Power transmission line intelligent line selection dynamic planning method based on segmentation consideration

Abstract

The invention discloses a power transmission line intelligent line selection dynamic planning method based on segmentation consideration, and relates to the field of electric power engineering, and the method comprises the steps: generating a potential path point sequence through a large language model according to starting point information, terminal point information and environment data, the starting point information and the terminal point information are used for building an overhead power transmission line, and the environment data are used for building a potential path point sequence; the environment data is used for indicating environment information of a target area where the overhead transmission line is located; performing two-dimensional path search on the potential path point sequence on a ground object layer through a target algorithm to generate an initial path of the overhead transmission line; performing three-dimensional elevation optimization on path nodes in the initial path according to digital elevation data to obtain a final path of the overhead transmission line; by adopting the scheme, the problems that manual line selection consumes time and labor, the planning efficiency is low and the line scheme quality is poor due to numerous influence factors and mutual restriction in the line selection planning process of the overhead transmission line are solved.

Description

Technical Field

[0001] This application relates to the field of power engineering, and more particularly, to an intelligent route selection dynamic programming method for transmission lines based on segmented consideration. Background Art

[0002] The route selection and planning of overhead transmission lines is a key link in power engineering design, and it is necessary to minimize project costs and environmental impacts as much as possible while meeting safety specifications. The traditional route selection process of transmission lines mainly relies on manual experience: designers plan a line corridor that avoids obstacles between the starting point and the ending point based on topographic maps, satellite images and other data. When planning the line, various ground object factors (such as roads, rivers, buildings, vegetation, etc.) and terrain elevation factors (terrain undulation) need to be considered. The route should not only minimize the line length as much as possible, but also avoid passing through restricted areas (such as densely built areas, nature reserves), and minimize intersections with linear facilities such as roads and railways. Due to numerous influencing factors and their mutual constraints, manual route selection often requires repeated weighing of multiple plans, which is time-consuming and laborious.

[0003] In view of the problems in the related art that due to numerous influencing factors and their mutual constraints in the route selection and planning process of overhead transmission lines, manual route selection is time-consuming and laborious, with low planning efficiency and poor line scheme quality, no effective solution has been proposed yet. Summary of the Invention

[0004] An embodiment of this application provides an intelligent route selection dynamic programming method for transmission lines based on segmented consideration, so as to at least solve the problems in the related art that due to numerous influencing factors and their mutual constraints in the route selection and planning process of overhead transmission lines, manual route selection is time-consuming and laborious, with low planning efficiency and poor line scheme quality.

[0005] According to an aspect of the embodiment of this application, there is provided an intelligent route selection dynamic programming method for transmission lines based on segmented consideration, including: generating a sequence of potential path points through a large language model according to starting point information, ending point information and environmental data, where the starting point information and the ending point information are used to build an overhead transmission line, and the environmental data is used to indicate the environmental information of the target area where the overhead transmission line is located; performing two-dimensional path search on the sequence of potential path points through a target algorithm on a ground object layer to generate an initial path of the overhead transmission line; and performing three-dimensional elevation optimization on path nodes in the initial path according to digital elevation data to obtain a final path of the overhead transmission line.

[0006] In an exemplary embodiment, a two-dimensional path search is performed on the potential path point sequence on the ground feature layer by a target algorithm to generate an initial path of the overhead transmission line, including: initializing and generating an exploration graph, where the starting point information is the root node of the exploration graph; determining a plurality of candidate points in the potential path point sequence whose distance from the root node is less than a first preset threshold as an initial priority sequence, where the plurality of candidate points are used to indicate a plurality of path directions of the initial path; circularly expanding the exploration graph by a potential point priority mechanism and / or a random node sampling method until the nodes in the exploration graph are connected to the path nodes corresponding to the end point information or the number of times of circular expansion reaches a preset iteration number; extracting a path sequence from the exploration graph according to the root node and the path nodes corresponding to the end point information, and generating the initial path according to the path sequence.

[0007] In an exemplary embodiment, circularly expanding the exploration graph by a potential point priority mechanism and / or a random node sampling method includes: when it is determined that there is a first path node in the potential path point sequence that has not been added to the exploration graph, determining a second path node with the highest priority among the first path nodes; determining whether the sub-path between a newly added third path node in the exploration graph and the second path node conforms to the line planning rule, where the line planning rule is used to indicate that the initial path does not overlap with the area where obstacles are located, and the sub-path is a straight line path; when it is determined that it conforms to the line planning rule, adding the second path node to the exploration graph; when it is determined that there is no first path node or it is determined that it does not conform to the line planning rule, randomly selecting a fourth path node in a first sub-region from the third path node to the path nodes corresponding to the end point information by the random node sampling method, and when it is determined that the sub-path between the third path node and the fourth path node conforms to the line planning rule, adding the fourth path node to the exploration graph, where the target region includes the first sub-region.

[0008] In an exemplary embodiment, the method further includes: when the number of failure times of expanding the exploration graph by the potential point priority mechanism reaches a second preset threshold, increasing the sampling frequency of the random node sampling method.

[0009] In an exemplary embodiment, before performing three-dimensional elevation optimization on the path nodes in the initial path according to the digital elevation data to obtain the final path of the overhead transmission line, the method further includes: identifying non-critical nodes among the multiple path nodes in the initial path to obtain a plurality of non-critical nodes, wherein the sub-path where the non-critical node is located does not overlap with the area where the obstacle is located, and the sub-path is a straight path; deleting the plurality of non-critical nodes from the initial path to obtain a simplified path; performing three-dimensional elevation optimization on the path nodes in the simplified path according to the digital elevation data to obtain the final path.

[0010] In an exemplary embodiment, performing three-dimensional elevation optimization on the path nodes in the simplified path according to the digital elevation data to obtain the final path includes: determining the elevation of a plurality of fifth path nodes in the simplified path according to the digital elevation data; for a sixth path node among the plurality of fifth path nodes, searching for a local terrain high point in a second sub-region where the sixth path node is located, and updating the sixth path node to the local terrain high point to obtain an updated plurality of fifth path nodes, wherein the local terrain high point is a legal construction area; determining the final path according to the updated plurality of fifth path nodes.

[0011] In an exemplary embodiment, determining the final path according to the updated plurality of fifth path nodes includes: sequentially determining the horizontal distance between two adjacent fifth path nodes among the updated plurality of fifth path nodes, and determining whether there is a target horizontal distance greater than a third preset threshold among the plurality of horizontal distances; in the case where it is determined that there is a target horizontal distance greater than the third preset threshold, adding one or more intermediate path nodes between the two fifth path nodes corresponding to the target horizontal distance to obtain a plurality of seventh path nodes, so that the horizontal distance between any two adjacent seventh path nodes is less than or equal to the third preset threshold; sequentially determining the slope value of the tower connection line corresponding to two adjacent seventh path nodes among the plurality of seventh path nodes, and determining whether there is a target slope value greater than a fourth preset threshold among the plurality of slope values; in the case where it is determined that there is a target slope value greater than the fourth preset threshold, adjusting the heights of the two seventh path nodes corresponding to the target slope value to obtain a plurality of eighth path nodes, so that the slope value corresponding to any two adjacent eighth path nodes is less than or equal to the fourth preset threshold; determining the final path according to the plurality of eighth path nodes.

