Transmission line intelligent route selection method based on adaptive resolution grid and improved ant colony algorithm

Abstract

The present application relates to a kind of transmission line intelligent route selection method based on adaptive resolution grid and improved ant colony algorithm, comprising: after remote sensing image rasterization, adaptive quadtree image segmentation is carried out, and complete adaptive resolution grid map is obtained;Establish hierarchical analysis model, obtain the construction cost of each grid area in transmission line, i.e.grid cost, establish adaptive resolution grid map cost model;Through improved ant colony algorithm, the transmission line path after inflection point correction is obtained.The present application uses improved ant colony algorithm to speed up the iterative convergence of algorithm, and is combined with adaptive resolution grid, the reduction of grid number can significantly improve the efficiency of algorithm;The size of adaptive resolution grid changes according to the change of terrain on remote sensing map, can avoid the problem that the position accuracy of planned route is low;Inflection point correction is carried out to the planned route, reduces the number of corner tower that needs to be built, saves construction cost.

Description

Technical Field

[0001] This invention relates to the field of intelligent transmission line selection technology, and in particular to an intelligent transmission line selection method based on adaptive resolution grid and improved ant colony algorithm. Background Technology

[0002] As an important component of power transmission and transformation projects, the route selection of power transmission lines is related to the planning and construction of the entire power grid and has a profound impact on the local economic development and residents' lives.

[0003] Traditional route selection methods are labor-intensive and time-consuming, with their main drawback being the significant influence of human factors. Faced with increasingly complex power grids, digital design is being widely applied across various fields, making traditional manual transmission line planning insufficient to meet the higher demands of power grid construction. Furthermore, traditional manual methods are not only time-consuming and labor-intensive but also struggle to ensure the timeliness of geographical information. Due to the long duration of manual work, changes in the geographical environment can lead to planned routes that no longer meet requirements. Utilizing computers and intelligent algorithms for transmission line route selection can significantly shorten the construction period. By combining multiple transmission line evaluation indicators, the required construction cost can be quickly calculated. Using scientific and rational layout planning is of great significance to power transmission and transformation projects.

[0004] Currently, existing methods for intelligent route selection include dynamic programming, Dijkstra's algorithm, genetic algorithms, and A* algorithms. These methods are suitable for practical problems with relatively simple objectives, uncomplicated geographical environments, and no complex cost calculations. Ant colony optimization, due to its intelligence and strong robustness, is suitable for route selection problems involving factors such as route length, number of corners, and construction costs in different environments, and is now widely used in transmission line planning and road planning. Furthermore, when the planning area for transmission lines is large, rasterizing the map results in a large number of grid cells, leading to slow planning speed and low algorithm efficiency, thus making intelligent route selection inefficient. While increasing the grid size and reducing the number of grid cells speeds up the algorithm, it reduces the accuracy of the planned route location. Summary of the Invention

[0005] To address the issues of slow operation and low route selection efficiency of ant colony algorithms, the present invention aims to provide an intelligent route selection method for power transmission lines based on adaptive resolution grids and an improved ant colony algorithm, which helps assist manual selection of routes that meet the required cost and accelerates route selection efficiency.

[0006] To achieve the above objectives, the present invention adopts the following technical solution: a method for intelligent transmission line selection based on adaptive resolution grid and improved ant colony algorithm, the method comprising the following sequential steps:

[0007] (1) Based on the remote sensing image, the remote sensing image is rasterized and then subjected to adaptive quadtree image segmentation to obtain a complete adaptive resolution raster map.

[0008] (2) Based on the four factors affecting the cost of transmission line selection, a hierarchical analysis model was established to obtain the transmission line construction cost, i.e., grid cost, in each grid area. An adaptive resolution grid map cost model was established based on the adaptive resolution grid map and grid cost.

[0009] (3) Based on the adaptive resolution raster map cost model, the transmission line path after inflection point correction is obtained by the improved ant colony algorithm.

[0010] Step (1) specifically includes the following steps:

[0011] (1a) Perform ground feature identification on the remote sensing image to obtain a ground feature identification map, and then read and convert it into a grayscale image;

[0012] (1b) Rasterize the grayscale image to obtain a grayscale image composed of regularly arranged initial grids, and perform adaptive quadtree image segmentation on each initial grid in turn.

[0013] (1c) Arrange each segmented initial grid cell according to the rasterization order in step (1b) to obtain a complete adaptive resolution raster map.

[0014] Step (2) specifically includes the following steps:

[0015] (2a) The four factors affecting the cost of transmission line selection are environmental factors, social factors, construction factors and operation and maintenance factors. The environmental factors, social factors, construction factors and operation and maintenance factors are further subdivided into multiple evaluation indicators. Among them, the environmental factors include forest land, water area and desertified surface, social factors include industrial land, residential area and urban planning impact, construction factors include existing line crossings, geological and meteorological conditions, and operation and maintenance factors include operation safety and maintenance costs.

