Space and risk determination method and device for hidden danger object of distribution network line

Abstract

The invention provides a space and risk determination method and device for hidden danger objects of a distribution network line, and relates to the technical field of distribution network line management and control. The method comprises the steps of obtaining a distribution network line and multi-modal data of a hidden danger object which forms potential threats to the distribution network line, performing feature extraction on the multi-modal data, and determining multi-dimensional features of the hidden danger object. And then, inputting the multi-dimensional features into a multi-task fusion model, and predicting an event category to which the hidden danger object belongs based on the multi-task fusion model. And next, obtaining geographic position data of the hidden danger object, and mapping the geographic position data and the belonging event category to a target GIS platform to determine the geographic distribution condition of the hidden danger object. And finally, determining a risk score of the hidden danger object based on the event category and the geographical distribution condition, and determining a risk level of the hidden danger object based on the risk score. The geographic distribution condition and the risk level of the hidden danger object can be accurately determined.

Description

Technical Field

[0001] The present application relates to the technical field of distribution network line management and control, and in particular to a method and device for determining the space and risk of hidden danger objects in distribution network lines. Background Art

[0002] Distribution lines are a critical component of the power grid, responsible for transporting electricity from substations to homes, businesses, and industrial users. However, because they are often exposed to the elements, they are susceptible to damage from various external forces, such as construction equipment and trees growing into contact with them. Therefore, it is necessary to analyze potential hazards that threaten the safety and normal operation of distribution lines.

[0003] Currently, two common methods are used to analyze potential hazards on distribution lines: line distance measurement based on laser measurement technology and measurement based on stereo vision. The laser distance measurement method uses a laser radar (LiDAR) to transmit laser pulses to the distribution line and the potential hazard, respectively. The time difference between the laser pulses' emission and return is measured to calculate the distances from the distribution line and the potential hazard to the LiDAR. The LiDAR then processes the calculated distances to determine the positions of the distribution line and the potential hazard in three-dimensional space. This position is then used to determine the risk of the potential hazard. The stereo vision measurement method uses multiple cameras to capture images of the same scene (including the distribution line and the potential hazard) from different angles. The differences between these images are then analyzed to calculate the three-dimensional spatial information of the distribution line and the potential hazard. The risk of the potential hazard is then determined based on the distance and relative position between the distribution line and the potential hazard.

[0004] However, the data on distribution network lines and hidden danger objects obtained by the above two methods may not be accurate enough, and relying solely on the obtained data to calculate the three-dimensional spatial information of distribution network lines and hidden danger objects makes it difficult to accurately locate the positions of distribution network lines and hidden danger objects, which in turn leads to inaccurate risk determination of hidden danger objects. Summary of the Invention

[0005] The present application provides a method and device for determining the space and risk of hidden danger objects for distribution network lines, which is used to solve the technical problems that the existing technology is difficult to accurately locate the positions of distribution network lines and hidden danger objects, and the risk determination of hidden danger objects is not accurate enough.

[0006] In a first aspect, the present application provides a method for determining the space and risk of hidden danger objects on distribution network lines, comprising:

[0007] Acquire multimodal data of a target area; wherein the target area includes a distribution network line and hidden danger objects that pose a potential threat to the distribution network line;

[0008] Performing feature extraction on the multimodal data to determine the multidimensional features of the hidden danger object; inputting the multidimensional features into a pre-built multi-task fusion model, and predicting the event category of the hidden danger object based on the multi-task fusion model; wherein the event category represents an event or activity associated with the hidden danger object;

[0009] Obtaining geographic location data of the hidden danger objects, and mapping the geographic location data and the event categories to which they belong to onto a target geographic information system (GIS) platform to determine the geographic distribution of the hidden danger objects within the target area;

[0010] Based on the event category and geographical distribution of the hidden danger object, a risk score of the hidden danger object is determined, and based on the risk score, a risk level of the hidden danger object is determined.

[0011] In one possible design, extracting features from the multimodal data to determine the multidimensional features of the hidden danger object includes:

[0012] Integrating the multimodal data to form a multidimensional heterogeneous data set;

[0013] The multidimensional heterogeneous data set is preprocessed, and the multidimensional features of the hidden danger object are extracted from the preprocessed multidimensional heterogeneous data set; wherein the multidimensional features include morphological features, spatial distribution features and dynamic behavior features.

[0014] In one possible design, the multidimensional heterogeneous data set includes image data, video data, lidar data, sensor data and GIS data.

[0015] Extracting multidimensional features of the hidden danger object from the preprocessed multidimensional heterogeneous data set includes:

[0016] Using a preset convolutional neural network, feature extraction is performed on the image data and the video data to obtain morphological features of the hidden danger object;

[0017] Based on the lidar data and the GIS data, a preset point cloud clustering algorithm is used to calculate the cluster density of the hidden danger objects in the target area; the cluster density is analyzed to determine the distribution and distribution pattern of the hidden danger objects, and based on the distribution and distribution pattern, the spatial distribution characteristics of the hidden danger objects are determined;

[0018] Based on the sensor data, a preset time series analysis algorithm is adopted to determine a moving track of the hidden object, and based on the moving track, a dynamic behavior feature of the hidden object is determined.

[0019] In a possible design, the prediction of the event category to which the hidden object belongs based on the multi-task fusion model comprises:

[0020] Based on the multi-task fusion model, the shape feature, the spatial distribution feature and the dynamic behavior feature are converted into structured shape feature vectors, spatial distribution feature vectors and dynamic behavior feature vectors.

[0021] The shape feature vectors, the spatial distribution feature vectors and the dynamic behavior feature vectors are subjected to weighting processing, and the weighted shape feature vectors, spatial distribution feature vectors and dynamic behavior feature vectors are input into a preset classifier included in the multi-task fusion model.

[0022] Based on the preset classifier, a probability that the hidden object belongs to a preset event category is predicted, all probabilities are sorted in descending order, and a preset event category corresponding to a probability ranked first is determined as the event category to which the hidden object belongs.

[0023] In a possible design, the mapping of the geographic location data and the event category to a target geographic information system (GIS) platform to determine the geographic distribution of the hidden object in the target area comprises:

[0024] The geographic location data and the event category are mapped to a target GIS platform, and the geographic location data and the event category are associated based on the target GIS platform to obtain geographic distribution data of the hidden object in the target area.

[0025] The geographic distribution data is subjected to format processing to generate structured geographic distribution information.

[0026] In a possible design, the multi-dimensional features of the hidden object include shape features, spatial distribution features and dynamic behavior features.

[0027] The determination of the risk score of the hidden object based on the event category to which the hidden object belongs and the geographic distribution of the hidden object comprises:

[0028] Based on the shape features, a first preset calculation formula is adopted to calculate a corresponding shape feature risk factor; based on the spatial distribution features, a second preset calculation formula is adopted to calculate a corresponding spatial distribution feature risk factor; and based on the dynamic behavior features, a third preset calculation formula is adopted to calculate a corresponding dynamic behavior feature risk factor.