[0012] According to another aspect of the embodiments of the present application, there is also provided a multi-stage line planning device based on large model recommendation and terrain optimization, including: a generation module, configured to generate a sequence of potential path points through a large language model according to starting point information, ending point information, and environmental data, wherein the starting point information and the ending point information are used to build an overhead transmission line, and the environmental data is used to indicate the environmental information of the target area where the overhead transmission line is located; a search module, configured to perform a two-dimensional path search on the sequence of potential path points through a target algorithm on a ground feature layer to generate an initial path of the overhead transmission line; and an optimization module, configured to perform three-dimensional elevation optimization on the path nodes in the initial path according to digital elevation data to obtain a final path of the overhead transmission line.

[0013] According to yet another aspect of the embodiments of the present application, there is also provided a computer-readable storage medium storing a computer program, wherein the computer program is configured to execute the above-mentioned intelligent route selection dynamic programming method for transmission lines based on segmented consideration when running.

[0014] According to yet another aspect of the embodiments of the present application, there is also provided an electronic device including a memory, a processor, and a computer program stored on the memory and executable on the processor, wherein the above-mentioned processor executes the above-mentioned intelligent route selection dynamic programming method for transmission lines based on segmented consideration through the computer program.

[0015] According to yet another aspect of the embodiments of the present application, there is also provided a computer program product including a computer program, and the steps of the methods described in various embodiments of the present application are implemented when the computer program is executed by a processor.

[0016] In an embodiment of the present application, an intelligent route selection dynamic programming method for transmission lines based on segmented consideration is proposed. This method first processes the starting point information, ending point information, and environmental data through a large language model to generate a sequence of potential path points. Among them, the starting point information and ending point information are respectively used to indicate the starting point and ending point of the overhead transmission line to be built, and the environmental data is used to indicate the environmental information in the target area where the overhead transmission line is located. Then, the target algorithm is used to recommend potential path points and quickly grow the path to obtain an initial path. Finally, the generated initial path is fine-tuned in terms of elevation and nodes are supplemented, thereby completing the fine optimization of the line and obtaining the final path. By adopting the above solution, the environmental understanding ability of the large language model and the path search algorithm are combined, and the digital elevation model is introduced in stages for optimization, realizing the full-automatic route selection planning of overhead lines. Through this technical solution, the route selection efficiency can be significantly improved, the dependence on manual experience can be reduced, and a safer, more reliable, economical and reasonable line scheme can be generated. Furthermore, it solves the problem in the related technology that due to the numerous and mutually restrictive influencing factors in the route selection planning process of overhead transmission lines, manual route selection is time-consuming and laborious, the planning efficiency is low, and the quality of the line scheme is poor. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The accompanying drawings herein are incorporated into the specification and form a part of the specification, showing embodiments consistent with the present application, and are used together with the specification to explain the principles of the present application.

[0018] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, for those of ordinary skill in the art, other drawings can also be obtained based on these drawings without creative efforts.

[0019] Figure 1 is a flowchart of an optional intelligent route selection dynamic programming method for transmission lines based on segmented consideration according to an embodiment of the present application; Figure 2 is a schematic flow diagram of an optional intelligent route selection dynamic programming method for transmission lines based on segmented consideration according to an embodiment of the present application; Figure 3 is a structural block diagram of an optional multi-stage line planning device based on large model recommendation and terrain optimization according to an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0020] To enable those skilled in the art to better understand the solution of this application, the following will clearly and completely describe the technical solution in the embodiments of this application in conjunction with the accompanying drawings in the embodiments of this application. Obviously, the described embodiments are only a part of the embodiments of this application, rather than all the embodiments. Based on the embodiments in this application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the scope of protection of this application.

[0021] It should be noted that the terms "first", "second", etc. in the description and claims of this application and the above-mentioned accompanying drawings are used to distinguish similar objects, and do not necessarily need to be used to describe a specific order or sequence. It should be understood that such data can be interchanged under appropriate circumstances so that the embodiments of this application described here can be implemented in an order other than those illustrated or described here. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or device that includes a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or are inherent to these processes, methods, products, or devices.

[0022] In related technologies, with the development of Geographic Information System (GIS) technology, automatic transmission line route selection solutions based on GIS have emerged. For example, there is a method of converting environmental constraints into grid raster data and applying the shortest path algorithm to search for the optimal path. Some research uses intelligent optimization algorithms to assist in route selection: such as combining the ant colony algorithm with GIS, generating a comprehensive cost layer by fusing multiple geographic layers through the fuzzy analytic hierarchy process on the ArcGIS platform, and then searching for the path with the lowest comprehensive cost. There is also research on improving the classic Dijkstra and A* algorithms, reducing invalid operations by setting an efficient priority search area, thereby improving the path search efficiency in a large-scale environment.

[0023] Improving the RRT (Rapidly-exploring Random Tree) path planning method (such as RRT*, Informed-RRT*) can achieve path optimization through sampling optimization, but still requires a large number of exploratory samplings in a vast space, consuming a large amount of computing resources. In addition, the processing methods for typical features such as buildings, roads, and rivers are relatively rough, easily resulting in the path crossing obstacles or nodes falling in non-constructible areas. Some deep learning-based methods are only used for classification and recognition and do not really participate in the path generation process.

[0024] In recent years, the rise of large language models (LLMs) has provided new ideas for path planning. Some works have attempted to combine the global reasoning ability of LLMs with traditional path algorithms, such as collaborating LLMs with the A* algorithm to plan paths, and using the macroscopic understanding of the environment by LLMs to guide the search direction of the algorithm. These methods have improved the efficiency and intelligence level of path planning to a certain extent.

[0025] Most existing transmission line planning tools focus on path optimization on a two-dimensional plane. After generating a preliminary line, manual three-dimensional elevation checking and tower location selection are carried out. It is usually necessary to ensure that the towers (line support points) avoid buildings and are preferably located at higher elevations to increase the clearance between the conductor and the ground. At the same time, the tower spacing should not be too large to control the sag and mechanical strength of the conductor. However, manual line selection is time-consuming and laborious and overly relies on manual experience.

[0026] To solve the technical problems existing in the related technologies, in this embodiment, an intelligent line selection dynamic programming method for transmission lines based on segmented consideration is provided. Figure 1 It is a flowchart of an optional intelligent line selection dynamic programming method for transmission lines based on segmented consideration according to an embodiment of the present application. The process includes the following steps S102 - S106: Step S102, through a large language model, generate a sequence of potential path points according to the starting point information, ending point information, and environmental data, where the starting point information and the ending point information are used to build an overhead transmission line, and the environmental data is used to indicate the environmental information of the target area where the overhead transmission line is located; Step S104, perform two-dimensional path search on the sequence of potential path points through a target algorithm on the ground feature layer to generate an initial path of the overhead transmission line; Step S106, perform three-dimensional elevation optimization on the path nodes in the initial path according to the digital elevation data to obtain the final path of the overhead transmission line.

[0027] Through the above steps, a dynamic programming method for intelligent route selection of transmission lines based on segmented consideration is proposed. This method first processes the starting point information, ending point information, and environmental data through a large language model to generate a sequence of potential path points. Among them, the starting point information and the ending point information are respectively used to indicate the starting point and the ending point of the overhead transmission line to be built, and the environmental data is used to indicate the environmental information in the target area where the overhead transmission line is located. Then, the target algorithm is used to recommend potential path points and quickly grow the path to obtain the initial path. Finally, the generated initial path is finely adjusted in elevation and nodes are supplemented to complete the fine optimization of the line and obtain the final path. By adopting the above solution, the environmental understanding ability of the large language model and the path search algorithm are combined, and the digital elevation model is introduced in stages for optimization, realizing the fully automatic route selection and planning of overhead lines. Through this technical solution, the route selection efficiency can be significantly improved, the dependence on manual experience can be reduced, and a safer, more reliable, economical and reasonable line plan can be generated. Furthermore, it solves the problems in the related technologies that due to the numerous influencing factors and mutual constraints in the route selection and planning process of overhead transmission lines, manual route selection is time-consuming and laborious, the planning efficiency is low, and the quality of the line plan is poor.

