Substation site optimization method based on remote sensing image recognition

Through remote sensing image recognition and GIS analysis technology, the site selection of substations is optimized, which solves the problems of low site selection efficiency and high cost in traditional methods, and realizes intelligent optimization of substation sites, improves site selection efficiency and reduces costs.

CN118966640APending Publication Date: 2024-11-15四川电力设计咨询有限责任公司
View PDF 0 Cites 1 Cited by

Patent Information

Application Number
CN202410997401.2
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-24
Publication Date
2024-11-15

AI Technical Summary

Technical Problem

Due to manual experience limitations, traditional substation site selection methods cannot traverse all feasible sites, resulting in the optimal solution not being discovered, reducing site selection efficiency and increasing cost.

Method used

Using a method based on remote sensing image recognition, combined with GIS analysis technology, through slope judgment and prohibited area identification, building complex identification methods are designed, entry and exit channel planning and excavation and filling cost calculation, and introducing factors for entry and exit routes of substations, using exhaustive search The algorithm realizes the best cost site selection.

Benefits of technology

It improves the efficiency of substation site selection, reduces site selection costs, helps designers quickly realize site selection in the indoor stage, and ensures optimal cost of site locations.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118966640A_ABST
    Figure CN118966640A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of substation site selection, in particular to a substation site optimization method based on remote sensing image recognition. According to the substation site selection method based on remote sensing image recognition, remote sensing satellite images and a GIS analysis technology are utilized, a feasible site selection area is determined through slope judgment and forbidden area recognition, a building group recognition method based on an adjacency operator is designed, and incoming and outgoing line channel planning, station entering road cost and excavation and filling cost calculation methods are optimized; according to the method, transformer substation incoming and outgoing line road factors are introduced as cost factors, cost optimization is taken as a target, intelligent optimization of a transformer substation site is realized, designers are helped to quickly realize site selection in an indoor stage, site selection efficiency is improved, and site selection cost is reduced.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of substation site selection, and in particular to a substation site optimization method based on remote sensing image recognition. Background Art

[0002] Traditional substation site selection usually adopts a combination of manual survey and computer-aided site selection. First, designers combine topographic maps and existing thematic data, such as meteorological conditions, planning areas, mining areas, basic farmland, etc., to select indoor sites. Since the information carried by topographic maps cannot fully reflect the actual distribution of on-site objects, designers need to conduct on-site surveys and then optimize the site. Usually, there are multiple feasible sites. Designers compare the costs of multiple candidate sites based on their own experience and make decisions after comprehensively considering various factors.

[0003] However, since designers cannot traverse all feasible sites and each designer has different experience and technical levels, the final recommended site plan may not be the optimal solution for substation site selection, thereby reducing site selection efficiency and increasing substation site selection costs. Summary of the invention

[0004] In order to solve the above technical problems, the present invention provides a substation site optimization method based on remote sensing image recognition, which is characterized by comprising the following steps:

[0005] S1. Obtain the site shape, size and site selection area of ​​the substation, collect remote sensing satellite images, digital elevation data and vector thematic data in the site selection area, and establish a raster network;

[0006] S2. Identify the ground features of each grid point in the grid network based on the remote sensing satellite image of the grid in the selected area, and assign ground feature costs to the ground features of each grid point based on the identification results of each grid point and the preset ground feature cost factors;

[0007] S3. Calculate the slope of each grid point in the grid network according to the digital elevation data of the grid in the site selection area, assign terrain cost factors to the slope of each grid point according to the slope calculation results of each grid point and the preset terrain cost factors, and make a feasibility judgment on the slope of each grid point in the grid network based on the preset safety slope and the site area in combination with the vector thematic data to obtain a feasible site selection area grid;

[0008] S4. Based on the grid of feasible site selection area, the shape and size of the substation site, a sliding search is performed at a given step length or north deflection angle to obtain a feasible site selection site, wherein the feasible site does not occupy the grid of the infeasible site selection area, the north deflection angle direction of the site is due north, and the north deflection angle can be adjusted according to a certain step difference;