[0029] Determining, based on the event category and geographical distribution of the hidden danger object, the contribution values ​​of the morphological characteristic risk factor, the spatial distribution characteristic risk factor, and the dynamic behavior characteristic risk factor to the risk score; assigning a first weight to the morphological characteristic risk factor, a second weight to the spatial distribution characteristic risk factor, and a third weight to the dynamic behavior characteristic risk factor based on the contribution values;

[0030] The risk score of the hidden danger object is obtained by multiplying the morphological characteristic risk factor by the first weight, the spatial distribution characteristic risk factor by the second weight, and the dynamic behavior characteristic risk factor by the third weight and summing the results.

[0031] In one possible design, determining the risk level of the hidden danger object based on the risk score includes:

[0032] If the risk score is less than or equal to a first score threshold, the risk level of the hidden danger object is determined to be a no-risk level;

[0033] If the risk score is greater than the first score threshold and less than or equal to the second score threshold, the risk level of the hidden danger object is determined to be a low risk level;

[0034] If the risk score is greater than the second scoring threshold and less than or equal to the third scoring threshold, the risk level of the hidden danger object is determined to be a medium risk level;

[0035] If the risk score is greater than a third score threshold, the risk level of the hidden danger object is determined to be a high risk level.

[0036] In one possible design, the method further includes:

[0037] If the geographical distribution indicates that the hidden danger object is located at a preset high-risk location within the target area, and / or the risk level indicates that the hidden danger object is risky, a hidden danger object handling request is sent to relevant staff.

[0038] In a second aspect, the present application provides a space and risk determination device for potential hazards of distribution network lines, comprising:

[0039] An acquisition module, configured to acquire multimodal data of a target area, wherein the target area includes distribution network lines and potential objects that pose a threat to the distribution network lines;

[0040] a processing module configured to extract features from the multimodal data to determine multidimensional features of the hidden danger object; input the multidimensional features into a pre-built multi-task fusion model, and predict the event category of the hidden danger object based on the multi-task fusion model; wherein the event category represents an event or activity associated with the hidden danger object;

[0041] The acquisition module is further configured to acquire geographic location data of the hidden danger object;

[0042] The processing module is further configured to map the geographic location data and the event category to a target geographic information system (GIS) platform to determine the geographic distribution of the hidden danger objects within the target area.

[0043] The determination module is configured to determine a risk score of the hidden danger object based on the event category and geographical distribution of the hidden danger object, and to determine a risk level of the hidden danger object based on the risk score.

[0044] In a possible design, the processing module further includes: an integration module, an extraction module,

[0045] The integration module is used to integrate the multimodal data to form a multidimensional heterogeneous data set;

[0046] The processing module is further configured to pre-process the multi-dimensional heterogeneous data set;

[0047] The extraction module is used to extract the multidimensional features of the hidden danger object from the preprocessed multidimensional heterogeneous data set; wherein the multidimensional features include morphological features, spatial distribution features and dynamic behavior features.

[0048] In one possible design, the multidimensional heterogeneous data set includes image data, video data, lidar data, sensor data and GIS data.

[0049] The extraction module is further configured to use a preset convolutional neural network to perform feature extraction on the image data and the video data to obtain morphological features of the hidden danger object;

[0050] The extraction module further includes: a calculation module for calculating the cluster density of the hidden danger objects in the target area using a preset point cloud clustering algorithm based on the lidar data and the GIS data;

[0051] The determination module is further configured to analyze the cluster density to determine the distribution and distribution pattern of the hidden danger objects, and determine the spatial distribution characteristics of the hidden danger objects based on the distribution and distribution pattern;

[0052] The determination module is further configured to determine the movement trajectory of the hidden danger object based on the sensor data and using a preset time series analysis algorithm, and to determine the dynamic behavior characteristics of the hidden danger object based on the movement trajectory.

[0053] In a possible design, the processing module further includes a conversion module, an input module, a prediction module, and a sorting module.

[0054] The conversion module is used to convert the morphological features, the spatial distribution features and the dynamic behavior features into structured morphological feature vectors, spatial distribution feature vectors and dynamic behavior feature vectors based on the multi-task fusion model;

[0055] The processing module is further configured to perform weighted processing on the morphological feature vector, the spatial distribution feature vector, and the dynamic behavior feature vector;

[0056] The input module is used to input the weighted morphological feature vector, spatial distribution feature vector and dynamic behavior feature vector into the preset classifier included in the multi-task fusion model;

[0057] The prediction module is configured to predict the probability that the hidden danger object belongs to a preset event category based on the preset classifier;

[0058] The sorting module is used to sort all probabilities in descending order;

[0059] The determining module is further configured to determine the preset event category corresponding to the probability ranked first as the event category to which the hidden danger object belongs.

[0060] In a possible design, the processing module further includes: a mapping module, an association module,

[0061] The mapping module is used to map the geographic location data and the corresponding event category to a target GIS platform;

[0062] The association module is used to associate the geographical location data with the event category based on the target GIS platform to obtain the geographical distribution data of the hidden danger objects in the target area;

[0063] The processing module is further configured to format the geographic distribution data to generate a structured geographic distribution situation.

[0064] In a possible design, the multi-dimensional characteristics of the hidden danger object include morphological characteristics, spatial distribution characteristics and dynamic behavior characteristics.

[0065] The computing module is further configured to calculate a corresponding morphological feature risk factor by using a first pre-designed calculation formula based on the morphological feature, calculate a corresponding spatial distribution feature risk factor by using a second pre-designed calculation formula based on the spatial distribution feature, and calculate a corresponding dynamic behavior feature risk factor by using a third pre-designed calculation formula based on the dynamic behavior feature.

[0066] The determining module is further configured to determine contribution values of the morphological feature risk factor, the spatial distribution feature risk factor, and the dynamic behavior feature risk factor to a risk score according to the event category to which the hidden danger object belongs and the geographic distribution condition.

[0067] The determining module further includes an assigning module configured to assign a first weight to the morphological feature risk factor, assign a second weight to the spatial distribution feature risk factor, and assign a third weight to the dynamic behavior feature risk factor based on the contribution values.

[0068] The processing module is further configured to multiply the morphological feature risk factor by the first weight, multiply the spatial distribution feature risk factor by the second weight, multiply the dynamic behavior feature risk factor by the third weight, and sum the products to obtain the risk score of the hidden danger object.

[0069] In a possible design, the determining module is further configured to:

[0070] If the risk score is less than or equal to a first score threshold, the risk level of the hidden danger object is determined as a no-risk level.

[0071] If the risk score is greater than the first score threshold and less than or equal to a second score threshold, the risk level of the hidden danger object is determined as a low-risk level.

