A safety distance warning method and system for substations

Abstract

The present invention provides a safety distance warning method and system for a substation, which relates to the field of safety distance warning technology. The method includes using an equipment processing model to determine the substation high-voltage cable information and multiple live equipment information in the station based on the substation panoramic image information and the substation design drawing information; constructing an equipment map; processing the equipment map based on a first graph neural network to determine multiple electrical risk point information and multiple electrical safety point information; constructing a road map; processing the road map based on a second graph neural network to determine the operating route of the equipment to be operated; determining multiple line segments and the safe activity range of the boom of each line segment based on the crane boom information and the operating route of the equipment to be operated; and issuing an early warning to the equipment to be operated based on the operating route of the equipment to be operated and the safe activity range of the boom of each line segment. This method can accurately determine the safe activity range of the equipment operation.

Description

Technical Field

[0001] The present invention relates to the technical field of safety distance warning, and in particular to a safety distance warning method and system for a substation. Background Art

[0002] In the power system, substations are key nodes for power transmission and distribution. The installation, maintenance, and upgrade and modification of their equipment often involve equipment operation. Currently, substation equipment operation faces many difficulties. Substations are densely populated with live equipment, and the distribution of various electrical risk points is complex. For example, there are strong electric field areas around high-voltage cables. If the equipment is too close during operation, it is very easy to cause safety accidents such as electric shock. Traditional operation route planning relies heavily on manual experience, which makes it difficult to comprehensively and accurately consider all electrical risk factors and cannot guarantee the safety of the operation line. During equipment operation, the range of movement of operating equipment such as crane booms must be strictly controlled. Once the safe range is exceeded, it may not only damage the equipment, but also pose a serious threat to the safety of live equipment and personnel in the surrounding stations. Existing methods make it difficult to accurately determine the safe range of movement of the boom in a complex electrical environment in real time.

[0003] Therefore, how to identify live equipment and issue distance threshold warnings to significantly improve the personal safety of workers is an urgent problem that needs to be solved. Summary of the Invention

[0004] The main technical problem solved by the present invention is how to identify live equipment and issue early warning of distance thresholds, which greatly improves the personal safety of operators and is an urgent problem to be solved.

[0005] According to a first aspect, the present invention provides a safety distance warning method for a substation, comprising: obtaining panoramic image information and design drawing information of the substation; determining high-voltage cable information of the substation and information of multiple live equipment within the station using a device processing model based on the panoramic image information and the design drawing information of the substation; constructing an equipment map, the equipment map comprising multiple live equipment nodes within the station and multiple edges between the live equipment nodes within the station, the node features of the live equipment nodes within the station being each live equipment information within the station and the panoramic image information of the substation, and the edges between the nodes being the distances between the live equipment within the station; processing the equipment map based on a first graph neural network to determine multiple electrical risk point information and multiple electrical safety point information; constructing a road map based on the multiple electrical risk points and the multiple electrical safety points; processing the road map based on a second graph neural network to determine an operating route of the equipment to be operated; determining multiple line segments and a safe range of boom activity of each line segment based on crane boom information and the operating route of the equipment to be operated; and issuing an early warning for the equipment to be operated based on the operating route of the equipment to be operated and the safe range of boom activity of each line segment.

[0006] In one possible implementation, constructing a road map based on the multiple electrical risk points and the multiple electrical safety points includes: constructing a road map, the road map including multiple electrical risk nodes, multiple electrical safety nodes and multiple edges between nodes, the node characteristics of the electrical risk nodes are the location and risk level of the electrical risk points, the electrical safety nodes are the location and safety level of the electrical safety points, and the panoramic image information of the substation, and the edges between nodes are the distances between the nodes.

[0007] In one possible implementation, the determining of multiple line segments and the safe range of boom movement of each line segment based on the crane boom information and the operating line of the equipment to be operated includes: obtaining risk path segments affected by multiple high-voltage cables and safe path segments affected by multiple high-voltage cables in the operating line based on the substation high-voltage cable information and the operating line of the equipment to be operated; determining multiple line segments and the safe range of boom movement of each line segment based on the risk path segments affected by the multiple high-voltage cables, the safe path segments affected by the multiple high-voltage cables, the crane boom information, the multiple electrical risk points, and the multiple electrical safety points.

[0008] In one possible implementation, the device processing model is a convolutional neural network.

[0009] According to the second aspect, the present invention provides a safety distance warning system for a substation, comprising: an information acquisition module for acquiring panoramic image information and design drawing information of the substation; an equipment processing module for determining high-voltage cable information of the substation and information of multiple live equipment within the station using an equipment processing model based on the panoramic image information and design drawing information of the substation; a graph construction module for constructing an equipment graph, the equipment graph comprising multiple live equipment nodes within the station and multiple edges between the live equipment nodes within the station, the node features of the live equipment nodes within the station being the live equipment information within each station and the panoramic image information of the substation, and the edges between the nodes being the distances between the live equipment within the station; a risk analysis module A block is used to process the equipment map based on the first graph neural network to determine multiple electrical risk point information and multiple electrical safety point information; a second map construction module is used to construct a road map based on the multiple electrical risk points and the multiple electrical safety points; a line planning module is used to process the road map based on the second graph neural network to determine the operating route of the equipment to be operated; a safety range determination module is used to determine multiple line segments and the safe activity range of the boom of each line segment based on the crane boom information and the operating route of the equipment to be operated; an early warning execution module is used to issue an early warning to the equipment to be operated based on the operating route of the equipment to be operated and the safe activity range of the boom of each line segment.