[0028] In an exemplary embodiment, a two-dimensional path search is performed on the sequence of potential path points through a target algorithm on the ground feature layer to generate the initial path of the overhead transmission line, including: initializing and generating an exploration graph, where the starting point information is the root node of the exploration graph; determining a plurality of candidate points in the sequence of potential path points whose distance from the root node is less than a first preset threshold as an initial priority sequence, where the plurality of candidate points are used to indicate a plurality of path directions of the initial path; circularly expanding the exploration graph through a potential point priority mechanism and / or a random node sampling method until the nodes in the exploration graph are connected to the path nodes corresponding to the ending point information or the number of times of the circular expansion reaches a preset iteration number; extracting a path sequence from the exploration graph according to the root node and the path nodes corresponding to the ending point information, and generating the initial path according to the path sequence.

[0029] This embodiment elaborates in detail a method for initial path planning of overhead transmission lines using large model recommendation and Rapidly-exploring Random Tree (RRT) algorithm. This method combines intelligent recommendation and random exploration, aiming to efficiently generate a preliminary line path that avoids obstacles and is as short and straight as possible.

[0030] 1. Initialize the exploration graph: The construction of the exploration graph starts with the setting of the root node, which is the starting point information of the transmission line. The starting point information includes the starting point coordinates (X_start, Y_start), as well as possible additional attributes such as altitude and terrain type.

[0031] Construct an empty exploration graph structure. Add the starting node to the tree as the root node of the tree, and label it as RRT_tree = {root}.

[0032] 2. Generate a sequence of potential path points: Call the large language model DeepSeek to generate a series of potential path points based on the starting point, ending point, and environmental information. These points are intelligently recommended according to factors such as terrain and obstacle distribution, and are used to indicate the most likely path directions.

[0033] The sequence of potential path points preliminarily determines the search priority and direction through the relative position relationship with the starting point. For example, for points P1, P2, P3... in the sequence, their distance and azimuth information from the starting point will guide the expansion direction of the exploration graph.

[0034] 3. Potential point priority mechanism and random node sampling: Initialization of the priority sequence: Select multiple candidate points from the sequence of potential path points whose distance from the root node is less than the first preset threshold (e.g., 100 meters) to form an initial priority sequence. These points will be given priority in the search process and used to guide the growth direction of the exploration graph.

[0035] Loop expansion: Before the exploration graph reaches the end point or the number of iterations does not exceed the preset value N_max, the algorithm performs loop expansion: Priority expansion: According to the priority sequence, select the nearest candidate point P_near as the expansion target. Find the nearest node N_near to P_near in the tree, and try to sample a new node P_new from N_near to P_near and check for intersections with obstacles.

[0036] Collision detection: If the new node P_new and its connection line to N_near do not intersect with obstacles such as buildings and rivers, add P_new to the exploration graph and update its path to the root node.

[0037] Random expansion: When the potential point priority mechanism expansion is blocked (e.g., P_new intersects with obstacles), the algorithm switches to the random mode and randomly samples a node R_rand in the free space. Find the nearest node N_rand to R_rand in the tree, and try to generate a new node R_new and check for collisions.

[0038] Priority update: Whenever a new node is successfully added, dynamically update the priority order in the sequence of potential path points according to its position information to ensure that subsequent expansions are closer to the target path.

[0039] 4. Path Generation and Extraction: Endpoint Determination: If the node generated during the loop expansion process is within a preset connection distance (e.g., 10 meters) from the node corresponding to the endpoint information, it is determined that the path is completed and the expansion is terminated.

[0040] Path Sequence Extraction: Once the exploration graph reaches the endpoint, starting from the root node, trace back the shortest path to the endpoint and extract the path sequence, which is a series of connected pole tower node positions.

[0041] Initial Path Generation: According to the extracted path sequence, generate a preliminary overhead transmission line path, that is, the line layout represented on a two-dimensional plane.

[0042] This embodiment utilizes the combination of potential path points recommended by the large model and the exploration graph, skillfully integrating intelligent guidance and random exploration, thereby achieving efficient and intelligent preliminary path planning for overhead transmission lines. By giving priority to recommended points, it avoids the problems of low efficiency and path deviation that may be caused by blind random search; while random node sampling ensures the comprehensiveness and flexibility of the search, ensuring that a feasible path can be found even under complex environmental conditions. This method significantly improves the automation level and planning quality of path planning, providing a solid foundation for subsequent elevation optimization and line engineering design.

[0043] Optionally, the loop expansion of the exploration graph through the potential point priority mechanism and / or the random node sampling method includes: in the case where it is determined that there is a first path node in the potential path point sequence that has not been added to the exploration graph, determining the second path node with the highest priority among the first path nodes; determining whether the sub-path between the newly added third path node and the second path node in the exploration graph conforms to the line planning rules, where the line planning rules are used to indicate that the initial path does not overlap with the area where the obstacle is located, and the sub-path is a straight-line path; in the case where it is determined that it conforms to the line planning rules, adding the second path node to the exploration graph; in the case where it is determined that there is no first path node or it does not conform to the line planning rules, randomly select a fourth path node in the first sub-region from the third path node to the path node corresponding to the endpoint information through the random node sampling method, and in the case where it is determined that the sub-path between the third path node and the fourth path node conforms to the line planning rules, adding the fourth path node to the exploration graph, where the target area includes the first sub-region.

[0044] In this embodiment, we have explained in detail how to perform cyclic expansion on the Rapidly-exploring Random Tree (RRT) through the potential point priority mechanism and the random node sampling method to efficiently generate a preliminary path that conforms to the route planning rules. The key point of this process lies in how to flexibly and intelligently expand the exploration graph to ensure that the path not only avoids obstacles but also quickly advances towards the end point.

[0045] 1. Expansion with potential point priority: Priority check: First, check whether there are still the first path nodes in the potential path point sequence that have not been added to the exploration graph. These nodes were previously recommended by the large model to guide the direction of path search.

[0046] Determine the point with the highest priority: If such points exist, the algorithm will select the point with the highest priority (the second path node), and this priority is determined based on factors such as the distance from the starting point, the relative positions between the recommended points, and the route planning rules.

[0047] Generate and verify the sub-path: The algorithm attempts to generate the straight-line path (sub-path) between the nearest third path node (the latest added node in the exploration graph) and the second path node, and checks whether this sub-path overlaps with any obstacle areas, including roads, rivers, buildings, etc.

[0048] Add the sub-path to the exploration graph: If the sub-path conforms to the route planning rules, that is, it does not overlap with obstacles, the second path node will be added to the exploration graph, expanding the scope of the tree, and at the same time updating the best path from the starting point to the current point.

[0049] 2. Expansion with random node sampling: When potential points are exhausted or do not conform to the rules: If all points in the potential path point sequence have been added to the exploration graph, or all sub-paths between potential points and nodes in the exploration graph do not conform to the route planning rules, the algorithm switches to the random node sampling mode.

[0050] Set the target area: Define a first sub-area from the third path node (the nearest node in the exploration graph) to the node corresponding to the end point information. This area is the key area currently searched by the algorithm.