[0072] If the risk score is greater than the second score threshold and less than or equal to a third score threshold, the risk level of the hidden danger object is determined as a medium-risk level.

[0073] If the risk score is greater than the third score threshold, the risk level of the hidden danger object is determined as a high-risk level.

[0074] In a possible design, the hidden danger object space and risk determination apparatus for a distribution network line further includes a sending module configured to send a hidden danger object processing request to relevant workers if the geographic distribution condition indicates that the hidden danger object is located at a preset high-risk position in the target area and / or the risk level indicates that the hidden danger object is at risk.

[0075] In a third aspect, the present application provides an electronic device comprising: at least one processor and a memory; the memory stores computer-executable instructions; the at least one processor executes the computer-executable instructions stored in the memory, so that the at least one processor executes the method described in the first aspect above and various possible designs.

[0076] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, in which computer-executable instructions are stored. When a processor executes the computer-executable instructions, the method described in the first aspect and various possible designs is implemented.

[0077] In a fifth aspect, an embodiment of the present application provides a computer program product, including a computer program, which, when executed by a processor, implements the method described in the first aspect and various possible designs.

[0078] The present application provides a method and apparatus for spatial and risk identification of hazardous objects for distribution lines. This method acquires multimodal data from a target area, extracts features from the multimodal data, and determines the multidimensional characteristics of the hazardous objects. The target area includes distribution lines and hazardous objects that pose a potential threat to them. By collecting data from multiple sources, it is possible to provide comprehensive information about the hazardous objects, achieving a comprehensive description of the hazardous objects and providing a reliable data foundation for subsequent determination of their geographic distribution. The multidimensional features are then input into a pre-created multi-task fusion model. Based on the multi-task fusion model, the event category to which the hazardous object belongs is predicted. The event category represents the events or activities associated with the hazardous object. Because the multi-task fusion model has powerful recognition capabilities, it can efficiently classify the events associated with the hazardous objects even in complex regional environments. Next, the geographic location data of the hazardous objects is acquired and mapped along with the event category to a target GIS platform to determine their geographic distribution within the target area. By combining the location data and event categories of hazardous objects with the target GIS platform, the geographic distribution of these objects can be visually displayed on a map. The target GIS platform can also precisely locate these objects using geographic coordinates, accurately determining their geographic distribution. Finally, based on the event categories and geographic distribution of these objects, a risk score is assigned to the object, and based on the risk score, the risk level is determined. Assuming the geographic distribution of these objects is accurate, the final risk level is also accurate. BRIEF DESCRIPTION OF THE DRAWINGS

[0079] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0080] Figure 1 A flow chart of a method for determining the space and risk of hidden danger objects in distribution network lines provided in an embodiment of the present application;

[0081] Figure 2 A schematic diagram of the structure of a device for determining the space and risk of hidden danger objects in distribution network lines provided in an embodiment of the present application;

[0082] Figure 3 This is a hardware structure diagram of the electronic device provided in an embodiment of the present application.

[0083] The above drawings illustrate specific embodiments of the present application, which will be described in more detail below. These drawings and the textual description are not intended to limit the scope of the present application in any way, but rather to illustrate the concepts of the present application to those skilled in the art by reference to specific embodiments. DETAILED DESCRIPTION

[0084] Exemplary embodiments will be described in detail herein, with examples illustrated in the accompanying drawings. In the following description, when referring to the drawings, identical numerals in different figures represent identical or similar elements, unless otherwise indicated. The embodiments described in the following exemplary embodiments are not intended to represent all embodiments consistent with the present application. Rather, they are merely examples of apparatus and methods consistent with certain aspects of the present application, as detailed in the appended claims.

[0085] The terms "first," "second," "third," "fourth," and so forth (if any) in the description and claims of the present invention and the accompanying drawings are used to distinguish similar objects and are not necessarily used to describe a particular order or sequential sequence. It should be understood that the terms used in this manner are interchangeable where appropriate, such that the embodiments of the present invention described herein can, for example, be practiced in an order other than that illustrated or described herein.

[0086] In the embodiments of this application, words such as "exemplary" or "for example" are used to indicate examples, illustrations, or descriptions. Any embodiment or design described in this application as "exemplary" or "for example" should not be construed as being preferred or advantageous over other embodiments or designs. Rather, the use of words such as "exemplary" or "for example" is intended to present the relevant concepts in a concrete manner.

[0087] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, stored data, displayed data, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of relevant data must comply with relevant laws, regulations and standards, and corresponding operation entrances must be provided for users to choose to authorize or refuse.

[0088] Distribution lines are a critical component of the power grid system, responsible for transporting electricity from substations to end users, including homes, businesses, and industries. This means that distribution lines are the final link in the power transmission chain, directly impacting users' electricity experience and the reliability of power supply.

[0089] Distribution lines are typically installed outdoors and directly exposed to the elements, making them vulnerable to damage from various external forces. For example, construction activities may accidentally damage distribution lines, and tree growth may cause branches to contact or overwhelm distribution lines.

[0090] Given these potential threats, it's essential to analyze potential hazards that threaten the safety and normal operation of distribution lines. By determining the spatial location and risk level of these hazards, appropriate preventative measures can be taken to ensure the safety and normal operation of distribution lines.

[0091] Currently, there are two common methods for analyzing potential hazards in distribution network lines:

[0092] (1) Line distance measurement method based on laser measurement technology. This method uses a laser radar to transmit laser pulses to the distribution network line and the hidden danger object respectively, and measures the time difference between the laser pulse emission and return, and calculates the distance between the distribution network line and the hidden danger object to the laser radar. The laser radar then processes the calculated distance to obtain the position of the distribution network line and the hidden danger object in three-dimensional space. The position of the distribution network line and the hidden danger object in three-dimensional space is further used to determine the risk of the hidden danger object.

[0093] (2) Stereoscopic vision-based measurement method. This method uses multiple cameras to capture images of the same scene (including distribution network lines and potential hazards) from different angles. The method then calculates the three-dimensional spatial information of the distribution network lines and potential hazards by analyzing the differences between the multiple images. The method then determines the risk of the potential hazards based on the distance and relative position between the distribution network lines and the potential hazards.

[0094] However, in the process of collecting data about distribution network lines and hidden danger objects using the above two methods, the accuracy of the data collected about distribution network lines and hidden danger objects may be insufficient due to factors such as equipment resolution limitations, environmental interference or measurement errors.

[0095] In addition, the above two methods only rely on the acquired data to calculate the three-dimensional spatial information of the distribution network lines and hidden danger objects, and lack a direct connection with the geographic coordinate system, which means that the calculated three-dimensional spatial information of the distribution network lines and hidden danger objects may not be accurately mapped to the real-world geographic location.

