A method and system for transmitting data of power transmission lines in a signal-free area based on a wireless bridge
By collecting and processing meteorological, regional and forest image data, and using convolutional neural networks to dynamically plan wireless bridge links, the problem of frequent interruptions of wireless communication networks in thunderstorms is solved, and the quality of data transmission is improved.
Patent Information
- Application Number
- CN202411320160.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-23
- Publication Date
- 2024-11-19
- Estimated Expiration
- 2044-09-23
AI Technical Summary
In thunderstorms, lightning current interference causes wireless communication networks to frequently switch links, causing network interruptions and affecting data transmission quality.
By collecting meteorological images, regional images and forest top view images, combining communication network data, the link between wireless bridges is dynamically planned using convolutional neural networks to avoid lightning current interference and reduce link switching frequency.
Dynamically plan wireless communication network links in thunderstorms to reduce network interruptions and improve data transmission quality.
Smart Images

Figure CN118843159B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of communication technology, and more specifically, to a method and system for transmitting data of a power transmission line in a signal-free area based on a wireless bridge. Background Art
[0002] In order to ensure the safe operation of the transmission line, multiple data collection nodes are installed along the transmission line. Each node is equipped with a variety of sensors to monitor key parameters such as temperature, vibration, and line loss in real time. These data collection nodes establish a communication link with the nearest bridge node through a wireless bridge, and finally converge to the central control room, which is equipped with redundant wireless bridges. When the path delay of a wireless bridge increases to a certain threshold, it automatically switches to a wireless bridge with a low path delay that can be connected.
[0003] In a smart grid project in a mountainous area, there are more than 50 days of thunderstorms each year. During thunderstorms, in areas with dense trees, lightning currents encounter resistance along the tree trunks and generate induced currents, forming local electromagnetic noise sources that interfere with radio transmission and reception. This interference only affects local areas, but the impact time is short, generally within a few minutes;
[0004] When switching the links between wireless bridges by the aforementioned method, there is a situation where the target wireless bridge is interfered by lightning current after the switching is completed. At this time, it is necessary to switch again. In thunderstorm weather, frequent link switching will occur in the wireless communication network, and each switching takes a certain amount of time, causing network interruption, resulting in frequent interruption of wireless communication, affecting the quality of data transmission. Summary of the invention
[0005] The present invention uses a data transmission method for power transmission lines in a signal-free area based on a wireless bridge to avoid signal interference from lightning current as much as possible when switching links, and to reduce the frequency of switching links as much as possible.
[0006] The present invention provides a method for transmitting data of a power transmission line in a signal-free area based on a wireless bridge, comprising the following steps:
[0007] Step 100, collecting meteorological images, regional images, forest bird's-eye view images, and communication network data of the coverage area of the wireless communication network composed of wireless bridges;
[0008] Step 200, pre-processing the collected meteorological images, regional images, forest overhead images, and communication network data and inputting them into a first neural network, outputting a value representing the selected action;
[0009] Step 300: Obtain the link between the wireless bridges as the target link state according to the selected action decoding, and adjust the link between the wireless bridges to the target link state.
[0010] In a preferred embodiment, the meteorological image includes: information on the distribution of thunderstorms, the direction of movement of thunderstorms, and the intensity of thunderstorms;
[0011] The collection is done at fixed time intervals;
[0012] Regional images include: terrain information, tree density, and altitude information;
[0013] Forest bird's-eye view images include information on vegetation cover type, range, and density.
[0014] Meteorological images are weather radar images obtained by integrating satellite meteorological data through radar technology;
[0015] Regional images are high-resolution satellite images obtained through remote sensing satellites;
[0016] The forest top-view images are high-resolution images obtained through drone aerial photography;
[0017] The communication network data includes: communication network topology map, wireless bridge nodes, geographical location coordinates, bandwidth, delay, and signal strength corresponding to the wireless bridge nodes.
[0018] In a preferred embodiment, a wireless bridge node corresponds to a wireless bridge, the edges between wireless bridge nodes represent the links between the wireless bridges, and the node characteristics of the wireless bridge node are obtained by encoding the geographic location coordinates, bandwidth, delay, and signal strength corresponding to the wireless bridge node.