[0051] Randomly select the fourth path node: Randomly select a point (the fourth path node) within the first sub-area, and check whether the sub-path from the third path node to this random point conforms to the route planning rules.

[0052] Verify and add the sub-path: If this randomly generated sub-path passes the obstacle detection, that is, it does not overlap with any obstacle areas, then the fourth path node will be added to the exploration graph to continue the tree expansion process.

[0053] 3. Loop condition and termination condition: Loop condition: As long as the exploration graph has not reached the node corresponding to the end information and the number of loops does not exceed the preset maximum number of iterations, the loop expansion will continue.

[0054] Termination condition: Once the node in the exploration graph is within the preset connection distance from the node corresponding to the end information, or the number of loops reaches the preset maximum value, the loop stops, and at this time, it is considered that the path planning is completed.

[0055] Summary: By combining the potential point first mechanism and the random node sampling method, the expansion process of the exploration graph achieves a balance between intelligence and flexibility. The potential point first expansion accelerates the path towards the end point, reducing blind search; while random sampling ensures the algorithm's adaptability to complex environments and avoids local optimal traps in path planning. This expansion strategy enables the exploration graph to grow efficiently. The finally generated path not only avoids all obstacles but also is as close to a straight line as possible, shortening the line distance and improving the intelligence and efficiency of overhead transmission line planning.

[0056] Optionally, the method further includes: when the number of failures in expanding the exploration graph through the potential point first mechanism reaches a second preset threshold, increasing the sampling frequency of the random node sampling method.

[0057] In this embodiment, we introduce how to dynamically adjust the algorithm strategy and increase the usage frequency of the random node sampling method when encountering difficulties in expanding the exploration graph based on the potential point first mechanism, so as to overcome the problem of limited local search and ensure the continuous progress and final success of path planning.

[0058] 1. Limitations of the potential point first mechanism: In the initial path planning of overhead transmission lines, although the potential point first mechanism is adopted to guide the expansion direction of the exploration graph, due to the complexity of the terrain and the uncertainty of obstacle distribution, there may be situations where consecutive expansion attempts based on potential points fail. The reasons for failure may include that the potential point is near an obstacle, the terrain is too rugged to directly connect lines, or other planning rules limit the expansion.

[0059] 2. Failure times threshold: To prevent the algorithm from falling into an infinite loop or being overly restricted by the potential point mechanism, this embodiment sets a second preset threshold for monitoring the number of consecutive failures. This threshold can be an empirical value obtained from experimental tests, such as 10 consecutive expansions based on potential points have failed.

[0060] 3. Increasing the random sampling frequency: When the number of failures in exploring graph expansion through the potential point priority mechanism reaches the second preset threshold, the algorithm automatically adjusts the strategy and increases the usage frequency of the random node sampling method.

[0061] This means that in the next loop expansion, the algorithm will adopt more random node sampling instead of relying solely on the guidance of potential points. Random sampling can explore potential paths in a broader free space, improving the diversity and possibility of the search. Especially in complex terrains and dense obstacle environments, it can help the algorithm break out of the local optimal state and discover new feasible paths.

[0062] 4. Restore the potential point priority mechanism: When the random node sampling successfully generates a new node and adds it to the RFT tree, the algorithm will re-evaluate the current exploration graph structure and environmental conditions. If the expansion of the potential point priority mechanism becomes feasible or effective again, the algorithm will restore the combined way of the original proportion of potential point priority expansion and random node sampling.

[0063] In this way, the algorithm can adopt a more flexible search strategy when the potential point expansion is blocked. Once the obstacles are cleared or the path conditions improve, it can reuse the intelligent recommended potential points for directional search to achieve efficient and intelligent path planning.

[0064] In summary, this embodiment enhances the path planning ability of the exploration graph in complex environments by dynamically adjusting the usage frequencies of the potential point priority mechanism and the random node sampling method, enabling the algorithm to more robustly handle various challenges and ensuring that even under adverse conditions, a preliminary overhead transmission line path that meets all planning rules can be finally generated. This method not only improves the efficiency of path planning but also enhances the robustness and generalization ability of the algorithm, and is one of the important means to improve the automation level in the field of intelligent path planning.

[0065] Optionally, before optimizing the three-dimensional elevation of the path nodes in the initial path according to the digital elevation data to obtain the final path of the overhead transmission line, the method further includes: identifying non-critical nodes among the multiple path nodes in the initial path to obtain multiple non-critical nodes, where the sub-path where the non-critical node is located does not overlap with the area where the obstacle is located, and the sub-path is a straight-line path; deleting the multiple non-critical nodes from the initial path to obtain a simplified path; and optimizing the three-dimensional elevation of the path nodes in the simplified path according to the digital elevation data to obtain the final path.

[0066] In this embodiment, we explored how to simplify the initial path of an overhead transmission line before performing 3D elevation optimization using digital elevation data, in order to eliminate non-critical nodes, reduce algorithm complexity, and ensure that the path after simplification still meets safety and regulatory requirements.

[0067] 1. Identification of non-critical nodes: After obtaining the initial path, this path may contain a large number of path nodes. Some of these nodes are path inflection points, which play a key role in obstacle avoidance and path direction; while others are merely by-products during the path growth process and have little or no impact on the final path.

[0068] Path analysis and sub-path inspection: The algorithm checks each straight-line sub-path (line segment) formed between every pair of adjacent nodes on the initial path and evaluates whether this sub-path overlaps with any obstacle areas (such as buildings, rivers, roads, etc.). If the sub-path does not overlap with obstacles and the straight-line segment of the path can directly pass through free space, then the path nodes on this straight-line segment are considered non-critical nodes.

[0069] 2. Marking of non-critical nodes: Mark all path nodes that meet the above conditions as non-critical nodes, and these nodes will be deleted before subsequent elevation optimization.

[0070] Path simplification: Node deletion and path reconstruction: After identifying all non-critical nodes, the algorithm deletes these nodes from the initial path, retaining only the critical nodes (i.e., path inflection points), as well as the starting point and the ending point.

[0071] Path reconstruction: After deleting non-critical nodes, straight-line paths are re-formed between adjacent critical nodes. The simplified path obtained in this way only contains nodes that have a decisive impact on the final layout of the line, greatly reducing the computational complexity of subsequent elevation optimization.

[0072] 3. 3D elevation optimization: Introduction of DEM data: After obtaining the simplified path, digital elevation model (DEM) data is used to optimize the elevation of each node in the path. The goal is to place each tower node at a relatively flat and appropriately elevated position to meet the line's clearance requirements from the ground and erection specifications, and to avoid engineering problems caused by the line being too low or too high.

[0073] Node fine-tuning and insertion: The algorithm will fine-tune the elevation of each critical node in the simplified path to make it fall on a relatively high ground, and insert intermediate points when necessary to control the tower spacing and the slope of the path, ensuring that the line is both safe and economical in 3D space.

[0074] 4. Generation of the final path: The simplified path after elevation optimization is the final path of the overhead transmission line. This path not only avoids all obstacles in the two-dimensional plane but also fully considers the terrain factors in the three-dimensional space, making the line layout more reasonable and more in line with the actual engineering requirements and safety specifications.

[0075] Summary: Through the path simplification and three-dimensional elevation optimization processes described in the above embodiments, the present invention provides an efficient and accurate method for planning the path of an overhead transmission line. Path simplification can effectively reduce the consumption of computing resources, while three-dimensional elevation optimization ensures the safety and economy of the line during actual construction. The final path obtained is the optimal solution after multiple optimizations including intelligent recommendation, obstacle avoidance, and terrain adaptability. This method not only improves the efficiency and quality of automated route selection planning but also reduces the dependence on manual adjustment, providing strong support for the rapid and intelligent design of overhead transmission lines.