[0010] In one possible implementation, the second graph construction module is also used to: construct a road graph, the road graph includes multiple electrical risk nodes, multiple electrical safety nodes and multiple edges between nodes, the node characteristics of the electrical risk node are the location and risk level of the electrical risk point, the electrical safety node is the location and safety level of the electrical safety point, and the panoramic image information of the substation, and the edges between nodes are the distances between the nodes.

[0011] In a possible implementation, the safety range determination module is also used to: obtain the risk path segments affected by multiple high-voltage cables and the safety path segments affected by multiple high-voltage cables in the operating line based on the high-voltage cable information of the substation and the operating line of the equipment to be operated; determine the multiple line segments and the safe activity range of the boom of each line segment based on the risk path segments affected by the multiple high-voltage cables, the safety path segments affected by the multiple high-voltage cables, the crane boom information, the multiple electrical risk points, and the multiple electrical safety points.

[0012] In one possible implementation, the device processing model is a convolutional neural network.

[0013] According to a third aspect, an embodiment of the present invention provides an electronic device comprising: a processor; a memory; and a computer program; wherein the computer program is stored in the memory and configured to be executed by the processor to implement the aforementioned method, the method comprising: obtaining panoramic image information of a substation and information on design drawings of a substation; determining high-voltage cable information of a substation and information on multiple in-station live equipment using an equipment processing model based on the panoramic image information of the substation and the information on the in-station live equipment; constructing an equipment map, the equipment map comprising multiple in-station live equipment nodes and multiple edges between the in-station live equipment nodes, and the node features of the in-station live equipment nodes are each station The network comprises the following components: information on live equipment in the substation and panoramic image information of the substation, with the edges between nodes being the distances between live equipment in the substation; processing the equipment map based on the first graph neural network to determine information on multiple electrical risk points and multiple electrical safety points; constructing a road map based on the multiple electrical risk points and the multiple electrical safety points; processing the road map based on the second graph neural network to determine the operating routes of the equipment to be operated; determining multiple line segments and the safe activity range of the boom of each line segment based on the crane boom information and the operating route of the equipment to be operated; and issuing early warnings for the equipment to be operated based on the operating route of the equipment to be operated and the safe activity range of the boom of each line segment.

[0014] According to the fourth aspect, this embodiment provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the aforementioned safety distance warning method for the substation, the method comprising: obtaining panoramic image information and substation design drawing information of the substation; using an equipment processing model to determine the substation high-voltage cable information and multiple in-station live equipment information based on the panoramic image information and the substation design drawing information of the substation; constructing an equipment map, the equipment map comprising multiple in-station live equipment nodes and multiple edges between the in-station live equipment nodes, the node features of the in-station live equipment nodes being each in-station live equipment information, substation live equipment information, and substation live equipment information. Panoramic image information of the power station, where the edges between nodes are the distances between energized equipment in the station; processing the equipment map based on the first graph neural network to determine information on multiple electrical risk points and multiple electrical safety points; constructing a road map based on the multiple electrical risk points and the multiple electrical safety points; processing the road map based on the second graph neural network to determine the operating routes of the equipment to be operated; determining multiple line segments and the safe activity range of the boom of each line segment based on the crane boom information and the operating route of the equipment to be operated; and issuing early warnings for the equipment to be operated based on the operating route of the equipment to be operated and the safe activity range of the boom of each line segment.

[0015] The present invention provides a safety distance warning method and system for a substation, the method comprising obtaining panoramic image information and substation design drawing information of the substation; determining substation high-voltage cable information and multiple in-station live equipment information using an equipment processing model based on the panoramic image information and the substation design drawing information of the substation; constructing an equipment map, the equipment map comprising multiple in-station live equipment nodes and multiple edges between the in-station live equipment nodes, the node features of the in-station live equipment nodes being each in-station live equipment information and substation panoramic image information, the edges between the nodes being the distances between the in-station live equipment; based on a first graph neural network The network processes the equipment map to determine multiple electrical risk point information and multiple electrical safety point information; constructs a road map based on the multiple electrical risk points and the multiple electrical safety points; processes the road map based on the second graph neural network to determine the operating route of the equipment to be operated; determines multiple line segments and the safe activity range of the boom of each line segment based on the crane boom information and the operating route of the equipment to be operated; and issues an early warning for the equipment to be operated based on the operating route of the equipment to be operated and the safe activity range of the boom of each line segment. This method can accurately determine the safe activity range of substation equipment operations. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 A schematic flow chart of a safety distance warning method for a substation provided by an embodiment of the present invention;