[0009] S5. Perform building group analysis on the grid network of the feasible station site and the surrounding area to obtain the building group area grid;

[0010] S6. According to the substation inlet and outlet information and the building complex area grid, along the inlet and outlet direction, the outlets are arranged at equal intervals from the middle to both sides, and the available channels are calculated and checked whether they meet the substation inlet and outlet information. If the outlet cannot meet the substation inlet and outlet information, repeat steps S4-S6, otherwise execute step S7; the substation inlet and outlet information includes: voltage level, number of loops, cross arm width, demolition distance, and reserved channel width;

[0011] S7. According to the feasible station sites, the cost of excavation and filling of the site is calculated to obtain the excavation and filling cost of the site;

[0012] S8. According to the feasible station site selection, the station access road planning and cost calculation are performed to obtain the station access road planning and station access road cost;

[0013] S9. Based on the cost calculation results in steps S7 and S8, a substation site selection cost target model is constructed, and all feasible substation sites are searched through an exhaustive search algorithm to obtain the substation site with the optimal cost.

[0014] Furthermore, the vector thematic data at least includes basic farmland vector data, primary forest land vector data, earthquake zone vector data, environmental protection area vector data and urban and rural planning area vector data.

[0015] Furthermore, the step S3 includes the following steps:

[0016] S301. Collect digital elevation data of the site selection area, divide the target site selection area into grids, and obtain a grid network of the target site selection area;

[0017] S302. Calculate the slope value of each grid in the grid network of the site selection area, compare it with the preset safety slope, and connect the adjacent grid nodes with a slope greater than the safety slope into several connected areas;

[0018] S303. According to the digital elevation data of the site selection area, the area of ​​each connected area is calculated, and the feasibility of the connected area is judged based on the site area. The grid whose connected area is smaller than the site area is combined with the grid containing the vector thematic data, and the grid is identified as an infeasible area grid;

[0019] S304: Superimpose and compare the infeasible area grid with the site selection area grid to obtain a feasible site selection area grid.

[0020] Furthermore, the building group analysis in step S5 adopts a building group analysis method based on an adjacency operator, which specifically includes the following steps:

[0021] S501. Establish a pane network in the feasible site selection area, calculate the building coefficient of each pane, and compare it with the building coefficient threshold to identify all the building panes in the feasible site selection area. The specific building coefficient comparison model is:

[0022]

[0023] Wherein t is the building coefficient, v is the number of building grids in the building pane; u is the total number of grids in the building pane; a is the building coefficient threshold, which is determined by actual site conditions; the building coefficient threshold is determined based on the demolition cost of the building in the pane; the pane step length is greater than the feasible site selection area grid step length, and one pane includes at least two feasible site selection area grids;

[0024] S502. According to the identified building panes, determine the set of building panes to be checked, initialize the detection coefficient of each building pane to be checked, randomly select multiple building panes to be checked as search starting points, and establish an initial building group set. The detection coefficient is used to represent the search status of the building pane. If the detection coefficient is 0, it means that the current building pane has not been detected. If the detection coefficient is 1, it means that the current building pane has been detected, and the detection coefficient is initialized to 0;

[0025] S503. Starting from the starting point, the adjacent operator is used to diffuse the search in all directions. The diffuse search in all directions is to search in the four directions of the upper, lower, left and right of the same horizontal plane of the starting window pane with the starting window pane as the center, add the adjacent building window panes to the corresponding building group set, merge the window panes with a detection coefficient of 1, and continue to diffuse until no new window panes are added;

[0026] The adjacency operator is used to determine whether adjacent panes are all building panes, and its calculation formula can be expressed as:

[0027] e ij =t ij ·t ij+1

[0028] Where e is the building window pane connectivity coefficient. If it is 1, the adjacent building window panes are connected, otherwise they are not connected. ij is the building coefficient of the i-th horizontal and j-th vertical window pane;

[0029] S504. Calculate the connectivity of all building panes, determine the building group according to the building group threshold, and finally obtain the building group area grid within the site selection area.