[0016] (2b) The grid cost is used as the target layer, environmental factors, social factors, construction factors and operation and maintenance factors are used as the factor layer, and the evaluation indicators are used as the evaluation indicator layer to construct a hierarchical model.

[0017] (2c) The hierarchical model is constructed into a hierarchical analysis model using the analytic hierarchy process (AHP). The evaluation indicators under each factor layer are quantitatively evaluated according to the scale of 0.1 to 0.9 in the AHP to obtain the evaluation data. Then, the priority relation matrix E is constructed on the evaluation data based on the fuzzy hierarchical analysis method (FAHP). The weight of the evaluation indicators under each factor layer is solved through the priority relation matrix E. The comprehensive weight of each evaluation indicator on the target line selection cost is solved. The grid cost is obtained based on the comprehensive weight.

[0018] (2d) Based on the raster cost, the adaptive resolution raster map cost model is obtained by combining the adaptive resolution raster map.

[0019] Step (3) specifically includes the following steps:

[0020] (3a) In the traditional ant colony algorithm, each grid cell on the map is used as a basic unit for path search. The traditional ant colony algorithm is improved by reading the map information in the adaptive resolution grid map cost model and using each adaptive resolution grid cell on the map as a basic unit for path search. The grid adopts a neighborhood structure with multiple centers. The rule for determining the number of centers is: the size is the minimum grid size L. m Place a center point within a grid that is 1 or 2 times the size of the minimum grid size L. m Four center points are placed in a grid that is 4 or 8 times larger than the grid size, and 16 center points are placed in a grid with the largest size. Each center point represents the grid it belongs to. The route searched by the ants is formed by connecting the center points.

[0021] (3b) Initialize the pheromone matrix τ = τ0. The coordinates of each value in the pheromone matrix τ are the corresponding positions on the grid map. Initialize the iteration number K, the number of ants M, the pheromone influence factor α, and the pheromone volatility factor ρ. Introduce the neighbor grid cost influence factor β and the direction guidance parameter influence factor γ as improvement strategies and initialize them.

[0022] (3c) Calculate p based on the probability formula for grid selection. abi The center point of the next grid is selected by a roulette wheel method, and points that have already been passed are not selected;

[0023] (3d) After ant m completes this path search, record the route cost and route length, m≤M;

[0024] (3e) After M ants have completed the search, update the pheromone matrix τ and the influence factor γ of the directional guidance parameter to complete the k-th iteration, k≤K;

[0025] (3f) Record the route with the minimum cost value after each iteration, and determine whether k is equal to K. If yes, proceed to step (3g); otherwise, jump to step (3c).

[0026] (3g) After all iterations are completed, output the route with the lowest cost value among all iteration results and perform inflection point correction to remove redundant inflection points, that is, output the transmission line path after inflection point correction.

[0027] Step (1b) specifically includes the following steps:

[0028] (1b1) Set the minimum grid size in the entire adaptive resolution raster map to L. m Set all initial grids to the first generation grid;

[0029] (1b2) If the current grid size is greater than L m If the grid is not divided, proceed to step (1b3); otherwise, stop dividing the grid and proceed to step (1b5).

[0030] (1b3) If the variance of the grayscale value of the current grid is greater than the variance threshold V f If the result is positive, segmentation is required, and the process jumps to step (1b4). Otherwise, it is determined whether the variance of the gray values ​​of the 4×4 sub-grids within the grid is less than the mean threshold V. s If the variance of the gray values ​​of all 4×4 sub-grids within a raster is less than the mean threshold V s If the current grid is not divided, proceed to step (1b5); otherwise, proceed to step (1b4).

[0031] (1b4) Divide the current grid into 4 second-generation sub-grids of the same size using a quadtree. These 4 sub-grids are of the same generation. The current grid is the previous generation grid of the 4 sub-grids. Jump to step (1b2) in turn to determine whether to divide the 4 second-generation sub-grids.

[0032] (1b5) If there are other grids in the same generation of the current grid that have not been segmented, then jump to step (1b2) to perform the segmentation judgment in turn; otherwise, perform the judgment on other grids in the previous generation and jump to step (1b2) until every grid in the entire image does not meet the segmentation judgment condition.