[0017] Figure 2A schematic diagram of a substation provided by an embodiment of the present invention;

[0018] Figure 3 A schematic diagram of a device map constructed according to an embodiment of the present invention;

[0019] Figure 4 A schematic diagram of a road map constructed according to an embodiment of the present invention;

[0020] Figure 5 A schematic diagram of a process for determining a plurality of line segments and a safe range of movement of a boom for each line segment provided by an embodiment of the present invention;

[0021] Figure 6 A schematic diagram of a crane provided in an embodiment of the present invention;

[0022] Figure 7 A schematic diagram of a safety distance warning system for a substation provided by an embodiment of the present invention; DETAILED DESCRIPTION

[0023] The present invention will be further described in detail below by means of specific embodiments in conjunction with the accompanying drawings. Similar elements in different embodiments are numbered with associated similar elements. In the following embodiments, many detailed descriptions are provided to enable the present invention to be better understood. However, those skilled in the art will readily appreciate that some of the features may be omitted under different circumstances, or may be replaced by other elements, materials, or methods. In some cases, some operations related to the present invention are not shown or described in the specification. This is to avoid the core of the present invention being overwhelmed by excessive descriptions, and for those skilled in the art, it is not necessary to describe these related operations in detail. They can fully understand the related operations based on the description in the specification and the general technical knowledge in the art.

[0024] In an embodiment of the present invention, there is provided Figure 1 A safety distance early warning method for a substation is shown, and the safety distance early warning method for a substation includes steps S1 to S8:

[0025] Step S1: Obtaining substation panoramic image information and substation design drawing information.

[0026] A substation is a place in the power system where voltage and current are transformed, and electrical energy is received and distributed. It consists of buildings, live equipment, and high-voltage cables. Figure 2 A schematic diagram of a substation provided by an embodiment of the present invention.

[0027] Substation panoramic images are captured using panoramic photography technology, providing a comprehensive and complete view of the substation. These images include visual information such as buildings, live equipment, lines, and the surrounding environment.

[0028] Substation design drawing information is used during the substation design phase to describe the substation's planning, layout, equipment installation, and line connection.

[0029] Step S2: Determine the substation high-voltage cable information and the information of multiple live devices in the substation using a device processing model based on the substation panoramic image information and the substation design drawing information.

[0030] The equipment processing model is a convolutional neural network. The input of the equipment processing model is the panoramic image information of the substation and the substation design drawing information. The output of the equipment processing model is the substation high-voltage cable information and information of multiple live equipment in the station.

[0031] A convolutional neural network (CNN) is a deep learning model that excels at processing grid-structured data, such as images. Convolutional neural networks are composed of convolutional layers, pooling layers, and fully connected layers. They can automatically extract features from data and perform classification or regression analysis.

[0032] Substation high-voltage cable information is the physical characteristics, electrical parameters, and spatial location of the high-voltage cables within the substation. This information includes the spatial location of the high-voltage cables, electrical technical parameters, installation method parameters, and operating status parameters.

[0033] The spatial location information of the high-voltage cable is the specific coordinates and direction of the high-voltage cable in the three-dimensional space of the substation.

[0034] Electrical technical parameters are rated indicators and operating parameters related to the electrical performance of high-voltage cables. The parameters include rated voltage, rated current carrying capacity, and cable model.

[0035] The laying method parameters are the installation and laying form and specifications of the high-voltage cable in the substation. For example, direct burial is 0.7 meters deep and uses MPP protective pipe with a diameter of 160 mm.

[0036] Operating status parameters are the real-time or rated status data of the high-voltage cable during actual operation. Operating status parameters include operating voltage, real-time current, and insulation resistance.

[0037] Live equipment within a power system is used for power generation, transmission, transformation, distribution, and consumption. It enables the production, transmission, conversion, control, distribution, and use of electrical energy. Examples include transformers for voltage conversion, circuit breakers and disconnectors for connecting and disconnecting circuits, and busbars for collecting and distributing electrical energy.

[0038] Multiple live equipment information is information about each of the individual live equipment within a substation, representing their physical characteristics, electrical parameters, spatial location, and operating status. This information includes their spatial location, type, electrical technical parameters, operating status, and energized condition parameters.

[0039] The spatial location information of the live equipment in the station is the specific coordinates and occupied space range of the live equipment in the three-dimensional space of the substation.

[0040] The type parameters of live equipment within the station include the functional classification and technical category of the live equipment within the station.

[0041] Electrical technical parameters are rated indicators related to the electrical performance of the equipment. For example, the rated capacity of the transformer is 10,000 kVA, the transformation ratio is 110 kV / 10.5 kV, and the no-load loss is 10 kW; the rated current of the circuit breaker is 2,500 A, the rated short-circuit breaking current is 50 kA, and the opening and closing time is ≤ 50 ms.