[0030] Further, the calculation of the cut and fill volume of the site address in step S7 includes the following steps:

[0031] S701. Based on the principle of cut and fill balance, calculate the cut and fill volume of the feasible site selection area through GIS digital elevation data;

[0032] S702. Inside and outside the feasible site selection area, perform slope cutting judgment and calculation according to a fixed step length and a set slope angle to obtain the retaining wall volume of the feasible site selection area;

[0033] S703. Carry out a leveling cost modeling for the feasible site selection area and calculate the cut and fill cost of the feasible site selection area. The leveling cost model is expressed as:

[0034] cost min (volume) = f * p1 + (c - f) * p2 + d * p4 + s1 * p5 + s2 * p6, c > f;

[0035] cost min (volume) = f * p1 + (f - c) * p3 + d * p4 + s1 * p5 + s2 * p6, c < f;

[0036] where cost min (volume) is the cut and fill cost of the feasible site selection area; c is the cut volume of land leveling, in m 3 ; f is the fill volume of land leveling, in m 3 ; d is the retaining wall volume, in m 3 ; s1 is the site occupation area, in m; s2 is the house demolition area within the site; p1 is the cut and fill balance, that is, when c = f, the comprehensive unit price of cut and fill, in yuan / m 3 ; p2 is the unit price of transporting surplus soil when the cut volume is greater than the fill volume, that is, c > f, in yuan / m 3 ; p3 is the unit price of purchasing external soil when the cut volume is less than the fill volume, that is, c < f, in yuan / m 3 ; p4 is the unit price of the retaining wall, in yuan / m 3 ; p5 is the unit price of land acquisition, in yuan / m 2 ; p6 is the unit price of house demolition, in yuan / m 2 .

[0037] Further, the slope cutting judgment and calculation in step S702 are carried out by cutting according to a certain step length d and a set slope angle α or filling according to a slope angle β;

[0038] If after cutting and filling, the slope surface at this step does not enter the prohibited construction area and the flood inundation line, it is judged that the slope surface is feasible and cutting and filling can be carried out;

[0039] If the slope enters the prohibited construction area and the flood level inundation line, the slope is judged to be unfeasible and a retaining wall is used instead. The retaining wall is a gravity reinforced concrete retaining wall.

[0040] Furthermore, to meet environmental protection requirements, the slope projection area should not be greater than 3% of the total site area. If it is greater than 3% of the total site area, it will be replaced with a retaining wall. Areas with large cut or fill volumes can be searched for retaining walls at a certain step length d until the 3% requirement is met.

[0041] The above conditions shall be considered at the same time until the excavation and filling are balanced. If the excavation and filling cannot be balanced, the excavation shall be considered to be greater than the filling. It is prohibited to have the excavation less than the filling and purchase earthwork from outside.

[0042] Furthermore, the access road calculation in step S8 is performed based on the station location and the distribution of surrounding roads, combined with the cost-based shortest path algorithm to calculate the access road cost, and its calculation model is:

[0043] cost min (road) = f'*P4;

[0044] Among them, cost min (road) is the optimal road cost; f' is the optimal road length; P4 is the comprehensive road cost.

[0045] Furthermore, the cost target model in step S9 is constructed based on an exhaustive search algorithm, and the model can be expressed as:

[0046] F=cost min (volume)+cost min (road);

[0047] Where F is the optimal substation site, cost min (volume) is the cost of cut and fill in the feasible site selection area, cost min (road) is the optimal road cost.