[0033] Step (2c) specifically includes the following steps:

[0034] (2c1) Construct a priority relationship matrix E1 based on the relative importance of forest land, water area, and desertified land surface in the evaluation index layer to environmental factors in the factor layer:

[0035] E1=(e ij ) n×n

[0036] Where i, j = 1, 2…n, n is the number of evaluation indicators corresponding to environmental factors, e ij The value is expressed on a scale from 0.1 to 0.9, and e ij +e ji =1. Similarly, we obtain the priority relationship matrix E2 of industrial land, residential areas, and urban planning impacts on social factors; the priority relationship matrix E3 of existing line crossings, geological and meteorological conditions on construction factors; the priority relationship matrix E4 of operation safety and maintenance costs on operation and maintenance factors; and the priority relationship matrix E5 of environmental factors, social factors, construction factors, and operation and maintenance factors on raster costs. c ;

[0037] Summing each row of matrix E1 (2c2) yields:

[0038]

[0039] Where i = 1, 2, ..., n, find... Obtain the fuzzy consistency matrix

[0040]

[0041] Similarly, we can obtain

[0042] (2c3) Calculate the weight q of the i-th element of forest land, water area, and desertified surface to the environmental factor at the next higher level. i :

[0043]

[0044] Where i = 1, 2…n, and μ ≤ (n-1) / 2, the weights ω1 = (q1, q2, q3) for the three evaluation indicators of forest land, water area, and desertified surface under environmental factors are derived. Similarly, the weights ω2 for the three evaluation indicators of industrial land, residential area, and urban planning impact under social factors are derived. The weights ω3 for the three evaluation indicators of existing line crossings, geological conditions, and meteorological conditions under construction factors are derived. The weights ω4 for the two evaluation indicators of operation safety and maintenance costs under operation and maintenance factors are derived. Finally, the weights ω for the four factors of environmental factors, social factors, construction factors, and operation and maintenance factors under grid cost are derived. c ;

[0045] (2c4) Calculate the overall weight of all evaluation indicators in the evaluation indicator layer on the raster cost as ω. c ′:

[0046] ω c ′=ω c ·[ω1,ω2,ω3,ω4] T

[0047] The rating levels for each evaluation indicator in the evaluation indicator layer are [s1, s2, Ks]. 11 ] T The grid cost c is represented by five levels, from 1 to 5. The formula for calculating the grid cost c is:

[0048] c = ω c ′·[s1,s2,Ks 11 ] T .

[0049] In step (3c), the grid selection probability is calculated according to the grid selection probability formula. Specifically, the following steps are included:

[0050] (3c1) The ant is currently located in grid cell a, and the next grid cell to be selected is b. i The set of candidate grid cells in the neighboring grid is g represents the number of grid cells to be selected, which is determined by the current position of the grid cell. For the ant to move from a to b in the k-th iteration... i pheromone concentration;

[0051] (3c2) From a to b i The reciprocal of the unit cost of the route, denoted as for:

[0052]

[0053] in, For ab i The unit cost of a segment path is obtained by the following formula:

[0054]

[0055] Where S is ab i The number of grid cells traversed by the segment path, l j For ab i The length of the segment path within the j-th grid cell, c j For ab i The raster cost of the j-th raster traversed by the segment path. For ab i The length of the path segment;

[0056] (3c3) Introduce the directional guidance parameter o, which is defined as follows:

[0057]

[0058] Where θ represents the distance from the current grid position a to the grid cell b to be selected. i The angle between the line connecting a and the line connecting a to the endpoint;

[0059] (3c4) The improved ant colony algorithm's grid selection probability is obtained from steps (3c2) and (3c3).

[0060]

[0061] in, That is, the ant moves from a to b. i The probability of.

[0062] Step (3e) specifically includes the following steps:

[0063] (3e1) The pheromone concentration is updated with each iteration, and the pheromone concentration of the grid is updated according to the following formula:

[0064]

[0065] Where, Δτ ab (k,m) represents the pheromone left by the m-th ant when it completes its search and moves from grid a to grid b in the k-th iteration. The calculation formula is:

[0066]

[0067] Where c(m) is the cost of the route obtained by ant m, and Q is the correction coefficient;

[0068] (3e2) After each iteration, γ is updated as follows:

[0069] γ(k+1)=(1-λ)·γ(k)

[0070] Where λ is the directional volatility coefficient, γ(k+1) is the control parameter of o in the (k+1)th iteration, and γ(k) is the control parameter of the kth iteration.

[0071] The step (3g) specifically includes the following steps:

[0072] (3g1) In the inflection point correction process, the processing proceeds from the starting point to the ending point. A route is formed by connecting f points: route = {v1, v2, ..., v} f The total cost of the route includes the route cost and the inflection point cost q.

