergraph and the second hypergraph respectively to obtain a first feature and a second feature;

[0121] The bird situation assessment module is used to input the first feature and the second feature into the bird situation assessment model, and determine the bird situation monitoring result according to the output value.

[0122] As a preferred implementation, in the embodiment of the present invention, the information acquisition module is specifically used for:

[0123] The historical multi-dimensional data is cleansed, and based on the data cleaning result, the first time series information and the second time series information are classified and stored.

[0124] As a preferred implementation manner, in an embodiment of the present invention, the first hypergraph construction module is specifically used for:

[0125] The radar position information within the target time period is defined as a first node, and the target first node is initialized to obtain a first node feature, wherein the first time series information at least includes the radar position information;

[0126] Based on the objective function, a first likelihood value between the target first node feature and other node features is calculated, wherein the objective function includes:

[0127]

[0128] in, represents the likelihood value, represents the first node feature of the target, Represents the dimension, Indicates other node features;

[0129] Sort the multiple first likelihood values ​​in descending order, and select the node corresponding to the first likelihood value of the number of front targets to generate a first hyperedge;

[0130] A first hypergraph is generated based on the nodes corresponding to the target first node feature and the first hyperedge.

[0131] As a preferred implementation manner, in an embodiment of the present invention, the second hypergraph construction module is specifically used for:

[0132] The flying bird position information within the target time period is defined as a second node, and the target second node is initialized to obtain a second node feature, wherein the first time series information at least includes the flying bird position information;

[0133] Based on the objective function, calculating a second likelihood value between the target second node feature and other node features;

[0134] Sort the multiple second likelihood numbers in descending order, and select the nodes corresponding to the second likelihood values ​​of the previous target number to generate a second hyperedge;

[0135] A second hypergraph is generated based on the nodes corresponding to the target second node feature and the second hyperedge.

[0136] As a preferred implementation, in the embodiment of the present invention, the hypergraph convolution module is specifically used for:

[0137] Determine the degree of vertices in a hypergraph and the degree of the hyperedge ,in, Represents a vertex The degree, Represents a hyperedge The degree, represents the number of vertices, represents the number of hyperedges, represents the adjacency matrix, represents the weight coefficient;

[0138] According to the degree of the node and the degree of the hyperedge to which it belongs, a hypergraph convolution is performed on the hypergraph, and the formula of the hypergraph convolution is:

[0139]

[0140] in, represents the degree matrix of the hyperedge, represents the degree matrix of the vertex, represents the weight matrix, represents the adjacency matrix, Indicates The vertex feature relationship of the layer, represents the learnable parameters, Indicates Vertex feature relationships of layers;

[0141] Fuse multiple vertex feature relationships to obtain a fused representation of the hypergraph , and define the fusion representation as the features after hypergraph convolution.

[0142] As a preferred implementation, in the embodiment of the present invention, the bird situation assessment module is specifically used for:

[0143] Construct a bird situation assessment module, the bird situation assessment module includes:

[0144]

[0145] in, represents the evaluation value, represents the first eigenvalue, represents the second eigenvalue, Represents the fusion variable between spatial dimension information and temporal dimension information, represents the number of features, represents the aggregate feature value, represents a learnable parameter.

[0146] As a preferred implementation, in the embodiment of the present invention, the bird situation assessment module is further used to:

[0147] Obtaining a first feature and a second feature;

[0148] Inputting the first feature and the second feature into a bird situation assessment model to obtain an output value;

[0149] In response to detecting that the output value is less than a preset threshold, a bird situation warning message is sent to the user terminal.

[0150] For the specific definition of the bird monitoring system based on radar data, please refer to the definition of the bird monitoring method based on radar data above, which will not be repeated here. Each module in the above-mentioned bird monitoring system based on radar data can be implemented in whole or in part by software, hardware and a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in the form of hardware, or can be stored in the memory of the computer device in the form of software, so that the processor can call and execute the operations corresponding to the above modules.

[0151] In one embodiment, a computer device is provided. The computer device may be a terminal, and its internal structure diagram may be as follows: Figure 3 As shown. The computer device includes a processor, a memory, a network interface, a display screen and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a bird monitoring method based on radar data is implemented. The display screen of the computer device can be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device can be a touch layer covering the display screen, or a key, trackball or touchpad set on the computer device shell, or an external keyboard, touchpad or mouse, etc.