[0048] The beneficial effects of the present invention are:

[0049] By providing a substation site optimization method based on remote sensing image recognition, using remote sensing satellite images and GIS analysis technology, the feasible site selection area is determined by slope judgment and prohibited area identification, and a building group identification method based on adjacency operator is designed. The planning of access channel, access road cost and excavation and filling cost calculation method are optimized, and the substation access road factor is introduced as a cost factor. With the goal of cost optimization, the intelligent optimization of substation site is realized, which helps designers to quickly realize site selection in the indoor stage, improve site selection efficiency and reduce site selection cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Figure 1 , a flow chart of a substation site optimization method based on remote sensing image recognition according to the present invention.

[0051] Figure 2 , a grid schematic diagram for calculating the grid slope of the substation site selection area in an embodiment of the present invention.

[0052] Figure 3 , a schematic diagram of the building complex judgment process in an embodiment of the present invention.

[0053] Figure 4 , schematic diagram of determining the building complex of the substation entrance and exit passages in an embodiment of the present invention.

[0054] Figure 5 , schematic diagram of substation inlet and outlet line planning in an embodiment of the present invention.

[0055] Figure 6 , schematic diagram of feasibility determination of incoming and outgoing lines of a substation in an embodiment of the present invention.

[0056] Figure 7 , a schematic diagram of calculating the grid fill and cut volume in the substation site selection area in an embodiment of the present invention.

[0057] Figure 8 , a side view of a downward slope outside a substation site in an embodiment of the present invention.

[0058] Fig. 9 , the segmentation diagram for calculating cut and fill within the station area in the embodiment of the present invention.

[0059] Fig.10 , a schematic diagram of a starting point and multiple end points in an embodiment of the present invention.

[0060] Fig.11 , spatial distribution of preferred substation sites in the embodiments of the present invention.

[0061] Fig.12 , a schematic diagram of the optimal site selection scheme in an embodiment of the present invention. DETAILED DESCRIPTION

[0062] In order to enable those skilled in the art to better understand the content of the present invention and make the purpose, technical solution and advantages of the present invention clearer, the present invention is further described in detail below in conjunction with the embodiments and the accompanying drawings. The illustrative embodiments of the present invention and their description are only used to explain the present invention and are not intended to further limit the present invention.

[0063] like Figure 1 As shown, an embodiment of the present invention provides a method for optimizing a substation site based on remote sensing image recognition, comprising the following steps:

[0064] S1. Obtain the substation voltage level, capacity, planned outgoing line and other conditions of the substation to be built, and determine the site shape, site area and site selection area. Use GIS geographic information system to load station selection high-definition remote sensing satellite images, DEM digital elevation data and collected vector thematic data such as planning areas, mining areas, ecologically sensitive areas, basic farmland, primary forest land, seismic zones, power grid spatial data, etc.

[0065] S2. Using the end-to-end semantic segmentation PSPNet network model, the remote sensing satellite images in the selected area are used to identify ground features, annotate them in raster form, and save them according to different raster codes to obtain the following: Figure 2 As shown in the image recognition results, the land features include at least houses, roads, rivers, and lakes.

[0066] The PSPNet network model is a deep learning model for semantic segmentation, which aims to segment an image into regions with semantic meaning, that is, to assign each pixel in the image to its corresponding semantic category. The idea of ​​pyramidization is used to capture information of different scales in the image to improve segmentation accuracy.

[0067] S3. Determine the safety slope in the current substation selection requirements: the flat slope within the station site shall not be greater than 15°.

[0068] S4. Collect DEM data in the site selection area and divide the site area into grids with a step size of 2m.

[0069] S5. Select a 3*3 grid, calculate the slope value corresponding to the current grid, express the elevations of the 9 grid center points as e5, e2, e6, e1, e, e3, e8, e4, e7, and calculate the slope value of grid e:

[0070]

[0071] Slope we Slope is the slope in the x direction; sn is the slope in the y direction; Cellsize is the grid interval length; Slope is the slope.

[0072] S6. Repeat step S5 to calculate the slope of all grids in the site selection area.