[0073] (3g2) First, process the line connecting the three points v1, v2, v3. If the total cost of the route connecting {v1, v2, v3} is higher than the total cost of the route connecting {v1, v3}, then remove point v2 from the route and process the line connecting the three points v1, v3, v4. Otherwise, do not process the points in the route and process v2, v3, v4.

[0074] (3g3) Repeat step (3g2) until the last point v is processed. f This yields the transmission line path after inflection point correction.

[0075] As can be seen from the above technical solution, the beneficial effects of the present invention are as follows: First, the improved ant colony algorithm accelerates the iterative convergence of the algorithm, and combined with the adaptive resolution grid, the reduction of the number of grids can significantly improve the efficiency of the algorithm; Second, the size of the adaptive resolution grid changes according to the terrain changes on the remote sensing map. The grid size is small and dense in areas with drastic terrain changes or many types of terrain, which can avoid the problem of low accuracy of the planned route position; Third, the inflection point correction of the planned route reduces the number of corner towers that need to be built, saving construction costs. Attached Figure Description

[0076] Figure 1 This is a flowchart of the method of the present invention;

[0077] Figure 2 This is a flowchart of the adaptive quadtree image segmentation process of the present invention;

[0078] Figure 3 This is a schematic diagram of the multi-center point neighborhood structure of the present invention;

[0079] Figure 4 This is a schematic diagram of the inflection point correction of the present invention. Detailed Implementation

[0080] like Figure 1 As shown, a smart transmission line selection method based on adaptive resolution grid and improved ant colony algorithm is presented. The method includes the following steps in sequence:

[0081] (1) Based on the remote sensing image, the remote sensing image is rasterized and then subjected to adaptive quadtree image segmentation to obtain a complete adaptive resolution raster map.

[0082] (2) Based on the four factors affecting the cost of transmission line selection, a hierarchical analysis model was established to obtain the transmission line construction cost, i.e., grid cost, in each grid area. An adaptive resolution grid map cost model was established based on the adaptive resolution grid map and grid cost.

[0083] (3) Based on the adaptive resolution raster map cost model, the transmission line path after inflection point correction is obtained by the improved ant colony algorithm.

[0084] Step (1) specifically includes the following steps:

[0085] (1a) Perform ground feature identification on the remote sensing image to obtain a ground feature identification map, and then read and convert it into a grayscale image;

[0086] (1b) Rasterize the grayscale image to obtain a grayscale image composed of regularly arranged initial grids, and perform adaptive quadtree image segmentation on each initial grid in turn.

[0087] (1c) Arrange each segmented initial grid cell according to the rasterization order in step (1b) to obtain a complete adaptive resolution raster map.

[0088] Step (2) specifically includes the following steps:

[0089] (2a) The four factors affecting the cost of transmission line selection are environmental factors, social factors, construction factors and operation and maintenance factors. The environmental factors, social factors, construction factors and operation and maintenance factors are further subdivided into multiple evaluation indicators. Among them, the environmental factors include forest land, water area and desertified surface, social factors include industrial land, residential area and urban planning impact, construction factors include existing line crossings, geological and meteorological conditions, and operation and maintenance factors include operation safety and maintenance costs.

[0090] (2b) The grid cost is used as the target layer, environmental factors, social factors, construction factors and operation and maintenance factors are used as the factor layer, and the evaluation indicators are used as the evaluation indicator layer to construct a hierarchical model.

[0091] (2c) The hierarchical model is constructed into a hierarchical analysis model using the analytic hierarchy process (AHP). The evaluation indicators under each factor layer are quantitatively evaluated according to the scale of 0.1 to 0.9 in the AHP to obtain the evaluation data. Then, the priority relation matrix E is constructed on the evaluation data based on the fuzzy hierarchical analysis method (FAHP). The weight of the evaluation indicators under each factor layer is solved through the priority relation matrix E. The comprehensive weight of each evaluation indicator on the target line selection cost is solved. The grid cost is obtained based on the comprehensive weight.

[0092] (2d) Based on the raster cost, the adaptive resolution raster map cost model is obtained by combining the adaptive resolution raster map.

[0093] like Figure 3 As shown, step (3) specifically includes the following steps:

[0094] (3a) In the traditional ant colony algorithm, each grid cell on the map is used as a basic unit for path search. The traditional ant colony algorithm is improved by reading the map information in the adaptive resolution grid map cost model and using each adaptive resolution grid cell on the map as a basic unit for path search. The grid adopts a neighborhood structure with multiple centers. The rule for determining the number of centers is: the size is the minimum grid size L. m Place a center point within a grid that is 1 or 2 times the size of the minimum grid size L.m Four center points are placed in a grid that is 4 or 8 times larger than the grid size, and 16 center points are placed in a grid with the largest size. Each center point represents the grid it belongs to. The route searched by the ants is formed by connecting the center points.