[0019] In a preferred embodiment, the first convolution module inputs a meteorological image and outputs meteorological features;
[0020] The second convolution module inputs the regional image and outputs the regional features;
[0021] The third convolution module inputs the forest top view image and outputs the forest features;
[0022] The graph pattern recognition module inputs communication network data and outputs graph coding features;
[0023] The feature combination module is used to splice meteorological features, regional features, forest features, and map coding features to form spliced features. The feature recognition module inputs the spliced features and outputs hidden features. The hidden features are input into the output module, and the output module outputs the value representing the selected action.
[0024] In a preferred embodiment, the calculation formula of the graph pattern recognition module is as follows:
[0025] ;
[0026] ;
[0027] Where D represents the input feature matrix, and a row vector of the input feature matrix represents the node feature of a wireless bridge node. represents the node encoding feature matrix, represents the transformation matrix, represents the sum of the adjacency matrix and the identity matrix, express The degree matrix of represents the trainable parameters, express Activation function;
[0028] The adjacency matrix indicates whether there is an edge between the wireless bridge nodes. If there is an edge between the i-th and j-th wireless bridge nodes, the element in the i-th row and j-th column of the adjacency matrix is 1, otherwise it is 0.
[0029] In a preferred embodiment, the calculation formula of the feature combination module is as follows:
[0030] ;
[0031] in , , They represent meteorological characteristics, regional characteristics and forest characteristics respectively, and X represents the combined characteristics.
[0032] In a preferred embodiment, the calculation formula of the feature recognition module is as follows:
[0033] ;
[0034] ;
[0035] ;
[0036] ;
[0037] ;
[0038] in, represents the initial output features, and They represent the 1st and Lth combined features of the input respectively, n≥L≥1, the Lth combined feature is generated by the data collected at the Lth moment, and the nth moment is the current moment. , and Both represent trainable weight parameters, , and and both represent trainable bias parameters, represents the dot product, Indicates The first intermediate feature, Indicates The second intermediate feature, Indicates A third intermediate feature, and Respectively represent and output features, tanh represents the tanh function, Represents the Sigmoid function.
[0039] In a preferred embodiment, the action is selected from the action space, which includes all selectable actions. An action is represented by an action matrix. If the element in the i-th row and j-th column of the action matrix is 1, it means that a link is established between the i-th and j-th wireless bridge nodes. If the element is 0, it means that no link is established.
[0040] The present invention provides a data transmission system for power transmission lines in a signal-free area based on a wireless bridge, comprising:
[0041] An information collection module collects meteorological images, regional images, forest bird's-eye views, and communication network data of the coverage area of the wireless communication network composed of wireless bridges;
[0042] The data processing module pre-processes the collected meteorological images, regional images, forest overhead images, and communication network data and inputs them into the first neural network to represent the value of the selected action;
[0043] The link updating module obtains the link between the wireless bridges as the target link state according to the selected action decoding, and adjusts the link between the wireless bridges to the target link state.
[0044] The beneficial effects of the present invention are:
[0045] The present invention can dynamically plan the links of the wireless communication network composed of wireless bridges according to meteorological information and forest image information in thunderstorm weather, so as to avoid frequent interruption of wireless communication and influence on data transmission quality. BRIEF DESCRIPTION OF THE DRAWINGS
[0046] Figure 1 It is a flow chart of a method for transmitting data of a power transmission line in a signal-free area based on a wireless bridge of the present invention; DETAILED DESCRIPTION
[0047] The subject matter described herein will now be discussed with reference to example implementations. It should be understood that the discussion of these implementations is only to enable those skilled in the art to better understand and implement the subject matter described herein, and the functions and arrangements of the elements discussed may be changed without departing from the scope of protection of the present specification. Various examples may omit, replace, or add various processes or components as needed. In addition, the features described in some examples may also be combined in other examples.
[0048] At least one embodiment of the present invention discloses a method for transmitting data through a power transmission line in a signal-free area based on a wireless bridge, such as Figure 1 As shown, the following steps are included:
[0049] Step 100, collecting meteorological images, regional images, forest bird's-eye view images, and communication network data of the coverage area of the wireless communication network composed of wireless bridges;
[0050] The data is collected at fixed time intervals.
[0051] Meteorological images include: information on the distribution of thunderstorms, their movement direction and intensity;
[0052] Regional images include: terrain information, tree density, and altitude information;
[0053] Forest bird's-eye view images include information on vegetation cover type, range, and density.