[0073] S7. After obtaining all grid slopes, divide the grids into two types: slope greater than 15° and slope less than 15°. Connect adjacent grid nodes with slopes less than 15° into several areas, and judge each connected area one by one. If the area of ​​the area is less than the set threshold, it means that the area is not feasible. The site selection areas with an average slope greater than 15° and an area less than the set threshold and the prohibited construction areas such as basic farmland, primary forest land, core protection area, urban and rural planning area, and housing complex are regarded as prohibited site selection areas, and the rest of the areas are feasible site selection areas.

[0074] S8. In the feasible site selection area, according to the given site shape and size, a sliding search for a feasible site is performed at a given step length or north deflection angle. The feasible site does not occupy the grid of the infeasible site selection area. The north deflection angle of the site is due north, and the north deflection angle can be adjusted according to a certain step difference.

[0075] S9. Perform building group analysis based on the feasible site searched in S8. The building group analysis steps used in this embodiment are:

[0076] First, a window pane with a step length of 50 meters is used to overlay the image grid extracted by artificial intelligence recognition to determine whether the window pane is a building window pane;

[0077] Secondly, the building coefficient t is calculated and it is determined whether it is greater than the building coefficient threshold a. In this embodiment, a=0.7. The sliding calculation is performed to determine whether each window pane in the site selection area and the surrounding area (determined by the channel length) is a building window pane. The specific building coefficient comparison model is:

[0078]

[0079] Wherein t is the building coefficient, v is the number of building grids in the building pane; u is the total number of grids in the building pane; a is the building coefficient threshold, which is determined by actual site conditions; the building coefficient threshold is determined based on the demolition cost of the building in the pane; the pane step length is greater than the feasible site selection area grid step length, and one pane includes at least two feasible site selection area grids;

[0080] Finally, the building group analysis method based on adjacency operator designed by the present invention is adopted. Figure 3 As a schematic diagram of the analysis process, the connectivity q of the building pane is calculated, and whether it is a building complex is determined according to the building complex threshold. If q is greater than the threshold, it means that the building area is a building complex. The adjacency operator calculation formula can be expressed as:

[0081] e ij =t ij ·t ij+1

[0082] Where e is the building window pane connectivity coefficient. If it is 1, the adjacent building window panes are connected, otherwise they are not connected.ij is the building coefficient of the i-th horizontal and j-th vertical window pane;

[0083] like Figure 4 As shown in the figure, there are 5 connected building areas. Assuming that the building group threshold q = 4, then Figure 4 Among them, 1, 2, and 4 are a building complex, and the number of window panes connecting 3 and 5 is only 3, so it is not a building complex.

[0084] S10. Planning and verification of inlet and outlet channels.

[0085] The specific steps of determining the substation channel planning in this embodiment are:

[0086] S1001. Determine the outgoing line information of the substation.

[0087] S1002. Specify the number of outgoing lines N, the outgoing line interval width B, the reserved channel width W, and the channel length L on the 4th side of the substation, where the reserved channel widths of single-circuit transmission lines corresponding to different voltage levels are shown in Table 1.

[0088] Table 1 Channel reserved width table

[0089] Voltage level (kV) Number of circuits Channel reserved width (m) 500 one 18 330 one 14 220 one 12 110 one 8

[0090] S1003. Arrange the outgoing wires at equal intervals according to the number of outgoing wires N and the outgoing wire interval width B.

[0091] S1004. Figure 5 As shown, a two-dimensional XY plane coordinate system is established with the midpoint of the line interval on this side as the origin. When the number of line outgoing N is an odd number, the line outgoing direction of the intervals with line numbers 1 to (N-1) / 2 is 90 to 180°; the line outgoing direction of the intervals with line numbers (N+1) / 2 is 45 to 135°; the line outgoing direction of the intervals with line numbers (N+3) / 2 to N is 0 to 90°. When N is an even number, the line outgoing direction of the intervals with line numbers 1 to N / 2 is 90 to 180°; the line outgoing direction of the intervals with line numbers N / 2+1 to N is 0 to 90°.