[0095] (3b) Initialize the pheromone matrix τ = τ0. The coordinates of each value in the pheromone matrix τ are the corresponding positions on the grid map. Initialize the iteration number K, the number of ants M, the pheromone influence factor α, and the pheromone volatility factor ρ. Introduce the neighbor grid cost influence factor β and the direction guidance parameter influence factor γ as improvement strategies and initialize them.

[0096] (3c) Calculate p based on the probability formula for grid selection. abi The center point of the next grid is selected by a roulette wheel method, and points that have already been passed are not selected;

[0097] (3d) After ant m completes this path search, record the route cost and route length, m≤M;

[0098] (3e) After M ants have completed the search, update the pheromone matrix τ and the influence factor γ of the directional guidance parameter to complete the k-th iteration, k≤K;

[0099] (3f) Record the route with the minimum cost value after each iteration, and determine whether k is equal to K. If yes, proceed to step (3g); otherwise, jump to step (3c).

[0100] (3g) After all iterations are completed, output the route with the lowest cost value among all iteration results and perform inflection point correction to remove redundant inflection points, that is, output the transmission line path after inflection point correction.

[0101] like Figure 2 As shown, step (1b) specifically includes the following steps:

[0102] (1b1) Set the minimum grid size in the entire adaptive resolution raster map to L. m Set all initial grids to the first generation grid;

[0103] (1b2) If the current grid size is greater than L m If the grid is not divided, proceed to step (1b3); otherwise, stop dividing the grid and proceed to step (1b5).

[0104] (1b3) If the variance of the grayscale value of the current grid is greater than the variance threshold V f If the result is positive, segmentation is required, and the process jumps to step (1b4). Otherwise, it is determined whether the variance of the gray values ​​of the 4×4 sub-grids within the grid is less than the mean threshold V. s If the variance of the gray values ​​of all 4×4 sub-grids within a raster is less than the mean threshold Vs If the current grid is not divided, proceed to step (1b5); otherwise, proceed to step (1b4).

[0105] (1b4) Divide the current grid into 4 second-generation sub-grids of the same size using a quadtree. These 4 sub-grids are of the same generation. The current grid is the previous generation grid of the 4 sub-grids. Jump to step (1b2) in turn to determine whether to divide the 4 second-generation sub-grids.

[0106] (1b5) If there are other grids in the same generation of the current grid that have not been segmented, then jump to step (1b2) to perform the segmentation judgment in turn; otherwise, perform the judgment on other grids in the previous generation and jump to step (1b2) until every grid in the entire image does not meet the segmentation judgment condition.

[0107] Step (2c) specifically includes the following steps:

[0108] (2c1) Construct a priority relationship matrix E1 based on the relative importance of forest land, water area, and desertified land surface in the evaluation index layer to environmental factors in the factor layer:

[0109] E1=(e ij ) n×n

[0110] Where i, j = 1, 2…n, n is the number of evaluation indicators corresponding to environmental factors, e ij The value is expressed on a scale from 0.1 to 0.9, and e ij +e ji =1. Similarly, we obtain the priority relationship matrix E2 of industrial land, residential areas, and urban planning impacts on social factors; the priority relationship matrix E3 of existing line crossings, geological and meteorological conditions on construction factors; the priority relationship matrix E4 of operation safety and maintenance costs on operation and maintenance factors; and the priority relationship matrix E5 of environmental factors, social factors, construction factors, and operation and maintenance factors on raster costs. c ;

[0111] Summing each row of matrix E1 (2c2) yields:

[0112]

[0113] Where i = 1, 2, ..., n, find... Obtain the fuzzy consistency matrix

[0114]

[0115] Similarly, we can obtain

[0116] (2c3) Calculate the weight q of the i-th element of forest land, water area, and desertified surface to the environmental factor at the next higher level. i :

[0117]

[0118] Where i = 1, 2…n, and μ ≤ (n-1) / 2, the weights ω1 = (q1, q2, q3) for the three evaluation indicators of forest land, water area, and desertified surface under environmental factors are derived. Similarly, the weights ω2 for the three evaluation indicators of industrial land, residential area, and urban planning impact under social factors are derived. The weights ω3 for the three evaluation indicators of existing line crossings, geological conditions, and meteorological conditions under construction factors are derived. The weights ω4 for the two evaluation indicators of operation safety and maintenance costs under operation and maintenance factors are derived. Finally, the weights ω for the four factors of environmental factors, social factors, construction factors, and operation and maintenance factors under grid cost are derived. c ;

[0119] (2c4) Calculate the overall weight of all evaluation indicators in the evaluation indicator layer on the raster cost as ω. c ′:

[0120] ω c ′=ω c ·[ω1,ω2,ω3,ω4] T

[0121] The rating levels for each evaluation indicator in the evaluation indicator layer are [s1, s2, Ks]. 11 ] T The grid cost c is represented by five levels, from 1 to 5. The formula for calculating the grid cost c is:

[0122] c = ω c ′·[s1,s2,Ks 11 ] T .