[0042] Operating status parameters are real-time or rated status data of live equipment in the station during actual operation. For example, the current load rate of the transformer is 85%, the oil temperature is 65°C, and the winding temperature is 80°C.

[0043] The live status parameters include whether the equipment is energized and the voltage level and phase of the energized parts.

[0044] Substation panoramic imagery contains intuitive visual scene features, such as the physical appearance, spatial layout, and relative position of equipment. The pixel-level details in this panoramic imagery provide convolutional neural networks with the shape and texture of equipment, such as the structure of transformer heat sinks and the color of cable insulation. Real-time panoramic imagery also reflects the actual installation status of equipment, such as tilt or damage, providing visual evidence for equipment identification. Substation design drawings contain standardized electrical parameters and design planning data. This data can be converted into equipment electrical properties and spatial location parameters by parsing drawing legends and symbols using convolutional neural networks. For example, cable routing can be derived from cable symbols, and type parameters can be derived from equipment legends. Combining this panoramic imagery with design drawings can address actual installation deviations, such as subtle discrepancies between the theoretical cable routing in the design drawings and the actual installation.

[0045] Convolutional neural networks (CNNs) automatically extract low-level features such as edges, shapes, and textures from an image's pixel matrix through a combination of multiple convolutional and pooling layers. These features are then gradually integrated into higher-level semantic features such as the equipment's outline and structure through a deep network. For example, they can identify the rectangular outline of a transformer or the routing of cables in a panoramic image. For design drawings, CNNs can analyze elements such as legend symbols and line annotations. By combining these elements with a pre-defined library of features for live substation equipment, such as transformer legends and standard styles for cable symbols, CNNs can convert two-dimensional images from substation design drawings into corresponding equipment types and parameters. The weight sharing and local connectivity of CNNs enable them to efficiently process large amounts of image data, adapting to the densely packed and complex layouts of substations. By training and optimizing parameters, CNNs can accurately distinguish the boundaries between high-voltage cables and other equipment, and then output information about the high-voltage cables and multiple live substation equipment.

[0046] Step S3: construct an equipment map. The equipment map includes multiple in-station powered equipment nodes and multiple edges between the in-station powered equipment nodes. The node features of the in-station powered equipment nodes are the information of each in-station powered equipment and the panoramic image information of the substation. The edges between the nodes are the distances between the in-station powered equipment.

[0047] The device graph is a data structure consisting of two components: nodes and edges. Each node contains node features, which are attributes of the node. These attributes include information about each device and a panoramic image of the substation. Node features describe the properties of each node. Edges describe the relationship between two nodes. Edges between nodes represent the distance between devices. Figure 3A schematic diagram of a device map constructed according to an embodiment of the present invention is provided. Figure 3 As shown, Figure 3 It includes in-station powered device node A, in-station powered device node B, in-station powered device node C, and in-station powered device node D. The edge between each two in-station powered device nodes is the distance between the two in-station powered device nodes.

[0048] Step S4: Processing the equipment map based on the first graph neural network to determine multiple electrical risk point information and multiple electrical safety point information.

[0049] A graph neural network (GNN) is a deep learning model that can process graph-structured data. It consists of three core components: a graph convolutional layer, a graph attention mechanism, and a fully connected prediction layer. The convolutional layer processes node and edge features, the graph attention mechanism dynamically weights the influence of neighboring nodes, and the fully connected prediction layer generates coordinates and risk safety parameters. By performing multi-layer nonlinear transformations and information transfer on node and edge features, the GNN can analyze and predict the risk or safety attributes of nodes in a device graph.

[0050] The input of the first graph neural network is the equipment map, and the output of the first graph neural network is multiple electrical risk point information and multiple electrical safety point information.

[0051] Electrical risk point information, output by the first graph neural network, represents discrete locations within the substation's three-dimensional space containing electrical safety hazards and their quantified risk attributes. This information includes the three-dimensional spatial coordinates of the electrical risk point, its risk level, its associated device identifier, and its risk type classification label.

[0052] The three-dimensional spatial position coordinates are composed of the X-axis coordinate value, the Y-axis coordinate value, and the Z-axis coordinate value. The coordinate unit is meter, where the Z-axis coordinate value represents the height of the electrical risk point vertically to the ground.

[0053] The risk type classification label is a standardized text enumeration value used to identify different electrical risk categories in the substation. The risk type classification label includes high-voltage discharge area, electromagnetic interference area, mechanical collision area, high-temperature radiation area, etc.

[0054] Electrical safety point information is output by the first graph neural network and describes the location and protection parameters of the safety zone within the substation's three-dimensional space. This information includes the three-dimensional spatial coordinates of the electrical safety point, its safety value, and the radius of the safety zone.

[0055] By constructing a device map, the spatial positional relationships and electrical correlations between energized equipment within a substation can be clearly reflected. This relationship information is crucial for electrical risk assessment, as the spatial distances and electrical parameters between devices directly affect the distribution of risk points. Using information about energized equipment within the station and panoramic substation images as node features, and the distances between devices as edge features, can more fully utilize the multidimensional data within the substation. This helps the first graph neural network model better understand the spatial dependencies and electrical correlations between devices, thereby improving the accuracy of locating risk points and safe points.