[0092] S1005. Calculate the channel width Di between buildings and from buildings to the station

[0093] S1006. Alternately judge each outgoing line channel from both sides to the center: the first step is to judge interval 1, the second step is to judge interval N, the third step is to judge interval 2, the fourth step is to judge interval N-1... until all interval outgoing line judgments are completed.

[0094] S1007. With L as the scanning radius and the interval as the center, perform a scan check at every 10° distance line Di between the building and each outgoing line direction as described in step S1504. If the channel distance value Di>W, it means that the outgoing line is feasible, and the distance value Di=Di-W is updated at the same time.

[0095] S1008. Use the same rule to determine whether other incoming and outgoing lines are feasible. If all incoming and outgoing lines are feasible, then the current substation site is feasible. The schematic diagram is as follows: Figure 6 , enter S11, otherwise return to S9 to search for other feasible sites.

[0096] S11. For the feasible site, calculate the excavation and filling of the land leveling in the site area and the excavation, filling or retaining wall of the land leveling outside the site area. The schematic diagram of land leveling outside the feasible site area in S11 is as follows: Figure 7 As shown; the schematic diagram of the slope settlement process is as follows Figure 8 shown.

[0097] Furthermore, in S11, the cut and fill calculation steps involved in this embodiment are as follows:

[0098] S1101. Divide the feasible site selection area into a group of small grids along the length and width of the site with a certain step size S.

[0099] S1102. Find the ground elevation corresponding to the four points on each small grid, such as Fig. 9 As shown, assuming that the four points are k0, k1, k2, and k3, the corresponding elevations are alt0, alt1, alt2, and alt3, and the elevation of the station plane is alt. The elevation differences between the four points and the station plane are:

[0100] h0=alt-alt0;

[0101] h1=alt-alt1;

[0102] h2=alt-alt2;

[0103] h3=alt-alt3;

[0104] Decompose a small grid into two triangles, namely T1 (k0, k1, k3) and T2 (k2, k1, k3), and calculate the volume formula based on the irregular triangular prism:

[0105] V=Area*(h1+h2+h3) / 3

[0106] Where V is the volume, Area is the projected area of ​​the triangular face, and h1, h2, and h3 are the heights of the three sides of the triangular prism.

[0107] Substituting into the formula, we get the volume of T1 is PArea*(h0+h1+h2) / 3, and the volume of T2 is PArea*(h0+h2+h2) / 3; where PArea is the projection area of ​​the triangle, PArea=s*s / 2; if V is greater than 0, it is the fill volume, and if it is less than 0, it is the excavation volume.

[0108] S1103. Repeat S1102 in this way to calculate the cut and fill volumes of all small grids, and sum them up to get the cut and fill volume of the station.

[0109] S1104. For the site selection area, excavation is performed according to a certain step length d and a set slope angle α, or filling is performed according to a slope angle β. If after excavation and filling, the slope surface under the step length does not enter the prohibited construction area and the flood inundation line, then the slope surface is judged to be feasible and excavation and filling can be performed. If the slope surface enters the prohibited construction area and the flood level inundation line, then the slope surface is judged to be infeasible and a retaining wall is used instead. The retaining wall is a gravity reinforced concrete retaining wall.

[0110] In addition, to meet environmental protection requirements, the slope projection area should not be greater than 3% of the total site area. If it is greater than 3% of the total site area, it will be replaced with a retaining wall. Areas with large excavation or filling volume can be searched at a certain step length d for retaining wall replacement until the 3% requirement is met.

[0111] The above conditions shall be considered at the same time until the excavation and filling are balanced. If the excavation and filling cannot be balanced, the excavation shall be considered to be greater than the filling. It is prohibited to have the excavation less than the filling and purchase earthwork from outside.