[0123] In step (3c), the grid selection probability p is calculated according to the grid selection probability formula. abi Specifically, the following steps are included:

[0124] (3c1) The ant is currently located in grid cell a, and the next grid cell to be selected is b. i The set of candidate grid cells in the neighboring grid is g represents the number of grid cells to be selected, which is determined by the current position of the grid cell. For the ant to move from a to b in the k-th iteration... i pheromone concentration;

[0125] (3c2) From a to b i The reciprocal of the unit cost of the route, denoted as for:

[0126]

[0127] in, For ab i The unit cost of a segment path is obtained by the following formula:

[0128]

[0129] Where S is ab i The number of grid cells traversed by the segment path, l j For ab i The length of the segment path within the j-th grid cell, c j For ab i The raster cost of the j-th raster traversed by the segment path. For ab i The length of the path segment;

[0130] (3c3) Introduce the directional guidance parameter o, which is defined as follows:

[0131]

[0132] Where θ represents the distance from the current grid position a to the grid cell b to be selected. i The angle between the line connecting a and the line connecting a to the endpoint;

[0133] (3c4) The improved ant colony algorithm's grid selection probability is obtained from steps (3c2) and (3c3).

[0134]

[0135] in, That is, the ant moves from a to b. i The probability of.

[0136] Step (3e) specifically includes the following steps:

[0137] (3e1) The pheromone concentration is updated with each iteration, and the pheromone concentration of the grid is updated according to the following formula:

[0138]

[0139] Where, Δτ ab (k,m) represents the pheromone left by the m-th ant when it completes its search and moves from grid a to grid b in the k-th iteration. The calculation formula is:

[0140]

[0141] Where c(m) is the cost of the route obtained by ant m, and Q is the correction coefficient;

[0142] (3e2) After each iteration, γ is updated as follows:

[0143] γ(k+1)=(1-λ)·γ(k)

[0144] Where λ is the directional volatility coefficient, γ(k+1) is the control parameter of o in the (k+1)th iteration, and γ(k) is the control parameter of the kth iteration.

[0145] like Figure 4 As shown, step (3g) specifically includes the following steps:

[0146] (3g1) In the inflection point correction process, the processing proceeds from the starting point to the ending point. A route is formed by connecting f points: route = {v1, v2, ..., v} f The total cost of the route includes the route cost and the inflection point cost q.

[0147] (3g2) First, process the line connecting the three points v1, v2, v3. If the total cost of the route connecting {v1, v2, v3} is higher than the total cost of the route connecting {v1, v3}, then remove point v2 from the route and process the line connecting the three points v1, v3, v4. Otherwise, do not process the points in the route and process v2, v3, v4.

[0148] (3g3) Repeat step (3g2) until the last point v is processed. f This yields the transmission line path after inflection point correction.

[0149] In summary, this invention employs an improved ant colony algorithm to accelerate the iterative convergence of the algorithm, and combines it with an adaptive resolution grid. The reduction in the number of grids significantly improves the efficiency of the algorithm. The size of the adaptive resolution grid varies according to the terrain changes on the remote sensing image. The grid size is small and dense in areas with drastic terrain changes or diverse terrain types, thus avoiding the problem of low accuracy in the planned route location. The inflection point correction of the planned route reduces the number of corner towers that need to be built, saving construction costs.