[0056] As a typical graph-structured data, the equipment graph contains a large number of live equipment nodes within the station and the connection relationships between nodes. The graph neural network can exchange and aggregate information between nodes through a message passing mechanism. The graph neural network can gradually transform the local information of each node into a feature representation containing global relationships through multi-layer graph convolution operations. In this process, the model can learn the spatial dependencies and electrical correlations between devices, such as identifying high-risk spatial locations such as areas near high-voltage cables and equipment overlapping areas, and calculate the risk or safety of each location based on safety distance standards and equipment operating status. In addition, the attention mechanism of the graph neural network can automatically focus on the equipment nodes and connection edges that are most critical to risk assessment. For example, it can focus on analyzing the spatial relationship between live equipment and the paths of equipment to be operated, so as to accurately locate electrical risk points and safety points.

[0057] Step S5: constructing a road map based on the multiple electrical risk points and the multiple electrical safety points.

[0058] A road map is constructed, wherein the road map includes multiple electrical risk nodes, multiple electrical safety nodes, and multiple edges between the nodes. The node features of the electrical risk nodes are the location and risk level of the electrical risk points, the electrical safety nodes are the location and safety level of the electrical safety points, and the panoramic image information of the substation. The edges between the nodes are the distances between the nodes.

[0059] Road maps are graph structure data that can be used to characterize the safety of working paths within substations. Figure 5 A schematic diagram of a road map constructed according to an embodiment of the present invention is provided. Figure 4 As shown, Figure 4 Including electrical risk node A, electrical risk node B, electrical risk node C, and electrical risk node D, the edge between every two electrical risk nodes is the distance between the two electrical risk nodes.

[0060] Step S6: Process the road map based on the second graph neural network to determine the operating route of the equipment to be operated.

[0061] The input of the second graph neural network is the road map, and the output of the second graph neural network is the operation route of the equipment to be operated.

[0062] The operating route of the equipment to be operated is determined by processing the road map using a second graph neural network. It connects the starting point and the end point, avoiding high-risk areas. The operating route of the equipment to be operated consists of multiple ordered path points, each of which has two-dimensional coordinates (x, y). The coordinates of the path points are derived from safe nodes or low-risk nodes in the road map, and the distance between adjacent path points must meet the mobility limitations of the operating equipment, such as the crane.

[0063] The road map abstracts risky and safe areas within a substation into electrical risk nodes and electrical safety nodes in a graph structure, using inter-node distances as edge features. This allows the safety constraints of risk and safety within the operating environment to be transformed into a computable graphical data structure. A graph convolutional network, through graph convolution and attention mechanisms, automatically learns the optimal path relationship between risky and safe nodes. Using node features as the basis for safety assessment and edges as path length constraints, the network dynamically balances risk aversion and path efficiency in its calculations, generating operating routes for equipment that meet safety distance requirements and meet the mobility of the equipment.

[0064] Step S7: determining a plurality of line segments and a safe range of movement of the boom of each line segment based on the crane boom information and the operation line of the equipment to be operated.

[0065] In some embodiments, Figure 5 A schematic diagram of a process for determining a plurality of line segments and a safe range of boom movement for each line segment provided in an embodiment of the present invention, wherein the process of determining a plurality of line segments and a safe range of boom movement for each line segment includes steps S71 to S72:

[0066] Step S71: Based on the substation high-voltage cable information and the operating line of the equipment to be operated, the risk path segments affected by multiple high-voltage cables and the safe path segments affected by multiple high-voltage cables in the operating line are obtained.

[0067] In some embodiments, a path analysis model can be used based on the substation high-voltage cable information and the operating line of the equipment to be operated to obtain the risk path segments affected by multiple high-voltage cables in the operating line and the safe path segments affected by multiple high-voltage cables. The path analysis model is a Transformer model. The input of the path analysis model is the substation high-voltage cable information and the operating line of the equipment to be operated. The output of the path analysis model is the risk path segments affected by multiple high-voltage cables in the operating line and the safe path segments affected by multiple high-voltage cables.

[0068] The Transformer model is a deep learning model based on the self-attention mechanism. It uses multi-head self-attention layers to capture long-range dependencies in input sequences. Through its encoder and decoder architecture, the Transformer model extracts and maps features from the input sequence and can be used to convert information such as spatial position sequences into classification results.

[0069] Risky paths affected by high-voltage cables are those segments of the operating route of the equipment being operated, as output by the path analysis model, that present safety hazards due to the horizontal overlap or vertical distance with the substation's high-voltage cables not meeting safety standards. These paths include the coordinates of the path's start and end points, the path length, the spatial coordinates and voltage level of the high-voltage cables, and the risk level.

[0070] The horizontal projection coincidence is the overlap ratio between the operating line and the cable laying path in the horizontal plane.