[0076] In an exemplary embodiment, performing three-dimensional elevation optimization on the path nodes in the simplified path according to the digital elevation data to obtain the final path includes: determining the altitude of a plurality of fifth path nodes in the simplified path according to the digital elevation data; for a sixth path node among the plurality of fifth path nodes, searching for a local terrain high point in a second sub-region where the sixth path node is located, and updating the sixth path node to the local terrain high point to obtain a plurality of updated fifth path nodes, where the local terrain high point is a legal construction area; and determining the final path according to the plurality of updated fifth path nodes.

[0077] In this embodiment, we have detailed how to use digital elevation data (Digital Elevation Model, DEM) to optimize the nodes of the overhead transmission line path in terms of three-dimensional elevation to obtain a final line path that fully takes into account the terrain and landforms. This process ensures that the line is not only reasonable in planar layout but also complies with engineering specifications in the vertical dimension, improving the safety and economy of the line layout.

[0078] 1. Determine the altitude of path nodes: Obtain the altitude information of all fifth path nodes in the simplified path (the path after removing non-critical nodes). This step is completed by querying the true elevation values corresponding to the coordinates of each node through DEM data, laying the foundation for subsequent elevation optimization.

[0079] 2. Search for local terrain high points and update nodes: Selection of target node: Select a node from the fifth path nodes as the sixth path node, which is usually the key tower position point that needs to be optimized in terms of elevation.

[0080] Definition of the second sub-region: Define a second sub-region around the sixth path node. The size and shape of this region can be set according to the terrain complexity and the requirements of the tower layout, and it is used to search for the best local terrain high point. For example, the sub-region can be a circular region centered on the sixth path node with a certain radius (such as 100 meters).

[0081] Search for the local terrain high point: Within the defined second sub-region, the algorithm searches for a point (local terrain high point) with the highest altitude and belonging to a legal construction area based on the DEM data. This involves analyzing the terrain information to determine which areas are suitable for tower construction, such as avoiding swamps, cliffs, or protected areas, etc.

[0082] Update of the node position: After finding the appropriate local terrain high point, update the position of the sixth path node to this high point, that is, raise the node vertically to obtain better terrain adaptability and a higher clearance of the conductor from the ground.

[0083] 3. Elevation optimization loop: The above steps are repeated for each fifth path node (sixth path node) in the simplified path to ensure that all tower position points have been appropriately elevation optimized. During the optimization process, it is necessary to continuously check the overall compliance of the updated path, including but not limited to tower spacing limits, slope requirements, etc.

[0084] Termination condition: When all nodes have completed elevation optimization, or when the upper limit of the preset optimization iteration times is reached, the algorithm stops execution.

[0085] 4. Determine the final path: After completing the elevation optimization of all fifth path nodes, the algorithm will reconstruct the entire path according to the updated node positions to form the final overhead transmission line path. This path not only avoids obstacles in the horizontal direction but also has been carefully adjusted in the vertical elevation to ensure that there is sufficient safety distance between the line and the ground and meets various engineering specification requirements.

[0086] This embodiment realizes the three-dimensional elevation optimization of the overhead transmission line path by combining DEM data and local search optimization strategies. This method not only ensures the construction feasibility of the line on complex terrains but also reduces the clearance problem of the conductor from the ground by raising the nodes to the local terrain high points, improving the safety and stability of the line operation. In addition, through the elevation optimization of each node, the economic rationality of the line under different terrain and landform conditions is also ensured, providing a higher level of technical support for automated path planning.

[0087] Based on the above steps, determining the final path according to the updated multiple fifth path nodes includes: sequentially determining the horizontal distances between two adjacent fifth path nodes among the updated multiple fifth path nodes, and determining whether there is a target horizontal distance greater than a third preset threshold among the multiple horizontal distances; in the case of determining that there is a target horizontal distance greater than the third preset threshold, adding one or more intermediate path nodes between the two fifth path nodes corresponding to the target horizontal distance to obtain multiple seventh path nodes, so that the horizontal distance between any two adjacent seventh path nodes is less than or equal to the third preset threshold; sequentially determining the slope values of the pole tower connection lines corresponding to two adjacent seventh path nodes among the multiple seventh path nodes, and determining whether there is a target slope value greater than a fourth preset threshold among the multiple slope values; in the case of determining that there is a target slope value greater than the fourth preset threshold, adjusting the heights of the two seventh path nodes corresponding to the target slope value to obtain multiple eighth path nodes, so that the slope values corresponding to any two adjacent eighth path nodes are less than or equal to the fourth preset threshold; determining the final path according to the multiple eighth path nodes.

[0088] In this embodiment, we deeply explore how to further adjust the path based on the updated fifth path nodes (i.e., the nodes after preliminary elevation optimization) to meet the engineering specifications of the pole tower spacing and the line slope, and finally determine the feasible path of the overhead transmission line. This process ensures the safety and economy of the line in actual construction by adding intermediate nodes and adjusting the node heights.

[0089] 1. Horizontal distance inspection and intermediate node addition: Horizontal distance measurement: First, measure the horizontal distances between two adjacent fifth path nodes in the updated path. This step aims to check whether the distance between the pole towers is too long and does not meet the engineering specification requirements.

[0090] Third preset threshold: Set a third preset threshold (for example, the maximum allowable pole tower spacing is 300 meters) to judge whether the horizontal distance between adjacent nodes exceeds this threshold.

[0091] Intermediate node addition: If it is detected that the target horizontal distance is greater than the third preset threshold, the algorithm will add one or more intermediate path nodes (seventh path nodes) between the two nodes to ensure that the horizontal distance between any two adjacent nodes is less than or equal to the third preset threshold. When adding nodes, the terrain elevation and obstacles also need to be considered.

[0092] Node position selection: The position selection of the intermediate node needs to combine DEM data to ensure that it is located in a legal construction area and is relatively flat to avoid unnecessary engineering difficulties and costs.

[0093] 2. Slope Value Inspection and Node Height Adjustment: Calculation of slope value: Next, the algorithm calculates the slope values of the connecting lines between adjacent seventh path nodes in the updated path in sequence. The slope value reflects the degree of inclination of the line on the terrain and is an important indicator to ensure the safe and stable operation of the line.

[0094] Fourth preset threshold: Set a fourth preset threshold (for example, the maximum allowable slope is 5%) to verify whether the slope of the line is too large, which may affect the project implementation and line performance.

[0095] Height adjustment: If it is detected that a certain target slope value is greater than the fourth preset threshold, the algorithm will finely adjust the heights of the two seventh path nodes involved in this slope value to adjust the slope within the allowable range. The node height adjustment needs to combine with DEM data to ensure that the adjusted nodes are still located in a higher legal construction area.

[0096] Generation of eighth path nodes: After adjustment, the generated nodes are called eighth path nodes. The path composed of this series of nodes meets the dual restrictions of the minimum tower spacing and the maximum allowable slope, further optimizing the engineering feasibility of the line.

[0097] 3. Final Path Determination: Overall inspection and confirmation: The algorithm will finally conduct a comprehensive inspection of the entire path to ensure that the distances and slopes between all adjacent eighth path nodes meet the preset engineering specifications, including but not limited to tower spacing and line slope.