Claims

1. A method for intelligent transmission line selection based on adaptive resolution grid and improved ant colony algorithm, characterized in that: The method includes the following steps in sequence: (1) Based on the remote sensing image, the remote sensing image is rasterized and then subjected to adaptive quadtree image segmentation to obtain a complete adaptive resolution raster map. (2) Based on the four factors affecting the cost of transmission line selection, a hierarchical analysis model is established to obtain the transmission line construction cost, i.e., grid cost, in each grid area. An adaptive resolution grid map cost model is established based on the adaptive resolution grid map and grid cost. (3) Based on the adaptive resolution raster map cost model, the transmission line path after inflection point correction is obtained by the improved ant colony algorithm; Step (3) specifically includes the following steps: (3a) In the traditional ant colony algorithm, each grid cell on the map is used as a basic unit for path search. The traditional ant colony algorithm is improved by reading the map information in the adaptive resolution grid map cost model and using each adaptive resolution grid cell on the map as a basic unit for path search. The grid adopts a neighborhood structure with multiple centers. The rule for determining the number of centers is: the size is the smallest grid size. Place a center point within a grid cell that is 1 or 2 times the size of the minimum grid cell. Four center points are placed in a grid that is 4 or 8 times larger than the grid size, and 16 center points are placed in a grid with the largest size. Each center point represents the grid it belongs to. The route searched by the ants is formed by connecting the center points. (3b) Initialize the pheromone matrix pheromone matrix Each value in the graph corresponds to a location on the grid map. The initialization parameters are: iteration count K, ant count M, and pheromone influence factor. pheromone volatile factors Introducing the neighbor grid cost impact factor Influence factors of directional guidance parameters As an improvement strategy, and initialized; (3c) Calculated according to the probability formula for grid selection The center point of the next grid is selected by a roulette wheel method, and points that have already been passed are not selected; (3d) After ant m completes this path search, it records the route cost and route length. ; (3e) After M ants have completed the search, the pheromone matrix is... Influence factors of directional guidance parameters Perform the update and complete the k-th iteration. ; (3f) Record the route with the minimum cost value after each iteration, and determine whether k is equal to K. If yes, proceed to step (3g); otherwise, jump to step (3c). (3g) After all iterations are completed, output the route with the minimum cost value among all iteration results and perform inflection point correction to remove redundant inflection points, that is, output the transmission line path after inflection point correction. The step (3g) specifically includes the following steps: (3g1) In the inflection point correction process, the process is carried out from the starting point to the ending point. A route is formed by connecting f points. The total cost of the route includes the route cost and the inflection point cost q; (3g2) First treatment A line connecting three points, if the line... The total cost of the route is higher than that of the connection. The total cost of the route is then calculated by removing the points from the route. Then process Connect the three points; otherwise, the points in the route are not processed, and then processing begins. ; (3g3) Repeat step (3g2) until the last point is processed. This yields the transmission line path after inflection point correction.

2. The intelligent transmission line selection method based on adaptive resolution grid and improved ant colony algorithm according to claim 1, characterized in that: Step (1) specifically includes the following steps: (1a) Perform ground feature identification on the remote sensing image to obtain a ground feature identification map, and then read and convert it into a grayscale image; (1b) Rasterize the grayscale image to obtain a grayscale image composed of regularly arranged initial grids, and perform adaptive quadtree image segmentation on each initial grid in turn. (1c) Arrange each segmented initial grid cell according to the rasterization order in step (1b) to obtain a complete adaptive resolution raster map.

3. The intelligent transmission line selection method based on adaptive resolution grid and improved ant colony algorithm according to claim 1, characterized in that: Step (2) specifically includes the following steps: (2a) The four factors affecting the cost of transmission line selection are environmental factors, social factors, construction factors and operation and maintenance factors. The environmental factors, social factors, construction factors and operation and maintenance factors are further subdivided into multiple evaluation indicators. Among them, the environmental factors include forest land, water area and desertified surface, social factors include industrial land, residential area and urban planning impact, construction factors include existing line crossings, geological and meteorological conditions, and operation and maintenance factors include operation safety and maintenance costs. (2b) The grid cost is used as the target layer, environmental factors, social factors, construction factors and operation and maintenance factors are used as the factor layer, and the evaluation indicators are used as the evaluation indicator layer to construct a hierarchical model; (2c) The hierarchical model is constructed into a hierarchical analysis model using the analytic hierarchy process (AHP). The evaluation indicators under each factor layer are quantitatively evaluated according to the scale of 0.1 to 0.9 in the AHP to obtain the evaluation data. Then, the priority relation matrix E is constructed on the evaluation data based on the fuzzy hierarchical analysis method (FAHP). The weight of the evaluation indicators under each factor layer is solved through the priority relation matrix E. The comprehensive weight of each evaluation indicator on the target line selection cost is solved. The grid cost is obtained based on the comprehensive weight. (2d) Based on the raster cost, the adaptive resolution raster map cost model is obtained by combining the adaptive resolution raster map.