[0054] Meteorological images are weather radar images obtained by integrating satellite meteorological data through radar technology;
[0055] Regional images are high-resolution satellite images obtained through remote sensing satellites;
[0056] The forest top-view images are high-resolution images obtained through drone aerial photography;
[0057] The communication network data includes: communication network topology map, wireless bridge nodes, geographical location coordinates, bandwidth, delay, and signal strength corresponding to the wireless bridge nodes.
[0058] A wireless bridge node corresponds to a wireless bridge, the edges between wireless bridge nodes represent the links between wireless bridges, and the node characteristics of the wireless bridge node are obtained by encoding the geographical location coordinates, bandwidth, delay, and signal strength corresponding to the wireless bridge node.
[0059] The node features can be encoded using general text encoding methods.
[0060] Communication network topology graph G=(V,E), where V represents the set of wireless bridge nodes and E represents the set of edges;
[0061] Step 200, pre-processing the collected meteorological images, regional images, forest overhead images, and communication network data and inputting them into a first neural network, outputting a value representing a selected action, the first neural network comprising a first convolution module, a second convolution module, a third convolution module, a graph pattern recognition module, a feature combination module, a feature recognition module, and an output module;
[0062] The first convolution module inputs the meteorological image and outputs the meteorological features;
[0063] The second convolution module inputs the regional image and outputs the regional features;
[0064] The third convolution module inputs the forest top view image and outputs the forest features;
[0065] The graph pattern recognition module inputs communication network data and outputs graph coding features;
[0066] The feature combination module is used to splice meteorological features, regional features, forest features, and map coding features to form spliced features. The feature recognition module inputs the spliced features and outputs hidden features. The hidden features are input into the output module, and the output module outputs the value representing the selected action.
[0067] The action is selected from the action space, which includes all selectable actions. An action is represented by an action matrix. If the element in the i-th row and j-th column of the action matrix is 1, it means that a link is established between the i-th and j-th wireless bridge nodes. If the element is 0, it means that no link is established.
[0068] In one embodiment of the present invention, the first convolution module, the second convolution module and the third convolution module all adopt convolutional neural networks, such as VGG (Visual Geometry Group).
[0069] In one embodiment of the present invention, the calculation formula of the graph pattern recognition module is as follows:
[0070] ;
[0071] ;
[0072] Where D represents the input feature matrix, and a row vector of the input feature matrix represents the node feature of a wireless bridge node. represents the node encoding feature matrix, represents the transformation matrix, represents the sum of the adjacency matrix and the identity matrix, express The degree matrix of represents the trainable parameters, express Activation function;
[0073] The adjacency matrix indicates whether there is an edge between the wireless bridge nodes. If there is an edge between the i-th and j-th wireless bridge nodes, the element in the i-th row and j-th column of the adjacency matrix is 1, otherwise it is 0.
[0074] In one embodiment of the present invention, the calculation formula of the feature combination module is as follows:
[0075] ;
[0076] in , , They represent meteorological characteristics, regional characteristics and forest characteristics respectively, and X represents the combined characteristics.
[0077] The calculation formula of the feature recognition module is as follows:
[0078] ;
[0079] ;
[0080] ;
[0081] ;
[0082] ;
[0083] in, represents the initial output features, and They represent the 1st and Lth combined features of the input respectively, n≥L≥1, the Lth combined feature is generated by the data collected at the Lth moment, and the nth moment is the current moment. , and Both represent trainable weight parameters, , and Both represent trainable bias parameters, represents the dot product, Indicates The first intermediate feature, Indicates The second intermediate feature, Indicates A third intermediate feature, and Respectively represent and output features, tanh represents the tanh function, Represents the Sigmoid function.
[0084] The calculation formula for the output module is as follows:
[0085] ;
[0086] represents the nth output feature of the feature recognition module, represents the trainable weight parameters, represents the trainable bias parameter, Represents the output vector, and the value of the i-th component of the output vector represents the Q value of the i-th action.
[0087] The action corresponding to the value of the largest component of the output vector is selected as the selected action.
[0088] Step 300: Obtain the link between the wireless bridges as the target link state according to the selected action decoding, and adjust the link between the wireless bridges to the target link state.