[0096] However, inaccurate positioning of distribution network lines and hidden danger objects will lead to the inability to accurately know the distribution of hidden danger objects, which will further lead to inaccurate risk determination of hidden danger objects.

[0097] To address the above technical issues, the inventors considered the issue of insufficient data accuracy when relying solely on a single data source. Based on this, the inventors conceived of integrating data from multiple sources regarding distribution network lines and potential hazards. Even if the data accuracy of a single source is insufficient, data from multiple sources can still provide relatively comprehensive information about distribution network lines and potential hazards. The inventors also considered that the three-dimensional spatial information of distribution network lines and potential hazards obtained solely through data calculations lacks a link to a geographic coordinate system, making it impossible to accurately map it to real-world geographic locations. Based on this, the inventors combined the collected multi-source data with a geographic information system (GIS). Leveraging the visualization and mapping capabilities of GIS, the inventors linked the multi-source data to the geographic coordinate system, accurately locating distribution network lines and potential hazards. Accurately locating distribution network lines and potential hazards allows for accurate understanding of their distribution and the determination of their risks.

[0098] The following specific embodiments describe in detail the technical solution of the present application and how the technical solution of the present application solves the above-mentioned technical problems. The following specific embodiments can be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments. The embodiments of the present application will be described below in conjunction with the accompanying drawings.

[0099] The embodiments of the present application provide a method for determining the space and risk of hidden danger objects in distribution network lines. Figure 1 A flow chart of a method for determining the space and risk of hidden danger objects in distribution network lines provided in an embodiment of the present application is shown as follows: Figure 1 As shown, the method for determining the space and risk of hidden danger objects for distribution network lines includes:

[0100] S101: Acquire multimodal data of a target area.

[0101] The target area refers to the area where the distribution network lines are located. Specifically, the target area includes the distribution network lines and potential hazards that pose a threat to the distribution network lines, such as trees, buildings, and construction facilities.

[0102] The multimodal data of this embodiment includes image data, video data, lidar data, sensor data and GIS data.

[0103] S102: Extract features from the multimodal data to determine the multidimensional features of the hidden danger object; input the multidimensional features into a pre-built multi-task fusion model, and predict the event category to which the hidden danger object belongs based on the multi-task fusion model.

[0104] Understandably, data from different modalities provides different perspectives and information, but they complement each other. By integrating data from multiple modalities, we can leverage the strengths of each modality to improve the accuracy of the overall analysis. Therefore, after acquiring multimodal data for the target area, we first integrate this data to form a multidimensional, heterogeneous dataset. This multidimensional, heterogeneous dataset also includes image data, video data, LiDAR data, sensor data, and GIS data.

[0105] Furthermore, to improve the quality of multidimensional heterogeneous datasets and eliminate the heterogeneity and inconsistency of multimodal data within them, preprocessing is required to obtain standardized datasets in a unified format. Preprocessing includes data denoising, data format conversion, data alignment and synchronization, and data enhancement.

[0106] It's worth noting that LiDAR data is typically represented as point clouds, and the method for data enhancement for point clouds differs from that for data from other modalities. Specifically, this embodiment uses a point cloud density reconstruction method to enhance LiDAR data. For example, a complete point cloud model can be generated using an existing Delaunay triangulation-based reconstruction algorithm. This algorithm is state-of-the-art and will not be described in detail here.

[0107] Furthermore, it's important to note that sensor data, because it involves multiple physical quantities and time series data, typically carries timestamps. Therefore, the approach to data alignment and synchronization for sensor data differs from that for data from other modalities. Specifically, sensor data can be interpolated and aligned based on timestamps. Then, linear interpolation can be used to fill in missing sensor data and remove anomalous sensor data.

[0108] After completing the preprocessing of the multidimensional heterogeneous data set, the multidimensional features of the hidden danger objects can be extracted from the preprocessed multidimensional heterogeneous data set, where the multidimensional features include morphological features, spatial distribution features and dynamic behavior features.

[0109] Specifically, a preset convolutional neural network is used to extract the edge contour features and texture features of hidden danger objects from image data and video data to obtain the morphological features of the hidden danger objects.

[0110] Based on LiDAR and GIS data, a preset point cloud clustering algorithm is used to calculate the density of potential hazards within the target area. Interpretatively, based on the density of potential hazard point cloud data, high-density areas are identified, and the density of potential hazard objects is calculated. High-density areas represent areas where potential hazard objects are concentrated. This density is then analyzed to determine the distribution of potential hazard objects, such as their location and size, as well as their distribution pattern, such as uniform or concentrated. Based on this distribution, the spatial distribution characteristics of the potential hazard objects are determined.

[0111] Based on the sensor data, a preset time series analysis algorithm is used to determine the movement trajectory of the hidden danger object. Based on this movement trajectory, the dynamic behavior characteristics of the hidden danger object are determined. It should be understood that sensor data is typically collected in the form of a time series, recording the state of the hidden danger object at different time points, such as position, velocity, acceleration, etc. Time series analysis algorithms are specifically designed to process this continuous time data and can identify and extract dynamic characteristics from it.

[0112] It should be noted that in order to achieve a comprehensive description of hidden danger objects, the above-mentioned morphological characteristics, spatial distribution characteristics and dynamic behavior characteristics can be integrated into a multidimensional feature space, so that the morphology, spatial distribution and dynamic behavior of hidden danger objects can be comprehensively considered under a unified framework.

[0113] Next, the morphological features, spatial distribution features, and dynamic behavior features are input into the multi-task fusion model. Based on the multi-task fusion model, the morphological features, spatial distribution features, and dynamic behavior features are converted into structured morphological feature vectors, spatial distribution feature vectors, and dynamic behavior vectors. The input layer of the multi-task fusion model includes morphological feature input units, spatial distribution feature input units, and dynamic behavior feature input units.

[0114] Explanatory: The morphological feature input unit uses a residual neural network. Its convolutional layers extract the input morphological features, obtaining the morphology of the hidden danger object, such as edge contours and texture. The residual blocks of the residual neural network are then used to capture deeper morphology, such as object color. The morphology captured by the convolutional layers is then fused with that captured by the residual blocks to form a high-dimensional morphological feature vector.

[0115] In the spatial distribution feature input unit, the actual distance between the hidden danger objects is calculated by constructing a graph model, wherein the nodes in the graph model represent the hidden danger objects, and the edges in the graph model represent the distance of the hidden danger objects in the graph model. Further, a clustering algorithm is used to analyze the actual distance between the hidden danger objects, and a corresponding adjacency matrix is generated for each node in the graph model according to the analysis result, and according to the adjacency matrix, a spatial distribution feature vector of each node, i.e., the hidden danger object, can be generated.

[0116] The dynamic behavior feature input unit adopts a bidirectional long short-term memory network to process the dynamic behavior features, to obtain the moving track features of the hidden danger objects, and then generate a dynamic behavior feature vector of the hidden danger object according to the moving track features.