[0071] The vertical distance is the vertical space between the plane where the operating line is located and the cable.

[0072] Safe paths affected by high-voltage cables are low-risk paths within the equipment's operating route, as output by the path analysis model. These paths include the start and end coordinates, path length, spatial coordinates of the high-voltage cables, voltage level, and safety level.

[0073] Substation high-voltage cable information includes parameters such as the cable's spatial coordinates, voltage level, and laying method. The operating route for the equipment to be operated is a path sequence consisting of planar coordinate points. By combining these two, the spatial relationship between the path and the high-voltage cable can be quantitatively assessed. For example, if the horizontal projection of the operating route overlaps with the high-voltage cable or the vertical distance falls below a safety threshold, electromagnetic interference or physical contact risks may arise. A safe path segment, on the other hand, meets these distance requirements.

[0074] The Transformer model's self-attention mechanism efficiently captures the spatial dependencies between high-voltage cable information and the coordinate sequence of the operating route. By encoding features such as cable location and voltage level along with pathpoint coordinates into a vector sequence, the Transformer model can learn the risk association between different path segments and cables. The Transformer model's multi-head attention mechanism allows the model to focus on key features from different dimensions, such as vertical height from the cable and horizontal overlap. This allows the model to map the input sequence into a classification result of risky and safe path segments, thereby identifying risky areas along the operating route and outputting risky and safe path segments affected by multiple high-voltage cables.

[0075] In some embodiments, the path analysis model includes a spatial position matching layer, a risk analysis layer, and a path segment classification layer. The input of the spatial position matching layer is the substation high-voltage cable information and the operating line of the equipment to be operated. The output of the spatial position matching layer is the voltage level association table of the line and the high-voltage cable, vertical distance data, and horizontal projection overlap area. The input of the risk analysis layer is the voltage level association table of the line and the high-voltage cable, vertical distance data, and horizontal projection overlap area. The output of the risk analysis layer is the risk score sequence, risk impact type, and path point importance index of each position on the line. The input of the path segment classification layer is the risk score sequence, risk impact type, and path point importance index of each position on the line. The output of the path segment classification layer is the risk path segment affected by multiple high-voltage cables in the operating line and the safe path segment affected by multiple high-voltage cables.

[0076] The voltage level association table of the high-voltage cable can record the voltage level of the high-voltage cable corresponding to each position of the operating line. Through the voltage level association table of the high-voltage cable, the rated voltage values ​​of the high-voltage cables near different sections on the line can be clearly defined.

[0077] The vertical distance data is a set of distance values ​​between each location point on the operating line and the high-voltage cable in the vertical direction.

[0078] The horizontal projection overlap area is the area where the projection of the operating line on the horizontal plane overlaps with the projection of the high-voltage cable on the horizontal plane.

[0079] The risk score sequence for each position on the operation line refers to the sequence formed by the risk scores assessed for each position point on the line according to the order of the operation line.

[0080] The risk impact type is the specific category of risk that each location point on the operation line may be exposed to. The risk impact types include high-voltage discharge, electromagnetic interference, etc.

[0081] The path point importance index is a numerical indicator used to measure the importance of each location point on the operation route during the operation process.

[0082] The spatial location matching layer extracts voltage level correlations, vertical distances, and horizontal projection overlap areas from the substation's high-voltage cable information and the operating lines. The risk analysis layer generates risk score sequences, risk types, and location importance coefficients for each line location. The path segment classification layer classifies risky and safe path segments. This layered approach allows for a more precise analysis of the impact of high-voltage cables on operating lines. By constructing multiple layers, the model's modularity is enhanced, allowing each layer to focus on optimizing specific functions. This in turn enhances the accuracy of the system's risk assessments in complex electrical environments and improves overall processing efficiency.

[0083] Step S72: determine multiple line segments and the safe movement range of the boom of each line segment based on the risk path segments affected by the multiple high-voltage cables, the safe path segments affected by the multiple high-voltage cables, the crane boom information, the multiple electrical risk points, and the multiple electrical safety points.

[0084] The crane is a mobile crane, and the crane includes a truck crane and a crawler crane. Figure 6 A schematic diagram of a crane provided in accordance with an embodiment of the present invention.

[0085] Crane boom information is used to describe the physical characteristics and motion parameters of the crane boom. Crane boom information includes boom length parameters, angular range of motion, maximum lifting capacity, boom cross-sectional dimensions, and boom material parameters.

[0086] By combining the crane boom information, the three-dimensional spatial activity boundaries of the boom in different line segments can be calculated to ensure that the boom maintains a safe distance from high-voltage cables, live equipment in the station, etc. during operation, while also meeting the boom's own mechanical performance limitations.