4. The intelligent transmission line selection method based on adaptive resolution grid and improved ant colony algorithm according to claim 2, characterized in that: Step (1b) specifically includes the following steps: (1b1) Set the minimum grid size in the entire adaptive resolution raster map to . Set all initial grids to the first generation grid; (1b2) If the current grid size is greater than If the grid is not divided, proceed to step (1b3); otherwise, stop dividing the grid and proceed to step (1b5). (1b3) If the variance of the grayscale value of the current raster is greater than the variance threshold If the result is positive, segmentation is required, and the process jumps to step (1b4). Otherwise, it is determined whether the variance of the gray values ​​of the 4×4 sub-grids within the grid is less than the mean threshold. If the variance of the gray values ​​of all 4×4 sub-grids within a given grid is less than the mean threshold... If the current grid is not divided, the process will proceed to step (1b5); otherwise, the process will proceed to step (1b4). (1b4) Divide the current grid into 4 second-generation sub-grids of the same size using a quadtree. These 4 sub-grids are of the same generation. The current grid is the previous generation grid of the 4 sub-grids. Jump to step (1b2) in turn to determine whether to divide the 4 second-generation sub-grids. (1b5) If there are other grids in the same generation of the current grid that have not been segmented, then jump to step (1b2) to perform the segmentation judgment in turn; otherwise, perform the judgment on other grids in the previous generation and jump to step (1b2) until every grid in the entire image does not meet the segmentation judgment condition.

5. The intelligent transmission line selection method based on adaptive resolution grid and improved ant colony algorithm according to claim 3, characterized in that: Step (2c) specifically includes the following steps: (2c1) Construct a priority relationship matrix by evaluating the relative importance of forest land, water area, and desertified land surface in the indicator layer to environmental factors in the factor layer. : ； Where i, j = 1, 2…n, and n is the number of evaluation indicators corresponding to the environmental factors. The values ​​are expressed on a scale from 0.1 to 0.9, and Similarly, the priority relationship matrix of the impact of industrial land, residential areas, and urban planning on social factors can be obtained. The priority relationship matrix of existing line crossings, geological and meteorological conditions on construction factors Matrix of the priority relationship between operational safety and maintenance costs on operational factors Priority matrix of environmental factors, social factors, construction factors, and operation and maintenance factors on raster costs ; (2c2) matrix Summing each row yields: ； Where i = 1,2…n, find The fuzzy consistency matrix is ​​obtained. : ； Similarly, we can obtain , , , ; (2c3) Calculate the weight of the i-th element of forest land, water area, and desertified surface with respect to the environmental factors at the next higher level. : ； Where i = 1,2…n, and This led to the determination of the weights of the three evaluation indicators—forest land, water area, and desertified surface—under environmental factors. Similarly, the weights of the three evaluation indicators—industrial land use, residential areas, and urban planning—under the influence of social factors can be derived. The weights of the three evaluation indicators under construction factors: existing line crossings, geological conditions, and meteorological conditions. The weights of the two evaluation indicators, operation safety and maintenance cost, under operation and maintenance factors. The weights of the four factors—environmental, social, construction, and operation and maintenance—under grid cost. ; (2c4) Calculate the comprehensive weight of all evaluation indicators in the evaluation indicator layer on the raster cost. : ； The rating levels for each evaluation indicator in the evaluation indicator layer are: The grid cost c is represented by five levels, from 1 to 5. The formula for calculating the grid cost c is: 。 6. The intelligent transmission line selection method based on adaptive resolution grid and improved ant colony algorithm according to claim 1, characterized in that: In step (3c), the grid selection probability is calculated according to the grid selection probability formula. Specifically, the following steps are included: (3c1) The ant is currently located at grid cell a, and the next grid cell to be selected is b. i The set of candidate grid cells in the neighboring grid is g represents the number of candidate grid cells, which is determined by the position of the current grid cell. For the ant to move from a to b in the k-th iteration... i pheromone concentration; (3c2) From a to b i The reciprocal of the unit cost of the route, denoted as for: ； in, For ab i The unit cost of a segment path is obtained by the following formula: ； Where S is ab i The number of grid cells traversed by the segment path. For ab i The length of the segment path within the j-th grid cell. For ab i The raster cost of the j-th raster traversed by the segment path. For ab i The length of the path segment; (3c3) Introduce the directional guidance parameter o, which is defined as follows: ； in, From the current grid position a to the grid cell to be selected b i The angle between the line connecting a and the line connecting a to the endpoint; (3c4) The improved ant colony algorithm's grid selection probability is obtained from steps (3c2) and (3c3). : ； in, That is, the ant moves from a to b. i The probability of.

7. The intelligent transmission line selection method based on adaptive resolution grid and improved ant colony algorithm according to claim 1, characterized in that: Step (3e) specifically includes the following steps: (3e1) The pheromone concentration is updated with each iteration, and the pheromone concentration of the grid is updated according to the following formula: ； in, The pheromone left by the m-th ant when it completes its search in the k-th iteration and moves from grid a to grid b is calculated using the following formula: ； in, Let Q be the cost of the route obtained by ant m's search, and Q be the correction factor. (3e2) After each iteration, To update, follow these steps: ； in, Here is the directional volatility coefficient. These are the control parameters for o in the (k+1)th iteration. represents the control parameters for the k-th iteration.

Images