[0117] Further, the morphological feature vector, the spatial distribution feature vector and the dynamic behavior feature vector are weighted processed by the full connection layer of the multi-task fusion model, and the weighted processed morphological feature vector, spatial distribution feature vector and dynamic behavior feature vector are input to the preset classifier located in the output layer of the multi-task fusion model.

[0118] The probability that the hidden danger object belongs to the preset event category is predicted based on the preset classifier, and all the probabilities are sorted in descending order. The preset event category corresponding to the first probability is determined as the event category to which the hidden danger object belongs. The event category to which the hidden danger object belongs represents an event or activity associated with the hidden danger object.

[0119] The process of determining the event category to which the hidden danger object belongs is explained through a specific example. Assuming that the hidden danger object is a construction facility for the construction of a distribution network line, the preset classifier finds through prediction that the construction facility may belong to three types of preset events, which are: the construction facility occupies a non-construction area, the construction facility occupies a main road, and the construction facility continuously occupies the main road for more than a preset time length. On this basis, the preset classifier continues to predict that the probability that the construction facility occupies a non-construction area is 0.7, the probability that the construction facility occupies a main road is 0.25, and the probability that the construction facility continuously occupies the main road for more than a preset time length is 0.05. As can be seen, the probability that the construction facility occupies a non-construction area is the highest, and therefore, the event category to which the construction facility belongs is determined as “the construction facility occupies a non-construction area”.

[0120] S103, obtaining geographical position data of the hidden danger object, and mapping the geographical position data and the event category to a target GIS platform to determine the geographical distribution of the hidden danger object in a target area.

[0121] Specifically, the geographic location data of potential hazards and their associated event categories are mapped to the target GIS platform. Based on the target GIS platform, the geographic location data and event categories are then associated to obtain the geographic distribution data of potential hazards within the target area. This geographic distribution data can include the identifier of the event category, geographic location coordinates, and information about the street or road to which the potential hazard belongs. This geographic distribution data is then formatted to generate a structured geographic distribution picture.

[0122] As you can understand, the target GIS platform is a system for capturing, storing, analyzing, and displaying geospatial data. By integrating the location data of hazardous objects and their associated event categories into the target GIS platform, data visualization and analysis can be achieved. Specifically, the target GIS platform can display the location data and associated event categories in the form of a map, allowing users to intuitively view the geographical distribution of hazardous objects.

[0123] S104: Determine a risk score for the hidden danger object based on the event category and geographical distribution of the hidden danger object, and determine a risk level for the hidden danger object based on the risk score.

[0124] It should be noted that the calculation of the risk score of a hidden danger object depends not only on the event category and geographical distribution of the hidden danger object, but also on the combination of relevant risk factors. In this embodiment, the relevant risk factors include morphological feature risk factors related to morphological features. , spatial distribution characteristic risk factors related to spatial distribution characteristics , and dynamic behavior risk factors associated with dynamic behavior .

[0125] Specifically, based on the morphological characteristics, the first preset calculation formula is used to calculate the corresponding morphological characteristic risk factor; based on the spatial distribution characteristics, the second preset calculation formula is used to calculate the corresponding spatial distribution characteristic risk factor; based on the dynamic behavior characteristics, the third preset calculation formula is used to calculate the corresponding dynamic behavior characteristic risk factor.

[0126] The first preset calculation formula is as follows:

[0127]

[0128] in, Indicates the The morphological characteristic values ​​of hidden danger objects, Preset parameters representing morphological feature values, represents the weighting coefficient of the morphological eigenvalue, Indicates the total number of morphological features.

[0129] The second preset calculation formula is as follows:

[0130]

[0131] in, Indicates the The first hidden danger object and the The distance between adjacent hidden danger objects, Indicates the The weight of adjacent hidden danger objects, Indicates the The total number of all adjacent potential hazards of a potential hazard object.

[0132] The third preset calculation formula is as follows:

[0133]

[0134] in, Indicates the A hidden danger object The dynamic behavior value at the moment, Indicates a time period.

[0135] After calculating the morphological characteristic risk factor, spatial distribution characteristic risk factor and dynamic behavior characteristic risk factor, the contribution value of the morphological characteristic risk factor, spatial distribution characteristic risk factor and dynamic behavior characteristic risk factor to the risk score is determined according to the event category and geographical distribution of the hidden danger object. Then, based on the contribution value, the first weight is assigned to the morphological characteristic risk factor. , assign the second weight to the spatial distribution characteristic risk factor , and assigning a third weight to the dynamic behavior characteristic risk factor .

[0136] Furthermore, the morphological characteristic risk factor is multiplied by the first weight, the spatial distribution characteristic factor is multiplied by the second weight, and the dynamic behavior characteristic risk factor is multiplied by the third weight, and the sum is calculated to obtain the risk score of the hidden danger object.

[0137]

[0138] in, Indicates the Risk score of each hazardous object, Indicates the The risk factor of the morphological characteristics of the hidden danger objects, Indicates the The spatial distribution characteristic risk factors of hidden danger objects are Indicates the The dynamic behavior characteristic risk factor of a hidden danger object.

[0139] Next, the risk level of the hidden danger object can be determined based on the risk score calculated above. Specifically, if the risk score is less than or equal to the first scoring threshold, the risk level of the hidden danger object is determined to be no risk. If the risk score is greater than the first scoring threshold and less than or equal to the second scoring threshold, the risk level of the hidden danger object is determined to be low risk. If the risk score is greater than the second scoring threshold and less than or equal to the third scoring threshold, the risk level of the hidden danger object is determined to be medium risk. If the risk score is greater than the third scoring threshold, the risk level of the hidden danger object is determined to be high risk.

[0140] It should be understood that if the geographical distribution indicates that the hidden danger object is located at a preset high-risk location within the target area, and / or the risk level indicates that the hidden danger object is risky, it means that the hidden danger object may pose a threat to the safety of the distribution network line. At this time, a hidden danger object processing request needs to be sent to the relevant staff.

[0141] It's worth noting that if multiple hazardous objects are located in pre-set high-risk locations or present a risk, all of them need to be addressed. In this case, the order of handling is determined based on the priority of the hazardous objects. Specifically, the priority of the hazardous objects awaiting treatment is determined based on their risk scores. The objects are ranked from highest to lowest risk, with the higher the ranking, the higher the priority.

[0142] It should be understood that the priority ranking of potential hazards awaiting treatment indicates the severity of the threat they pose to the distribution network lines. The higher the ranking of the potential hazards, the greater the threat they pose to the distribution network lines. Therefore, the top-ranked potential hazards should be addressed first, and then the lower-ranked ones should be addressed in descending order of priority to optimize emergency response capabilities to potential hazards.