[0087] In some embodiments, a segment planning model can be used to determine multiple line segments and the safe range of boom movement for each line segment based on the risk path segments affected by the multiple high-voltage cables, the safe path segments affected by the multiple high-voltage cables, the crane boom information, the multiple electrical risk points, and the multiple electrical safety points. The segment planning model is a deep neural network model. The input of the segment planning model is the risk path segments affected by the multiple high-voltage cables, the safe path segments affected by the multiple high-voltage cables, the crane boom information, the multiple electrical risk points, and the multiple electrical safety points. The output of the segment planning model is multiple line segments and the safe range of boom movement for each line segment.

[0088] A deep neural network model includes a deep neural network (DNN). A deep neural network may include multiple processing layers, where each processing layer is composed of multiple neurons, and each neuron is capable of performing a matrix transformation on data.

[0089] Line segments are path units that are segmented using the segment planning model based on the impact of high-voltage cables and the safety risk level. A line segment includes the coordinates of the segment's start and end points, the segment's length, and its risk level. For example, a line segment with start coordinates of (5 meters, 10 meters) and end coordinates of (15 meters, 10 meters) corresponds to a high-risk path segment affected by high-voltage cables. The risk level is high, indicating that the spatial relationship between the segment and the high-voltage cables poses a safety hazard.

[0090] The boom's safe range of motion, output by the segment planning model, defines the safe three-dimensional boundary of the crane's boom's movement when performing operations on a specific line segment. This range includes the horizontal swing angle range, vertical elevation angle range, rotation angle range, and minimum safe distance from surrounding live equipment within the station. For example, within a certain safe path segment, the boom's elevation angle range is 30° to 60°, its rotation angle range is 0° to 270°, and its minimum safe distance from surrounding live equipment within the station is 3 meters.

[0091] Deep neural networks can transform data such as risky path segments, safe path segments affected by multiple high-voltage cables, crane boom information, multiple electrical risk points, and multiple electrical safety points into physically meaningful numerical features. Through layer-by-layer computation within hidden layers, the deep neural network automatically learns the spatial distance constraints between the operating route and high-voltage cables, the relationship between the crane boom's physical parameters and electrical safety distances, and the impact of electrical risk points and safety points on path safety. Based on the optimized parameters from training, the output layer divides the operating route into multiple segments according to power safety standards and generates boom safe range parameters for each segment that meet the crane's mechanical performance and electrical safety requirements.

[0092] Step S8: providing an early warning for the equipment to be operated based on the operating route of the equipment to be operated and the safe range of movement of the boom in each route segment.

[0093] When the operating route of the equipment to be operated and the safe range of movement of the boom in each route segment are determined, the system needs to monitor the spatial position and movement status of the boom on the operating route in real time, and compare the real-time coordinates and angle parameters of the boom with the safe range parameters of the corresponding route segment, such as three-dimensional space boundaries, angle limits, and safety distance thresholds. If the boom position or movement parameters exceed the safe range boundary, the early warning mechanism is triggered and sound and light signals are used to prompt the risk, and the specific location is marked, thereby realizing early warning of the equipment to be operated.

[0094] Based on the same inventive concept, Figure 7 A schematic diagram of a safety distance warning system for a substation provided in an embodiment of the present invention, wherein the safety distance warning system for a substation includes:

[0095] Information acquisition module 71, used to obtain substation panoramic image information and substation design drawing information;

[0096] An equipment processing module 72 is configured to determine information about high-voltage cables in the substation and information about multiple live devices in the substation using an equipment processing model based on the substation panoramic image information and the substation design drawing information;

[0097] A graph construction module 73 is used to construct an equipment graph. The equipment graph includes multiple in-station powered equipment nodes and multiple edges between the in-station powered equipment nodes. The node features of the in-station powered equipment nodes are the information of each in-station powered equipment and the panoramic image information of the substation. The edges between the nodes are the distances between the in-station powered equipment.

[0098] a risk analysis module 74 for processing the equipment map based on a first graph neural network to determine a plurality of electrical risk point information and a plurality of electrical safety point information;

[0099] A second map construction module 75 is configured to construct a road map based on the plurality of electrical risk points and the plurality of electrical safety points;

[0100] A route planning module 76 is configured to process the road map based on a second graph neural network to determine an operating route for the equipment to be operated;

[0101] A safety range determination module 77 is configured to determine a plurality of line segments and a safe range of movement of the boom of each line segment based on the crane boom information and the operation line of the equipment to be operated;

[0102] The early warning execution module 78 is used to issue an early warning to the equipment to be operated based on the operation line of the equipment to be operated and the safe range of movement of the boom of each line segment.

[0103] It should be noted that, in order to simplify the presentation of this specification and facilitate understanding of one or more embodiments of the invention, the foregoing descriptions of the embodiments of this specification sometimes combine multiple features into a single embodiment, figure, or description thereof. However, this disclosure method does not imply that the subject matter of this specification requires more features than those recited in the claims. In fact, an embodiment may have fewer features than all of the features of a single disclosed embodiment.

[0104] Finally, it should be understood that the embodiments described in this specification are intended only to illustrate the principles of the embodiments of this specification. Other variations may also fall within the scope of this specification. Therefore, by way of example and not limitation, alternative configurations of the embodiments of this specification may be considered consistent with the teachings of this specification. Accordingly, the embodiments of this specification are not limited to the embodiments explicitly described and illustrated in this specification.