[0089] In one embodiment of the present invention, the first neural network is trained using a Q learning method, and the state of the Q learning is defined as ten consecutively collected meteorological images, regional images, forest overhead images, and communication network data of the coverage area of the wireless communication network composed of wireless bridges;
[0090] In each round of training, the current state is input, and the first neural network selects an action from the action space;
[0091] Execute an action, move to the next state and receive a reward;
[0092] The reward function for Q learning is:
[0093]
[0094] in represents the reward value after executing the selected action at time k, is the average signal strength over a period of time after the action is performed, is the maximum possible value of the signal strength, is the average bandwidth over a period of time after the action is executed, is the maximum possible value of bandwidth, is the average delay over a period of time after the action is executed, is the maximum possible value of the delay, is the number of link switches within a period of time after the action is executed, is the maximum possible number of switching times, , , , Corresponding to each weight coefficient respectively, the sum of the four weight coefficients is 1. The default values are 0.2, 0.2, 0.2, and 0.4.
[0095] A period of time after an action is performed is the period of time between two moments in time.
[0096] At least one embodiment of the present invention discloses a wireless bridge-based power transmission line data transmission system in a signal-free area, comprising:
[0097] An information collection module collects meteorological images, regional images, forest bird's-eye views, and communication network data of the coverage area of the wireless communication network composed of wireless bridges;
[0098] The data processing module pre-processes the collected meteorological images, regional images, forest overhead images, and communication network data and inputs them into the first neural network to represent the value of the selected action;
[0099] The link updating module obtains the link between the wireless bridges as the target link state according to the selected action decoding, and adjusts the link between the wireless bridges to the target link state.
[0100] The above describes an embodiment of the present invention, but this embodiment is not limited to the above-mentioned specific implementation mode. The above-mentioned specific implementation mode is merely illustrative and not restrictive. Under the guidance of this embodiment, ordinary technicians in this field can also make more forms of equivalent embodiments, all of which are within the protection of this embodiment.
Claims
1. A method for transmitting data in a power transmission line in a signal-free area based on a wireless bridge, characterized in that: The following steps are involved: Step 100, collecting meteorological images, regional images, forest bird's-eye view images, and communication network data of the coverage area of the wireless communication network composed of wireless bridges; Step 200, pre-process the collected meteorological image, regional image, forest bird's-eye view image, and communication network data and input them into the first neural network, and output the value representing the selected action; the first neural network includes a first convolution module, a second convolution module, a third convolution module, a graph pattern recognition module, a feature combination module, a feature recognition module, and an output module; the first convolution module inputs the meteorological image and outputs the meteorological feature; the second convolution module inputs the regional image and outputs the regional feature; the third convolution module inputs the forest bird's-eye view image and outputs the forest feature; the graph pattern recognition module inputs the communication network data and outputs the graph coding feature; the feature The combination module is used to combine meteorological features, regional features, forest features, and image coding features to form a spliced feature. The feature recognition module inputs the spliced feature and outputs the hidden feature. The hidden feature is input into the output module, and the output module outputs the value representing the selected action. The calculation formula of the image pattern recognition module is as follows: ; ; Where D represents the input feature matrix, and a row vector of the input feature matrix represents the node feature of a wireless bridge node. represents the node encoding feature matrix, represents the transformation matrix, represents the sum of the adjacency matrix and the identity matrix, express The degree matrix of represents the trainable parameters, express Activation function; the adjacency matrix indicates whether there is an edge between the wireless bridge nodes. If there is an edge between the i-th and j-th wireless bridge nodes, the element in the i-th row and j-th column of the adjacency matrix is 1, otherwise it is 0; the calculation formula of the feature combination module is as follows: ;in , , represent meteorological characteristics, regional characteristics and forest characteristics respectively, and X represents the combined characteristics; The calculation formula of the recognition module is as follows: ; ; ; ; ;in, represents the initial output features, and They represent the 1st and Lth combined features of the input respectively, n≥L≥1, the Lth combined feature is generated by the data collected at the Lth moment, and the nth moment is the current moment. , and Both represent trainable weight parameters, , and Both represent trainable bias parameters, represents the dot product, Indicates The first intermediate feature, Indicates The second intermediate feature, Indicates A third intermediate feature, and Respectively represent and output features, tanh represents the tanh function, Represents the Sigmoid function; the calculation formula of the output module is as follows: ; represents the nth output feature of the feature recognition module, represents the trainable weight parameters, represents the trainable bias parameter, represents an output vector, and the value of the i-th component of the output vector represents the Q value of the i-th action; step 300, decoding the link between the wireless bridges according to the selected action as the target link state, and adjusting the link between the wireless bridges to the target link state.