[0143] The present application provides a spatial and risk determination method for potential hazards of distribution lines. This method acquires multimodal data from a target area, extracts features from the multimodal data, and determines the multidimensional characteristics of the potential hazards. The target area includes distribution lines and potential hazards posing a threat to them. The multidimensional characteristics include morphological features, spatial distribution features, and dynamic behavior features. By collecting data from multiple sources, a relatively comprehensive understanding of potential hazards can be provided, enabling a comprehensive description of the potential hazards and providing a reliable data foundation for subsequent determination of their geographic distribution. The multidimensional features are then input into a pre-established multi-task fusion model. Based on the multi-task fusion model, the multidimensional features are converted into structured multidimensional feature vectors, which are then weighted. The weighted multidimensional feature vectors are then input into a pre-set classifier within the multi-task fusion model. The pre-set classifier predicts the probability that the potential hazard belongs to a pre-set event category, and the pre-set event category with the highest probability is determined as the event category to which the potential hazard belongs. Because the multi-task fusion model has powerful recognition capabilities, it can efficiently classify potential hazard events even in complex regional environments. Next, the geographic location data of the hazardous object is obtained and mapped to the target GIS platform along with its event category. Based on the target GIS platform, the geographic location data is associated with its event category to obtain the geographic distribution data of the hazardous object within the target area. This geographic distribution data is then formatted to generate a structured geographic distribution. Combining the geographic location data and event category of the hazardous object with the target GIS platform allows for a visual display of the geographic distribution of the hazardous object on a map. The target GIS platform can also precisely locate the hazardous object using its geographic coordinates, thus accurately determining its geographic distribution. Furthermore, a pre-set calculation formula is used to calculate multidimensional feature risk factors associated with the multidimensional features. Based on the event category and geographic distribution of the hazardous object, the contribution of each multidimensional feature risk factor to the risk score is determined. Based on the contribution value, a corresponding weight is assigned to each multidimensional feature risk factor. Each multidimensional feature risk factor is then multiplied by the corresponding weight and summed to obtain the risk score of the hazardous object. Finally, based on the range of the risk score threshold, the risk level of the hazardous object is determined. Risk levels include no risk, low risk, medium risk, and high risk. On the premise that the geographical distribution of hidden danger objects is accurate, the risk level of the hidden danger objects finally determined is also accurate.

[0144] Figure 2 A schematic diagram of the structure of the space and risk determination device for hidden danger objects in distribution network lines provided in an embodiment of the present application is shown as follows: Figure 2As shown, the space and risk determination device 200 for hidden danger objects of distribution network lines includes: an acquisition module 201, a processing module 202, and a determination module 203;

[0145] The acquisition module 201 is used to acquire multimodal data of a target area; wherein the target area includes distribution network lines and hidden danger objects that pose a potential threat to the distribution network lines;

[0146] Processing module 202 is configured to extract features from the multimodal data to determine the multidimensional features of the hidden danger object; input the multidimensional features into a pre-built multi-task fusion model, and predict the event category of the hidden danger object based on the multi-task fusion model; wherein the event category represents the event or activity associated with the hidden danger object;

[0147] The acquisition module 201 is also used to obtain the geographical location data of the hidden danger object;

[0148] The processing module 202 is further configured to map the geographic location data and the corresponding event category to a target geographic information system (GIS) platform to determine the geographic distribution of the hidden danger objects within the target area.

[0149] The determination module 203 is configured to determine a risk score of the hidden danger object based on the event category and geographical distribution of the hidden danger object, and to determine a risk level of the hidden danger object based on the risk score.

[0150] In a possible design, the processing module 202 further includes: an integration module 204, an extraction module 205,

[0151] Integration module 204, for integrating multimodal data to form a multidimensional heterogeneous data set;

[0152] The processing module 202 is further used to pre-process the multi-dimensional heterogeneous data set;

[0153] The extraction module 205 is used to extract multidimensional features of hidden danger objects from the preprocessed multidimensional heterogeneous data set; wherein the multidimensional features include morphological features, spatial distribution features and dynamic behavior features.

[0154] In one possible design, the multidimensional heterogeneous data set includes image data, video data, lidar data, sensor data, and GIS data.

[0155] The extraction module 205 is further configured to use a preset convolutional neural network to perform feature extraction on the image data and the video data to obtain morphological features of the hidden danger object;

[0156] The extraction module 205 further includes a calculation module 206 configured to calculate, based on the lidar data and the GIS data, a clustering density of the hidden hazard object in the target region by using a preset point cloud clustering algorithm.

[0157] The determination module 203 is further configured to analyze the clustering density, determine a distribution condition and a distribution mode of the hidden hazard object, and determine a spatial distribution feature of the hidden hazard object based on the distribution condition and the distribution mode.

[0158] The determination module 203 is further configured to determine a moving track of the hidden hazard object by using a preset time series analysis algorithm based on the sensor data, and determine a dynamic behavior feature of the hidden hazard object based on the moving track.

[0159] In a possible design, the processing module 202 further includes a conversion module 207, an input module 208, a prediction module 209, and a sorting module 210.

[0160] The conversion module 207 is configured to convert the morphological feature, the spatial distribution feature, and the dynamic behavior feature into a structured morphological feature vector, a structured spatial distribution feature vector, and a structured dynamic behavior feature vector based on the multi-task fusion model.

[0161] The processing module 202 is further configured to perform weighted processing on the morphological feature vector, the spatial distribution feature vector, and the dynamic behavior feature vector.

[0162] The input module 208 is configured to input the morphological feature vector, the spatial distribution feature vector, and the dynamic behavior feature vector after the weighted processing to a preset classifier included in the multi-task fusion model.

[0163] The prediction module 209 is configured to predict a probability that the hidden hazard object belongs to a preset event category based on the preset classifier.

[0164] The sorting module 210 is configured to sort all the probabilities in a descending order.

[0165] The determination module 203 is further configured to determine, as a belonging event category of the hidden hazard object, a preset event category corresponding to a probability ranked first.

[0166] In a possible design, the processing module 202 further includes a mapping module 211 and an association module 212.

[0167] The mapping module 211 is configured to map the geographic location data and the belonging event category to a target GIS platform.

[0168] The association module 212 is configured to associate the geographic location data and the belonging event category based on the target GIS platform, to obtain geographic distribution data of the hidden hazard object in the target region.

[0169] The processing module 202 is further configured to format the geographic distribution data to generate structured geographic distribution information.

[0170] In a possible design, the multi-dimensional characteristics of hidden danger objects include morphological characteristics, spatial distribution characteristics and dynamic behavior characteristics.