Claims

1. A safety distance early warning method for a substation, characterized in that: include: Obtain substation panoramic image information and substation design drawing information; Determine substation high-voltage cable information and multiple in-station live equipment information using a device processing model based on the substation panoramic image information and the substation design drawing information; Construct an equipment graph, which includes multiple in-station powered equipment nodes and multiple edges between them. The node features of the in-station powered equipment nodes are the information of each in-station powered equipment and the panoramic image information of the substation. The edges between the nodes are the distances between the in-station powered equipment. Processing the equipment map based on a first graph neural network to determine a plurality of electrical risk point information and a plurality of electrical safety point information; Building a road map based on the multiple electrical risk points and the multiple electrical safety points; Processing the road map based on a second graph neural network to determine an operating route for the equipment to be operated; Determining multiple line segments and a safe boom movement range of each line segment based on the crane boom information and the operation route of the equipment to be operated, wherein the determining multiple line segments and a safe boom movement range of each line segment based on the crane boom information and the operation route of the equipment to be operated includes: Based on the substation high-voltage cable information and the operating line of the equipment to be operated, a risk path segment affected by multiple high-voltage cables and a safe path segment affected by multiple high-voltage cables in the operating line are obtained; Determine a plurality of line segments and a safe range of movement of the boom of each line segment based on the risk path segments affected by the plurality of high-voltage cables, the safe path segments affected by the plurality of high-voltage cables, the crane boom information, the plurality of electrical risk points, and the plurality of electrical safety points; An early warning of the equipment to be operated is performed based on the operating route of the equipment to be operated and the safe activity range of the boom in each route segment.

2. The safety distance early warning method for a substation according to claim 1, characterized in that: The constructing of a road map based on the plurality of electrical risk points and the plurality of electrical safety points includes: A road map is constructed, wherein the road map includes multiple electrical risk nodes, multiple electrical safety nodes, and multiple edges between the nodes. The node features of the electrical risk nodes are the location and risk level of the electrical risk points, the node features of the electrical safety nodes are the location and safety level of the electrical safety points, and the panoramic image information of the substation. The edges between the nodes are the distances between the nodes.

3. The safety distance early warning method for a substation according to claim 1, characterized in that: The device processing model is a convolutional neural network.

4. A safety distance warning system for a substation, characterized in that: include: Information acquisition module, used to obtain substation panoramic image information and substation design drawing information; An equipment processing module, configured to determine information about high-voltage cables in the substation and information about multiple live devices in the substation using an equipment processing model based on the substation panoramic image information and the substation design drawing information; A graph construction module is used to construct an equipment graph. The equipment graph includes multiple in-station live equipment nodes and multiple edges between the in-station live equipment nodes. The node features of the in-station live equipment nodes are the information of each in-station live equipment and the panoramic image information of the substation. The edges between the nodes are the distances between the in-station live equipment. a risk analysis module, configured to process the equipment map based on a first graph neural network to determine a plurality of electrical risk point information and a plurality of electrical safety point information; A second map construction module is configured to construct a road map based on the plurality of electrical risk points and the plurality of electrical safety points; a route planning module, configured to process the road map based on a second graph neural network to determine an operating route for the equipment to be operated; A safety range determination module is used to determine a plurality of line segments and a safe range of movement of the boom of each line segment based on the crane boom information and the operation line of the equipment to be operated. The safety range determination module is further used to: Based on the substation high-voltage cable information and the operating line of the equipment to be operated, a risk path segment affected by multiple high-voltage cables and a safe path segment affected by multiple high-voltage cables in the operating line are obtained; Determine a plurality of line segments and a safe range of movement of the boom of each line segment based on the risk path segments affected by the plurality of high-voltage cables, the safe path segments affected by the plurality of high-voltage cables, the crane boom information, the plurality of electrical risk points, and the plurality of electrical safety points; The early warning execution module is used to issue an early warning to the equipment to be operated based on the operating line of the equipment to be operated and the safe activity range of the boom of each line segment.

5. The safety distance warning system for a substation according to claim 4, characterized in that: The second graph construction module is further used for: A road map is constructed, wherein the road map includes multiple electrical risk nodes, multiple electrical safety nodes, and multiple edges between the nodes. The node features of the electrical risk nodes are the location and risk level of the electrical risk points, the node features of the electrical safety nodes are the location and safety level of the electrical safety points, and the panoramic image information of the substation. The edges between the nodes are the distances between the nodes.

6. The safety distance warning system for a substation according to claim 4, characterized in that: The device processing model is a convolutional neural network.

7. An electronic device, characterized in that: include: processor; Memory; And a computer program; wherein, the computer program is stored in the memory and is configured to be executed by the processor to implement the safety distance warning method for the substation according to any one of claims 1 to 3.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the safety distance warning method for a substation according to any one of claims 1 to 3 is implemented.

Images