[0152] Those skilled in the art will understand that Figure 3 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0153] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the following steps when executing the computer program:

[0154] S1: Acquire historical multidimensional data, where the historical multidimensional data at least includes first time series information of radar and second time series information of flying birds, wherein the time nodes of the first time series information and the second time series information correspond one to one, and the time series information includes spatial dimension information and time dimension information;

[0155] S2: Initializing and generating a first node feature according to the first time series information, generating a first hyperedge based on the first node feature, and generating a first hypergraph based on the nodes corresponding to the first node feature and the first hyperedge;

[0156] S3: Initialize and generate a second node feature according to the second time series information, generate a second hyperedge based on the second node feature, and generate a second hypergraph based on the nodes corresponding to the second node feature and the second hyperedge;

[0157] S4: performing hypergraph convolution on the first hypergraph and the second hypergraph respectively to obtain a first feature and a second feature;

[0158] S5: Input the first feature and the second feature into a bird situation assessment model, and determine the bird situation monitoring result according to the output value.

[0159] In one embodiment, a computer readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:

[0160] S1: Acquire historical multidimensional data, where the historical multidimensional data at least includes first time series information of radar and second time series information of flying birds, wherein the time nodes of the first time series information and the second time series information correspond one to one, and the time series information includes spatial dimension information and time dimension information;

[0161] S2: Initializing and generating a first node feature according to the first time series information, generating a first hyperedge based on the first node feature, and generating a first hypergraph based on the nodes corresponding to the first node feature and the first hyperedge;

[0162] S3: Initialize and generate a second node feature according to the second time series information, generate a second hyperedge based on the second node feature, and generate a second hypergraph based on the nodes corresponding to the second node feature and the second hyperedge;

[0163] S4: performing hypergraph convolution on the first hypergraph and the second hypergraph respectively to obtain a first feature and a second feature;

[0164] S5: Input the first feature and the second feature into a bird situation assessment model, and determine the bird situation monitoring result according to the output value.

[0165] In one embodiment, a computer program product is provided, the computer program product comprising a computer program, the computer program when executed by a processor implements the following steps:

[0166] S1: Acquire historical multidimensional data, where the historical multidimensional data at least includes first time series information of radar and second time series information of flying birds, wherein the time nodes of the first time series information and the second time series information correspond one to one, and the time series information includes spatial dimension information and time dimension information;

[0167] S2: Initializing and generating a first node feature according to the first time series information, generating a first hyperedge based on the first node feature, and generating a first hypergraph based on the nodes corresponding to the first node feature and the first hyperedge;

[0168] S3: Initialize and generate a second node feature according to the second time series information, generate a second hyperedge based on the second node feature, and generate a second hypergraph based on the nodes corresponding to the second node feature and the second hyperedge;

[0169] S4: performing hypergraph convolution on the first hypergraph and the second hypergraph respectively to obtain a first feature and a second feature;

[0170] S5: Input the first feature and the second feature into a bird situation assessment model, and determine the bird situation monitoring result according to the output value.

[0171] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0172] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0173] The above-mentioned embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the invention patent. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application.