[0171] The calculation module 206 is further configured to calculate a corresponding morphological feature risk factor based on the morphological feature using a first preset calculation formula; calculate a corresponding spatial distribution feature risk factor based on the spatial distribution feature using a second preset calculation formula; and calculate a corresponding dynamic behavior feature risk factor based on the dynamic behavior feature using a third preset calculation formula;

[0172] The determination module 203 is further configured to determine the contribution of the morphological characteristic risk factor, the spatial distribution characteristic risk factor, and the dynamic behavior characteristic risk factor to the risk score based on the event category and geographical distribution of the hidden danger object;

[0173] The determination module 203 further includes: an allocation module 213 for allocating a first weight to the morphological characteristic risk factor, a second weight to the spatial distribution characteristic risk factor, and a third weight to the dynamic behavior characteristic risk factor based on the contribution value;

[0174] The processing module 202 is further configured to multiply the morphological characteristic risk factor by the first weight, the spatial distribution characteristic risk factor by the second weight, and the dynamic behavior characteristic risk factor by the third weight, and sum them up to obtain a risk score for the hidden danger object.

[0175] In one possible design, the determination module 203 is further configured to:

[0176] If the risk score is less than or equal to the first score threshold, the risk level of the hidden danger object is determined to be a no-risk level;

[0177] If the risk score is greater than the first scoring threshold and less than or equal to the second scoring threshold, the risk level of the hidden danger object is determined to be a low risk level;

[0178] If the risk score is greater than the second scoring threshold and less than or equal to the third scoring threshold, the risk level of the hidden danger object is determined to be a medium risk level;

[0179] If the risk score is greater than the third score threshold, the risk level of the hidden danger object is determined to be a high risk level.

[0180] In a possible design, the space and risk determination apparatus for the hidden object of the distribution network line further includes a sending module 214 configured to send a hidden object processing request to a relevant worker if the geographic distribution indicates that the hidden object is located at a preset high-risk position in the target area and / or the risk level indicates that the hidden object has a risk.

[0181] The space and risk determination apparatus for the hidden object of the distribution network line provided in the embodiments of the present application can be used to execute the space and risk determination method for the hidden object of the distribution network line in any of the above embodiments, and has similar implementation principles and technical effects, which will not be repeated here.

[0182] It should be noted that the division of each module of the above apparatus is only a logical functional division, and all or part of the modules can be integrated into one physical entity, or can be physically separated. The modules can all be implemented in the form of software invoked by a processing element; or all be implemented in the form of hardware; or part of the modules are implemented in the form of software invoked by a processing element, and part of the modules are implemented in the form of hardware. In addition, all or part of the modules can be integrated together, or can be independently implemented. The processing element described herein can be an integrated circuit having a signal processing capability. In the implementation process, each step of the above method or each module can be completed by an integrated logic circuit of hardware in the processing element or an instruction in the form of software.

[0183] Figure 3 A structural schematic diagram of an electronic device provided in the embodiments of the present application is shown in FIG. 3. Figure 3 As shown in FIG. 3, the electronic device can include a transceiver 31, a processor 32, and a memory 33.

[0184] The processor 32 executes computer execution instructions stored in the memory, so that the processor 32 executes the scheme in the above embodiments. The processor 32 can be a general-purpose processor, including a central processing unit CPU, a network processor NP, etc.; and can also be a digital signal processor DSP, an application-specific integrated circuit ASIC, a field programmable gate array FPGA or other programmable logic device, a discrete gate or transistor logic device, a discrete hardware component.

[0185] The memory 33 is connected with the processor 32 through a system bus and completes mutual communication, and the memory 33 is used to store computer program instructions.

[0186] The transceiver 31 can be used for communication interaction with other devices.

[0187] The system bus can be a Peripheral Component Interconnect (PCI) bus or an Extended Industry Standard Architecture (EISA) bus, among others. System buses can be divided into address buses, data buses, and control buses. For ease of illustration, the diagram uses only a single thick line, but this does not imply a single bus or type of bus. Transceivers enable communication between the database access device and other computers (such as clients, read-write libraries, and read-only libraries). Memory may include random access memory (RAM) and non-volatile memory.

[0188] The electronic device provided in the embodiments of the present application can be used to execute the method provided in any of the above embodiments. Its implementation principles and technical effects are similar and will not be repeated here.

[0189] An embodiment of the present application further provides a computer-readable storage medium, in which computer instructions are stored. When the computer instructions are executed on a computer, the computer executes the method provided in any of the above embodiments.

[0190] An embodiment of the present application also provides a computer program product, which includes a computer program stored in a computer-readable storage medium. At least one processor can read the computer program from the computer-readable storage medium, and when at least one processor executes the computer program, it can implement the method provided in any of the above embodiments.

[0191] In the several embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of modules is only a logical function division. In actual implementation, there may be other division methods, such as multiple modules can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the mutual coupling or direct coupling or communication connection shown or discussed can be an indirect coupling or communication connection through some interface, device or module, which can be electrical, mechanical or other forms.

[0192] Modules described as separate components may or may not be physically separate, and components shown as modules may or may not be physical units, that is, they may be located in one place or distributed across multiple network elements. Some or all of these modules may be selected to implement the solution of this embodiment based on actual needs.

[0193] In addition, the functional modules in the various embodiments of the present application may be integrated into a single processing unit, or each module may exist physically separately, or two or more modules may be integrated into a single unit. The above-mentioned modules may be implemented in the form of hardware or hardware plus software functional units.

[0194] The integrated modules implemented in the form of software function modules can be stored in a computer-readable storage medium. The software function modules stored in a storage medium include a number of instructions for causing a computer device (which can be a personal computer, server, or network device, etc.) or a processor to execute some of the steps of the methods of various embodiments of the present application.

[0195] It should be understood that the processor described above may be a central processing unit (CPU), other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), etc. A general-purpose processor may be a microprocessor or any conventional processor. The steps of the method disclosed in the present invention may be directly executed by a hardware processor or by a combination of hardware and software modules within the processor.

[0196] The memory may include a high-speed RAM memory, and may also include non-volatile storage NVM, such as at least one disk memory, and may also be a USB flash drive, a mobile hard disk, a read-only memory, a magnetic disk or an optical disk.

[0197] A bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus. Buses can be categorized as address buses, data buses, and control buses. For ease of illustration, the buses in the drawings of this application are not limited to just one bus or just one type of bus.

[0198] The storage medium may be implemented by any type of volatile or non-volatile memory device, or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. The storage medium may be any available medium that can be accessed by a general-purpose or special-purpose computer.

[0199] An exemplary storage medium is coupled to a processor so that the processor can read information from the storage medium and write information to the storage medium. Of course, the storage medium can also be an integral part of the processor. The processor and the storage medium can be located in an application specific integrated circuit (ASIC). Of course, the processor and the storage medium can also exist as discrete components in an electronic control unit or a main control device.

[0200] Those skilled in the art will appreciate that all or part of the steps in the above-described method embodiments can be implemented using hardware associated with program instructions. The aforementioned program can be stored in a computer-readable storage medium. When executed, the program performs the steps of the above-described method embodiments. The aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks.