[0098] Path determination: Once it is confirmed that all conditions meet the requirements, the algorithm will determine the final overhead transmission line path based on the position information of the eighth path nodes. This path is not only reasonable in planar layout but also carefully adjusted in vertical elevation and terrain slope, ensuring the safety and economy of the line in actual construction.

[0099] Through the above steps of adding nodes and finely adjusting heights based on horizontal distance restrictions and slope control, this embodiment provides a method for refining path planning and enhancing engineering adaptability. This method not only ensures the construction feasibility of the line in complex terrain but also effectively controls the erection cost and operation risk of the line through node adjustment and optimization, contributing to the technological progress in the field of automated path planning and engineering design.

[0100] It should be noted that during the implementation process of the above line planning method, the above line planning rules need to be followed. The line planning rules specifically include but are not limited to: 1. The path can cross linear features such as roads and rivers, but the path nodes (tower bases) shall not be set in the road or river areas; 2. The path is strictly prohibited from passing through the building area, and the path nodes shall not be located on the building either; 3. When the path crosses a road, the crossing angle should be as close to 90° (vertical crossing) as possible to reduce the path length along the road direction and reduce the potential safety hazards of line operation.

[0101] In an optional embodiment, the present application combines Figure 2 to introduce the implementation process of the above-mentioned intelligent route selection dynamic programming method for transmission lines considering in segments, as Figure 2 shown, which specifically includes the following steps: 2.1. Plan on the ground feature factor layer; 2.2. Call the large language model to generate potential point routes: According to the starting point, ending point and surrounding environment information, call the api of the large language model (DeepSeek, LLM, etc.), and use few-shot prompt engineering for interaction. Output a list of several possible potential path points that each planning task may pass through. The environmental information can be input as text descriptions (such as terrain and landform, obstacle distribution) or structured data as model prompts, and the LLM gives recommended points (i.e., the above-mentioned potential path point sequence) that conform to experience such as being close to the ridge line and far from the building complex.

[0102] 2.3. Conduct sampling-based planning: Construct an exploration graph structure G for path growth (i.e., the above-mentioned exploration graph), and add the starting point to the graph G as the root node. When the ending point is not found and the number of iterations does not exceed the maximum number of attempts N_max, continuously expand the exploration graph G. The expansion process alternately adopts two modes: guided and random: 2.3.1. Guided expansion: Select the candidate point t with the highest priority from the potential point list as the target. Find the existing node p_near in the exploration graph G that is closest to t, and extend a certain step length from p_near in the direction pointing to t to generate a new node p_new.

[0103] Collision detection: Call the collisionFree function to check whether the connection line from p_near to p_new passes through obstacles. If the connection line has no collision in the current ground feature layer (i.e., there is no building blockage and no improper crossing along ground facilities), it is considered that the expansion is successful, add p_new to the graph G, and record the connection relationship between p_near and p_new; otherwise, if a collision is detected (for example, p_new falls inside a building or the connection line passes through a building), this guided expansion is abandoned.

[0104] 2.3.2 Random Expansion: When the guided expansion is blocked or random exploration needs to be inserted according to a predetermined strategy, a point p_rand is randomly sampled from the passable free space. Similarly, find the nearest node p_near in graph G and try to connect to get p_new. After ensuring no obstacles through collision detection, add this p_new to graph G.

[0105] 2.3.3 Dynamic Adjustment: Whenever a new node is successfully added, the priority order of the remaining candidate points in the potential point list is updated in real time. For example, according to the distance between the candidate point and the new node, or whether the candidate point has been included in the current path, the priority of more promising points can be enhanced to improve the success rate of subsequent guided expansion.

[0106] After each new node is added, detect the distance from this node p_new to the end point. If the distance is less than the preset threshold ϵ (indicating that it is very close to the end point), it is considered that the path has been basically connected, and the loop expansion is terminated.

[0107] 2.4 Avoid and Do Not Cross Buildings, Try to Cross Roads Vertically: Follow the following special constraint rules during the path planning process: 1) The path can cross linear features such as highways and rivers, but the path nodes (tower bases) shall not be set in the road or river area.

[0108] 2) The path is strictly prohibited from passing through the building area, and the path nodes cannot be located on the building either.

[0109] 3) When the path crosses a highway, the crossing angle should be as close to 90° (vertical crossing) as possible to reduce the path length along the highway direction and reduce the safety hazards of line operation.

[0110] 2.5 Simplify the Road and Retain the Inflection Points: After the loop ends (successfully reaching the end point or reaching the maximum number of iterations), extract a path sequence from the exploration graph G that leads from the start point to the end point. Then simplify this path: remove the redundant intermediate nodes in the path and only retain the key inflection points, so that there are no obstacle crossings for the straight-line paths between any adjacent inflection points. Simplification can be achieved by iteratively checking whether the straight lines of each section of the path collide with obstacles, and removing unnecessary bending points, thereby obtaining an initial line with a streamlined tower layout.

[0111] 2.6 Planning on the DEM Data Layer: After completing the planar path planning on the feature layer and obtaining the preliminary path, enter the second stage of elevation optimization. In this stage, the digital elevation model (dem_merge data) is used to vertically adjust the positions of the inflection points in the path and supplement path points when necessary to meet the terrain elevation difference requirements.

[0112] 2.7. Fine-tune the inflection points to higher altitudes: Obtain the altitude of each inflection point in the initial path (provided by the DEM). For each inflection point, search for local terrain high points in its vicinity and slightly move the inflection point to a location with a higher altitude and within a legal construction area in the adjacent area. This treatment enables the poles and towers to be erected on hills or highlands as much as possible, improving the clearance of the line from the ground.

[0113] 2.8. Add intermediate points and fine-tune the intermediate points: 2.8.1: Insert intermediate points over long distances: Measure the horizontal distance between adjacent poles and towers. If it is found that the span exceeds a predetermined maximum threshold (i.e., the distance between two poles and towers is too long, which is not conducive to the suspension of the conductor and mechanical stability), then add an intermediate node between the two points. The selection principle for the newly added intermediate points is the same as described above, and it is preferred to fall on a safe location with a higher altitude within this section, so as to reduce the span and increase the support height at the same time; 2.8.2. Slope change inspection: For adjacent pole and tower node pairs on the path, calculate the slope or vertical angle change of the line connecting them. If it is found that the slope of a certain section of the line is too large (for example, the line drops suddenly from a high mountain to a low valley), then the position of the intermediate inflection point can be appropriately adjusted (slightly lowering or raising a certain point) to smooth the slope change and ensure that the line gradient meets the requirements of the engineering specifications.

[0114] After the above adjustments, the final path is obtained, and then the final path is verified to ensure that the intermediate terrain height of each section of the line is not higher than the height at the endpoints of the section, that is, the ground elevation at the center of any span does not exceed the altitude of the foundations of the two end poles and towers. In this way, the entire line is optimized in three-dimensional space, avoiding obstacles and conforming to the terrain undulations, forming an optimized path that meets the safety distance and erection requirements.

[0115] Through the description of the above implementation manners, those skilled in the art can clearly understand that the method according to the above embodiments can be implemented by means of software plus a necessary general hardware platform. Of course, it can also be implemented by hardware, but in many cases, the former is a better implementation manner. Based on such an understanding, the technical solution of this application, in essence, or the part that makes a contribution to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), including several instructions for causing a terminal device (which can be a mobile phone, computer, server, or network device, etc.) to execute the methods of the various embodiments of this application.