Claims

1. A bird monitoring method based on radar data, characterized in that: The method comprises: Acquire historical multidimensional data, where the historical multidimensional data includes at least first time series information of the radar and second time series information of the bird, wherein the time nodes of the first time series information and the second time series information correspond one to one, and the time series information includes spatial dimension information and time dimension information; Initialize and generate a first node feature according to the first time series information, generate a first hyperedge based on the first node feature, and generate a first hypergraph based on the node corresponding to the first node feature and the first hyperedge; Initialize and generate a second node feature according to the second time series information, generate a second hyperedge based on the second node feature, and generate a second hypergraph based on the nodes corresponding to the second node feature and the second hyperedge; Performing hypergraph convolution on the first hypergraph and the second hypergraph respectively to obtain a first feature and a second feature; Inputting the first feature and the second feature into a bird situation assessment model, and determining a bird situation monitoring result according to the output value; The first time series information includes at least radar position information, and initializing a first node feature according to the first time series information, generating a first hyperedge based on the first node feature, and generating a first hypergraph based on the node corresponding to the first node feature and the first hyperedge includes: The radar position information within the target time period is defined as a first node, and the target first node is initialized to obtain a first node feature; Based on the objective function, a first likelihood value between the target first node feature and other node features is calculated, wherein the objective function includes: in, represents the likelihood value, represents the first node feature of the target, Represents the dimension, Indicates other node features; Sort the multiple first likelihood values ​​in descending order, and select the node corresponding to the first likelihood value of the number of front targets to generate a first hyperedge; Generate a first hypergraph based on the node corresponding to the target first node feature and the first hyperedge; The second time series information includes at least the bird position information, initializing the second node feature according to the second time series information, generating a second hyperedge based on the second node feature, and generating a second hypergraph based on the node corresponding to the second node feature and the second hyperedge, including: The flying bird position information within the target time period is defined as a second node, and the target second node is initialized to obtain the second node feature; Based on the objective function, calculating a second likelihood value between the target second node feature and other node features; Sort the multiple second likelihood numbers in descending order, and select the nodes corresponding to the second likelihood values ​​of the previous target number to generate a second hyperedge; A second hypergraph is generated based on the nodes corresponding to the target second node feature and the second hyperedge.

2. The bird monitoring method based on radar data according to claim 1 is characterized in that: After acquiring the historical multidimensional data, the method further includes: The historical multi-dimensional data is cleansed, and based on the data cleaning result, the first time series information and the second time series information are classified and stored.

3. The bird monitoring method based on radar data according to claim 2 is characterized in that: Perform hypergraph convolution on the hypergraph, and obtain the corresponding features including: Determine the degree of vertices in a hypergraph and the degree of the hyperedge ,in, Represents a vertex The degree, Represents a hyperedge The degree, represents the number of vertices, represents the number of hyperedges, represents the adjacency matrix, represents the weight coefficient; According to the degree of the node and the degree of the hyperedge to which it belongs, a hypergraph convolution is performed on the hypergraph, and the formula of the hypergraph convolution is: in, represents the degree matrix of the hyperedge, represents the degree matrix of the vertex, represents the weight matrix, represents the adjacency matrix, Indicates The vertex feature relationship of the layer, represents the learnable parameters, Indicates Vertex feature relationships of layers; Fuse multiple vertex feature relationships to obtain a fused representation of the hypergraph , and define the fusion representation as the features after hypergraph convolution.

4. The bird monitoring method based on radar data according to claim 3 is characterized in that: The bird situation assessment model includes: in, represents the evaluation value, represents the first eigenvalue, represents the second eigenvalue, Represents the fusion variable between spatial dimension information and temporal dimension information, represents the number of features, represents the aggregate feature value, represents a learnable parameter.

5. The bird monitoring method based on radar data according to claim 4 is characterized in that: Inputting the first feature and the second feature into the bird situation assessment model, and determining the bird situation monitoring result according to the output value includes: Obtaining a first feature and a second feature; Inputting the first feature and the second feature into a bird situation assessment model to obtain an output value; In response to detecting that the output value is less than a preset threshold, a bird situation warning message is sent to the user terminal.

6. A bird monitoring system based on radar data for implementing the bird monitoring method based on radar data as claimed in any one of claims 1 to 5, characterized in that: The system comprises: An information acquisition module, used to acquire historical multi-dimensional data, wherein the historical multi-dimensional data at least includes first time series information of radar and second time series information of flying birds, wherein the time nodes of the first time series information and the second time series information correspond one to one, and the time series information includes spatial dimension information and time dimension information; A first hypergraph construction module, configured to initialize and generate a first node feature according to the first time series information, generate a first hyperedge based on the first node feature, and generate a first hypergraph based on the node corresponding to the first node feature and the first hyperedge; A second hypergraph construction module, configured to initialize and generate a second node feature according to the second timing information, generate a second hyperedge based on the second node feature, and generate a second hypergraph based on the node corresponding to the second node feature and the second hyperedge; A hypergraph convolution module, used to perform hypergraph convolution on the first hypergraph and the second hypergraph respectively to obtain a first feature and a second feature; The bird situation assessment module is used to input the first feature and the second feature into the bird situation assessment model, and determine the bird situation monitoring result according to the output value.

Images