[0201] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some or all of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the scope of the technical solutions of the embodiments of the present application.

Claims

1. A method for determining the space and risk of hidden danger objects in distribution network lines, characterized in that: include: Acquire multimodal data of a target area; wherein the target area includes a distribution network line and hidden danger objects that pose a potential threat to the distribution network line; Performing feature extraction on the multimodal data to determine the multidimensional features of the hidden danger object; inputting the multidimensional features into a pre-built multi-task fusion model, and predicting the event category of the hidden danger object based on the multi-task fusion model; wherein the event category represents an event or activity associated with the hidden danger object; Obtaining geographic location data of the hidden danger objects, and mapping the geographic location data and the event categories to which they belong to onto a target geographic information system (GIS) platform to determine the geographic distribution of the hidden danger objects within the target area; Based on the event category and geographical distribution of the hidden danger object, a risk score of the hidden danger object is determined, and based on the risk score, a risk level of the hidden danger object is determined.

2. The method according to claim 1, characterized in that The extracting features from the multimodal data to determine the multidimensional features of the hidden danger object includes: Integrating the multimodal data to form a multidimensional heterogeneous data set; The multidimensional heterogeneous data set is preprocessed, and the multidimensional features of the hidden danger object are extracted from the preprocessed multidimensional heterogeneous data set; wherein the multidimensional features include morphological features, spatial distribution features and dynamic behavior features.

3. The method according to claim 2, characterized in that The multidimensional heterogeneous data set includes image data, video data, lidar data, sensor data and GIS data. Extracting multidimensional features of the hidden danger object from the preprocessed multidimensional heterogeneous data set includes: Using a preset convolutional neural network, feature extraction is performed on the image data and the video data to obtain morphological features of the hidden danger object; Based on the lidar data and the GIS data, a preset point cloud clustering algorithm is used to calculate the cluster density of the hidden danger objects in the target area; the cluster density is analyzed to determine the distribution and distribution pattern of the hidden danger objects, and based on the distribution and distribution pattern, the spatial distribution characteristics of the hidden danger objects are determined; Based on the sensor data, a preset time series analysis algorithm is used to determine the movement trajectory of the hidden danger object, and based on the movement trajectory, the dynamic behavior characteristics of the hidden danger object are determined.

4. The method according to claim 2, characterized in that The predicting the event category of the hidden danger object based on the multi-task fusion model includes: Based on the multi-task fusion model, the morphological features, the spatial distribution features, and the dynamic behavior features are converted into structured morphological feature vectors, spatial distribution feature vectors, and dynamic behavior feature vectors; Performing weighted processing on the morphological feature vector, the spatial distribution feature vector, and the dynamic behavior feature vector; inputting the weighted morphological feature vector, the spatial distribution feature vector, and the dynamic behavior feature vector into a preset classifier included in the multi-task fusion model; Based on the preset classifier, the probability that the hidden danger object belongs to the preset event category is predicted; all probabilities are sorted in order from large to small, and the preset event category corresponding to the probability ranked first is determined as the event category to which the hidden danger object belongs.

5. The method according to claim 1, wherein Mapping the geographic location data and the event category to a target geographic information system (GIS) platform to determine the geographic distribution of the hidden danger objects within the target area includes: Mapping the geographic location data and the event category to a target GIS platform; and associating the geographic location data with the event category based on the target GIS platform to obtain geographic distribution data of the hidden danger objects within the target area; The geographic distribution data is formatted to generate structured geographic distribution information.

6. The method according to claim 1, characterized in that The multi-dimensional characteristics of the hidden danger objects include morphological characteristics, spatial distribution characteristics and dynamic behavior characteristics. Determining the risk score of the hidden danger object based on the event category and geographical distribution of the hidden danger object includes: Based on the morphological characteristics, a first preset calculation formula is used to calculate the corresponding morphological characteristic risk factor; based on the spatial distribution characteristics, a second preset calculation formula is used to calculate the corresponding spatial distribution characteristic risk factor; based on the dynamic behavior characteristics, a third preset calculation formula is used to calculate the corresponding dynamic behavior characteristic risk factor; Determining, based on the event category and geographical distribution of the hidden danger object, the contribution values ​​of the morphological characteristic risk factor, the spatial distribution characteristic risk factor, and the dynamic behavior characteristic risk factor to the risk score; assigning a first weight to the morphological characteristic risk factor, a second weight to the spatial distribution characteristic risk factor, and a third weight to the dynamic behavior characteristic risk factor based on the contribution values; The risk score of the hidden danger object is obtained by multiplying the morphological characteristic risk factor by the first weight, the spatial distribution characteristic risk factor by the second weight, and the dynamic behavior characteristic risk factor by the third weight and summing the results.

7. The method according to claim 1, characterized in that Determining the risk level of the hidden danger object based on the risk score includes: If the risk score is less than or equal to the first scoring threshold, the risk level of the hidden danger object is determined to be a no-risk level; If the risk score is greater than the first score threshold and less than or equal to the second score threshold, the risk level of the hidden danger object is determined to be a low risk level; If the risk score is greater than the second scoring threshold and less than or equal to the third scoring threshold, the risk level of the hidden danger object is determined to be a medium risk level; If the risk score is greater than a third score threshold, the risk level of the hidden danger object is determined to be a high risk level.

8. The method according to any one of claims 1 to 7, characterized in that The method further comprises: If the geographical distribution indicates that the hidden danger object is located at a preset high-risk location within the target area, and / or the risk level indicates that the hidden danger object is risky, a hidden danger object handling request is sent to relevant staff.

9. A device for determining the space and risk of hidden danger objects in distribution network lines, characterized in that: include: An acquisition module, configured to acquire multimodal data of a target area, wherein the target area includes distribution network lines and potential objects that pose a threat to the distribution network lines; a processing module configured to extract features from the multimodal data to determine multidimensional features of the hidden danger object; input the multidimensional features into a pre-built multi-task fusion model, and predict the event category of the hidden danger object based on the multi-task fusion model; wherein the event category represents an event or activity associated with the hidden danger object; The acquisition module is further configured to acquire geographic location data of the hidden danger object; The processing module is further configured to map the geographic location data and the event category to a target geographic information system (GIS) platform to determine the geographic distribution of the hidden danger objects within the target area. The determination module is configured to determine a risk score of the hidden danger object based on the event category and geographical distribution of the hidden danger object, and to determine a risk level of the hidden danger object based on the risk score.

10. An electronic device, characterized in that: include: a processor, and a memory communicatively connected to the processor; The memory stores computer-executable instructions; The processor executes the computer-executable instructions stored in the memory to implement the method for determining the space and risk of hidden danger objects for distribution network lines according to any one of claims 1 to 8.

11. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the method for determining the space and risk of hidden danger objects for distribution network lines according to any one of claims 1 to 8.

Images