[0116] The embodiments of this application also provide a multi-stage line planning device based on large model recommendation and terrain optimization, as Figure 3 shown, Figure 3 is a structural block diagram of an optional multi-stage line planning device based on large model recommendation and terrain optimization according to the embodiments of this application. The device includes: A generation module 32 is configured to generate a sequence of potential path points based on starting point information, ending point information, and environmental data through a large language model, wherein the starting point information and the ending point information are used to construct an overhead transmission line, and the environmental data is used to indicate environmental information of a target area where the overhead transmission line is located; A search module 34 is configured to perform two-dimensional path search on the sequence of potential path points through a target algorithm on a ground feature layer to generate an initial path of the overhead transmission line; An optimization module 36 is configured to perform three-dimensional elevation optimization on path nodes in the initial path according to digital elevation data to obtain a final path of the overhead transmission line.

[0117] Through the above device, first, the starting point information, the ending point information, and the environmental data are processed by a large language model to generate a sequence of potential path points, wherein the starting point information and the ending point information are respectively used to indicate the starting point and the ending point of the overhead transmission line to be constructed, and the environmental data is used to indicate environmental information in the target area where the overhead transmission line is located; then, a target algorithm is used to recommend potential path points and quickly grow the path to obtain an initial path; finally, elevation fine-tuning and node supplementation are performed on the generated initial path, so as to complete the fine optimization of the line and obtain a final path; by adopting the above solution, the environmental understanding ability of the large language model and the path search algorithm are combined, and the digital elevation model is introduced in stages for optimization, so as to realize the full-automatic route selection planning of the overhead line; through this technical solution, the route selection efficiency can be significantly improved, the dependence on manual experience can be reduced, and a safer, more reliable, and more economical and reasonable line plan can be generated; furthermore, the problem in the related technology that due to numerous influencing factors and mutual constraints in the route selection planning process of the overhead transmission line, manual route selection is time-consuming and laborious, the planning efficiency is low, and the quality of the line plan is poor is solved.

[0118] In an exemplary embodiment, the above search module 34 is further configured to initialize and generate an exploration graph, wherein the starting point information is the root node of the exploration graph; determine a plurality of candidate points in the sequence of potential path points whose distance from the root node is less than a first preset threshold as an initial priority sequence, wherein the plurality of candidate points are used to indicate a plurality of path directions of the initial path; perform cyclic expansion on the exploration graph through a potential point priority mechanism and / or a random node sampling method until a node in the exploration graph is connected to a path node corresponding to the ending point information or the number of times of the cyclic expansion reaches a preset iteration number; extract a path sequence from the exploration graph according to the root node and the path node corresponding to the ending point information, and generate the initial path according to the path sequence.

[0119] In an exemplary embodiment, the above search module 34 is further configured to, when it is determined that there is a first path node in the potential path point sequence that has not been added to the exploration graph, determine a second path node with the highest priority among the first path nodes; determine whether the sub-path between a third path node newly added to the exploration graph and the second path node conforms to the line planning rule, where the line planning rule is used to indicate that the initial path does not overlap with the area where the obstacle is located, and the sub-path is a straight line path; when it is determined that the line planning rule is met, add the second path node to the exploration graph; when it is determined that there is no first path node or it does not conform to the line planning rule, randomly select a fourth path node in a first sub-region from the third path node to the path node corresponding to the end point information by the random node sampling method, and when it is determined that the sub-path between the third path node and the fourth path node conforms to the line planning rule, add the fourth path node to the exploration graph, where the target area includes the first sub-region.

[0120] Further, the above search module 34 is further configured to increase the sampling frequency of the random node sampling method when the number of failures in expanding the exploration graph by the potential point priority mechanism reaches a second preset threshold.

[0121] Optionally, the above optimization module 36 is further configured to identify non-critical nodes among multiple path nodes in the initial path to obtain multiple non-critical nodes, where the sub-path where the non-critical nodes are located does not overlap with the area where the obstacle is located, and the sub-path is a straight line path; delete the multiple non-critical nodes from the initial path to obtain a simplified path; perform three-dimensional elevation optimization on the path nodes in the simplified path according to the digital elevation data to obtain the final path.

[0122] In an exemplary embodiment, the above optimization module 36 is further configured to determine the altitude of multiple fifth path nodes in the simplified path according to the digital elevation data; for a sixth path node among the multiple fifth path nodes, search for a local terrain high point in a second sub-region where the sixth path node is located, and update the sixth path node to the local terrain high point to obtain multiple updated fifth path nodes, where the local terrain high point is a legal construction area; determine the final path according to the multiple updated fifth path nodes.

[0123] Optionally, the above optimization module 36 is further configured to sequentially determine the horizontal distance between two adjacent fifth path nodes among the updated multiple fifth path nodes, and determine whether there is a target horizontal distance greater than a third preset threshold among the multiple horizontal distances; in the case of determining that there is a target horizontal distance greater than the third preset threshold, add one or more intermediate path nodes between the two fifth path nodes corresponding to the target horizontal distance to obtain multiple seventh path nodes, so that the horizontal distance between any two adjacent seventh path nodes is less than or equal to the third preset threshold; sequentially determine the slope values of the pole tower connection lines corresponding to two adjacent seventh path nodes among the multiple seventh path nodes, and determine whether there is a target slope value greater than a fourth preset threshold among the multiple slope values; in the case of determining that there is a target slope value greater than the fourth preset threshold, adjust the heights of the two seventh path nodes corresponding to the target slope value to obtain multiple eighth path nodes, so that the slope values corresponding to any two adjacent eighth path nodes are less than or equal to the fourth preset threshold; determine the final path according to the multiple eighth path nodes.

[0124] An embodiment of the present application further provides a storage medium, which includes a stored program, wherein the above program runs to execute the method of any one of the above.

[0125] Optionally, in this embodiment, the above storage medium may be set to store program codes for executing the following steps: S1. Through a large language model, generate a sequence of potential path points according to the starting point information, the ending point information, and the environmental data, wherein the starting point information and the ending point information are used to build an overhead transmission line, and the environmental data is used to indicate the environmental information of the target area where the overhead transmission line is located; S2. Perform two-dimensional path search on the sequence of potential path points through a target algorithm on the ground object layer to generate an initial path of the overhead transmission line; S3. Perform three-dimensional elevation optimization on the path nodes in the initial path according to the digital elevation data to obtain the final path of the overhead transmission line.

[0126] An embodiment of the present application further provides an electronic device, including a memory and a processor, wherein a computer program is stored in the memory, and the processor is configured to run the computer program to execute the steps in any one of the above method embodiments.

[0127] Optionally, the above electronic device may further include a transmission device and an input / output device, wherein the transmission device is connected to the above processor, and the input / output device is connected to the above processor.

[0128] Optionally, in this embodiment, the above processor may be configured to execute the following steps through a computer program: S1. Generate a sequence of potential path points through a large language model based on the starting point information, ending point information, and environmental data, where the starting point information and the ending point information are used to construct an overhead transmission line, and the environmental data is used to indicate the environmental information of the target area where the overhead transmission line is located; S2. Perform a two-dimensional path search on the sequence of potential path points on the ground feature layer through a target algorithm to generate an initial path of the overhead transmission line; S3. Optimize the three-dimensional elevation of the path nodes in the initial path according to the digital elevation data to obtain the final path of the overhead transmission line.

[0129] Optionally, in this embodiment, the above storage medium may include, but is not limited to: various media such as USB flash drives, read-only memories (ROM), random access memories (RAM), external hard drives, magnetic disks, or optical discs that can store program codes.

[0130] An embodiment of the present application also provides a computer program product, including a non-volatile computer-readable storage medium, where the non-volatile computer-readable storage medium stores a computer program product, and when the computer program is executed by a processor, it implements the steps of the methods in various embodiments of the present application.