2. According to claim 1, a method for transmitting data in a power transmission line in a signal-free area based on a wireless bridge, characterized in that: Meteorological images include: information on the distribution of thunderstorms, the direction of movement and intensity of thunderstorms; data is collected at fixed time intervals; regional images include: information on terrain, tree density, and altitude; forest bird's-eye view images include: information on vegetation coverage type, range, and density; communication network data include: communication network topology, wireless bridge nodes, geographical location coordinates corresponding to wireless bridge nodes, bandwidth, delay, and signal strength.
3. A method for transmitting data in a power transmission line in a signal-free area based on a wireless bridge according to claim 2, characterized in that: A wireless bridge node corresponds to a wireless bridge, the edges between wireless bridge nodes represent the links between wireless bridges, and the node characteristics of the wireless bridge node are obtained by encoding the geographical location coordinates, bandwidth, delay, and signal strength corresponding to the wireless bridge node.
4. The method for transmitting data in a power transmission line in a signal-free area based on a wireless bridge according to claim 1, characterized in that: The action is selected from the action space, which includes all selectable actions. An action is represented by an action matrix. If the element in the i-th row and j-th column of the action matrix is 1, it means that a link is established between the i-th and j-th wireless bridge nodes. If the element is 0, it means that no link is established.
5. A data transmission system for power transmission lines in a signal-free area based on a wireless bridge, characterized in that: include: An information collection module collects meteorological images, regional images, forest bird's-eye views, and communication network data of the coverage area of the wireless communication network composed of wireless bridges; The data processing module pre-processes the collected meteorological images, regional images, forest overhead images, and communication network data and inputs them into the first neural network to represent the value of the selected action; the first neural network includes a first convolution module, a second convolution module, a third convolution module, a graph pattern recognition module, a feature combination module, a feature recognition module, and an output module; the first convolution module inputs the meteorological image and outputs the meteorological feature; the second convolution module inputs the regional image and outputs the regional feature; the third convolution module inputs the forest overhead image and outputs the forest feature; the graph pattern recognition module inputs the communication network data and outputs the graph coding feature; the feature The combination module is used to combine meteorological features, regional features, forest features, and image coding features to form a spliced feature. The feature recognition module inputs the spliced feature and outputs the hidden feature. The hidden feature is input into the output module, and the output module outputs the value representing the selected action. The calculation formula of the image pattern recognition module is as follows: ; ; Where D represents the input feature matrix, and a row vector of the input feature matrix represents the node feature of a wireless bridge node. represents the node encoding feature matrix, represents the transformation matrix, represents the sum of the adjacency matrix and the identity matrix, express The degree matrix of represents the trainable parameters, express Activation function; the adjacency matrix indicates whether there is an edge between the wireless bridge nodes. If there is an edge between the i-th and j-th wireless bridge nodes, the element in the i-th row and j-th column of the adjacency matrix is 1, otherwise it is 0; the calculation formula of the feature combination module is as follows: ;in , , represent meteorological characteristics, regional characteristics and forest characteristics respectively, and X represents the combined characteristics; The calculation formula of the recognition module is as follows: ; ; ; ; ;in, represents the initial output features, and They represent the 1st and Lth combined features of the input respectively, n≥L≥1, the Lth combined feature is generated by the data collected at the Lth moment, and the nth moment is the current moment. , and Both represent trainable weight parameters, , and Both represent trainable bias parameters, represents the dot product, Indicates The first intermediate feature, Indicates The second intermediate feature, Indicates A third intermediate feature, and Respectively represent and output features, tanh represents the tanh function, Represents the Sigmoid function; the calculation formula of the output module is as follows: ; represents the nth output feature of the feature recognition module, represents the trainable weight parameters, represents the trainable bias parameter, represents the output vector, and the value of the i-th component of the output vector represents the Q value of the i-th action; The link updating module obtains the link between the wireless bridges as the target link state according to the selected action decoding, and adjusts the link between the wireless bridges to the target link state.