[0131] Optionally, in this embodiment, the above computer program may be configured to implement the following steps when executed by a processor: S1. Generate a sequence of potential path points through a large language model based on the starting point information, ending point information, and environmental data, where the starting point information and the ending point information are used to construct an overhead transmission line, and the environmental data is used to indicate the environmental information of the target area where the overhead transmission line is located; S2. Perform a two-dimensional path search on the sequence of potential path points on the ground feature layer through a target algorithm to generate an initial path of the overhead transmission line; S3. Optimize the three-dimensional elevation of the path nodes in the initial path according to the digital elevation data to obtain the final path of the overhead transmission line.

[0132] Optionally, the specific examples in this embodiment may refer to the examples described in the above embodiments and optional implementation manners, and will not be elaborated herein.

[0133] Obviously, those skilled in the art should understand that the various modules or steps of the present application described above can be implemented by a general-purpose computing device. They can be concentrated on a single computing device or distributed over a network composed of multiple computing devices. Optionally, they can be implemented by program codes executable by the computing device. Thus, they can be stored in a storage device and executed by the computing device. And in some cases, the steps shown or described can be executed in a sequence different from that here, or they can be separately fabricated into individual integrated circuit modules, or multiple modules or steps among them can be fabricated into a single integrated circuit module for implementation. In this way, the present application is not limited to any specific combination of hardware and software.

[0134] The above are only the preferred embodiments of the present application and are not intended to limit the present application. For those skilled in the art, the present application can have various changes and modifications. Any modification, equivalent replacement, improvement, etc. made within the principle of the present application shall be included within the protection scope of the present application.

Claims

1. An intelligent line selection dynamic programming method for transmission lines based on segmented consideration, characterized in that, Including: Generating a sequence of potential path points by a large language model based on starting point information, ending point information, and environmental data, where the starting point information and the ending point information are used to construct an overhead transmission line, and the environmental data is used to indicate environmental information of a target area where the overhead transmission line is located; Performing a two-dimensional path search on the sequence of potential path points on a ground feature layer by a target algorithm to generate an initial path of the overhead transmission line; Performing three-dimensional elevation optimization on path nodes in the initial path according to digital elevation data to obtain a final path of the overhead transmission line.

2. The intelligent route selection dynamic programming method for transmission lines based on segmented consideration according to claim 1, characterized in that Performing a two-dimensional path search on the sequence of potential path points on a ground feature layer by a target algorithm to generate an initial path of the overhead transmission line, including: Initializing and generating an exploration graph, where the starting point information is the root node of the exploration graph; Determining a plurality of candidate points with a distance less than a first preset threshold from the root node in the sequence of potential path points as an initial priority sequence, where the plurality of candidate points are used to indicate a plurality of path directions of the initial path; Circularly expanding the exploration graph by a potential point priority mechanism and / or a random node sampling method until a node in the exploration graph is connected to a path node corresponding to the ending point information or the number of times of circular expansion reaches a preset iteration number; Extracting a path sequence from the exploration graph according to the root node and the path node corresponding to the ending point information, and generating the initial path according to the path sequence.

3. The intelligent line selection dynamic programming method for transmission lines based on piecewise consideration according to claim 2, wherein Circularly expanding the exploration graph by a potential point priority mechanism and / or a random node sampling method, including: When it is determined that there is a first path node in the sequence of potential path points that has not been added to the exploration graph, determining a second path node with the highest priority among the first path nodes; Determining whether a sub-path between a newly added third path node and the second path node in the exploration graph conforms to a line planning rule, where the line planning rule is used to indicate that the initial path does not overlap with an area where an obstacle is located, and the sub-path is a straight-line path; When it is determined that the line planning rule is met, adding the second path node to the exploration graph; When it is determined that there is no first path node or it is determined that the line planning rule is not met, randomly selecting a fourth path node in a first sub-region from the third path node to the path node corresponding to the ending point information by the random node sampling method, and when it is determined that the sub-path between the third path node and the fourth path node conforms to the line planning rule, adding the fourth path node to the exploration graph, where the target area includes the first sub-region.

4. The intelligent route selection dynamic programming method for transmission lines based on segmented consideration according to claim 3, characterized in that The method further includes: When the number of failure times of expanding the exploration graph by the potential point priority mechanism reaches a second preset threshold, increasing the sampling frequency of the random node sampling method.

5. The intelligent route selection dynamic programming method for transmission lines based on segmented consideration according to claim 1, characterized in that Before performing three-dimensional elevation optimization on path nodes in the initial path according to digital elevation data to obtain a final path of the overhead transmission line, the method further includes: Identify non-critical nodes among multiple path nodes in the initial path to obtain multiple non-critical nodes, where the sub-path where the non-critical node is located does not overlap with the area where the obstacle is located, and the sub-path is a straight path; Delete the multiple non-critical nodes from the initial path to obtain a simplified path; Perform three-dimensional elevation optimization on the path nodes in the simplified path according to the digital elevation data to obtain the final path.

6. The intelligent line selection dynamic programming method for transmission lines based on segmented consideration according to claim 5, characterized in that, Performing three-dimensional elevation optimization on the path nodes in the simplified path according to the digital elevation data to obtain the final path, including: Determine the elevation of multiple fifth path nodes in the simplified path according to the digital elevation data; For the sixth path node among the multiple fifth path nodes, search for local terrain high points in the second sub-region where the sixth path node is located, and update the sixth path node to the local terrain high point to obtain multiple updated fifth path nodes, where the local terrain high point is a legal construction area; Determine the final path according to the multiple updated fifth path nodes.

7. The intelligent line selection dynamic programming method for transmission lines based on segmented consideration according to claim 6, characterized in that Determining the final path according to the multiple updated fifth path nodes, including: Successively determine the horizontal distances between adjacent two of the multiple updated fifth path nodes, and determine whether there is a target horizontal distance greater than a third preset threshold among the multiple horizontal distances; When it is determined that there is a target horizontal distance greater than the third preset threshold, add one or more intermediate path nodes between the two fifth path nodes corresponding to the target horizontal distance to obtain multiple seventh path nodes, so that the horizontal distance between any two adjacent seventh path nodes is less than or equal to the third preset threshold; Successively determine the slope values of the tower pole connection lines corresponding to adjacent two of the multiple seventh path nodes, and determine whether there is a target slope value greater than a fourth preset threshold among the multiple slope values; When it is determined that there is a target slope value greater than the fourth preset threshold, perform height adjustment on the two seventh path nodes corresponding to the target slope value to obtain multiple eighth path nodes, so that the slope values corresponding to any two adjacent eighth path nodes are less than or equal to the fourth preset threshold; Determine the final path according to the multiple eighth path nodes.

8. A multi-stage route planning device based on large model recommendation and terrain optimization, characterized in that, Including: A generation module for generating a sequence of potential path points through a large language model according to the starting point information, the ending point information, and the environmental data, where the starting point information and the ending point information are used to build an overhead transmission line, and the environmental data is used to indicate the environmental information of the target area where the overhead transmission line is located; A search module for performing two-dimensional path search on the sequence of potential path points on the ground feature layer through a target algorithm to generate the initial path of the overhead transmission line; An optimization module for performing three-dimensional elevation optimization on the path nodes in the initial path according to the digital elevation data to obtain the final path of the overhead transmission line.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored program, where the program executes the method described in any one of claims 1 to 7 when running.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the method according to any one of claims 1 to 7.

Images