[0112] Further, in S11, when the retaining wall is used in this embodiment, it is necessary to determine the type of retaining wall (cut retaining wall, fill retaining wall), the ground elevation alt_m of the midpoint m of the station edge S1 where the retaining wall is located, if the arithmetic value of alt_m is greater than the reference height alt of the station site plane, it means that S1 is a cut retaining wall, otherwise, it is a fill retaining wall, and the wall height is calculated at the same time, and the wall height calculation formula is:

[0113] H=|alt_m-alt|+1.5

[0114] 1.5 is the default embedded depth of retaining wall, in m. The unit volume of retaining wall m can be determined by checking the standard design table based on retaining wall type and wall height. 3 The parameters of the excavation retaining wall and the fill retaining wall are shown in Table 2 and Table 3 respectively:

[0115] Table 2 Excavation retaining wall parameters

[0116]

[0117]

[0118] Table 3 Parameters of Fill Retaining Wall

[0119] model Wall height(m) <![CDATA[Volume (m 3 )]]> ZBZ 2 1.35 ZJB2.5 2.5 2.01 ZBZ 3 2.76 ZJB3.5 3.5 3.63 ZBZ 4 4.57 ZJB4.5 4.5 5.61 ZB5 5 7.46 ZJB5.5 5.5 8.82 ZB6 6 10.28 ZJB6.5 6.5 11.81 ZBZ 7 13.42 ZJB7.5 7.5 15.19 ZB8 8 16.98

[0120] S1105. Perform leveling cost modeling on the feasible site selection area, and calculate the filling and excavation costs of the feasible site selection area. The leveling cost model is expressed as:

[0121] cost min (volume) = f * p1 + (c - f) * p2 + d * p4 + s1 * p5 + s2 * p6, c > f;

[0122] cost min (volume) = f * p1 + (f - c) * p3 + d * p4 + s1 * p5 + s2 * p6, c < f;

[0123] Where cost min (volume) is the filling and excavation cost of the feasible site selection area; c is the excavation volume of land leveling, in m 3 ; f is the filling volume of land leveling, in m 3 ; d is the retaining wall volume, in m 3 ; s1 is the occupied area of the site, in m; s2 is the demolished area of houses within the site; p1 is the balance of filling and excavation, that is, when c = f, the comprehensive unit price of filling and excavation, in yuan / m 3 ; p2 is the unit price of transporting surplus soil out when the excavation volume is greater than the filling volume, that is, c > f, in yuan / m 3 ; p3 is the unit price of purchasing external soil when the excavation volume is less than the filling volume, that is, c < f, in yuan / m 3 ; p4 is the unit price of the retaining wall, in yuan / m 3 ; p5 is the unit price of land acquisition, in yuan / m 2 ; p6 is the unit price of house demolition, in yuan / m 2 .

[0124] S12. According to the location of the substation site and the distribution of surrounding roads, combine with the shortest path algorithm based on cost to plan and calculate the cost of the access road to the substation.

[0125] The basic conditions that the access road to the substation needs to meet are shown in Table 4:

[0126] Table 4 Parameters Table of Access Roads to Substations of Different Voltage Levels

[0127]

[0128] Furthermore, in S12, the shortest path algorithm based on cost in this example is described as follows:

[0129] The traditional Digest algorithm has only one starting point and one end point. In the process of path finding, the algorithm ends when the end point is found. For station road optimization, the starting point is the center of the station site, and the end point may be any road within a certain radius r from the station site. In this case, there may be multiple "potential end points". In the figure below, all grid centers that intersect with the road are "potential end points". The algorithm ends when the first "potential end point" is found.

[0130] The entire area is interpolated into a grid grid with a fixed step size. All grid points that are passed by roads are marked as "potential endpoints". When any such point is found during the algorithm, it is considered to be the endpoint and the algorithm can be terminated. Starting from the center of the station, start looking for the next point. The next point must be a neighbor node of the current node and must meet the maximum slope and other conditions. According to the traditional Digest algorithm, the nodes are traversed step by step until the first "potential endpoint", that is, the first road point, is found. The algorithm ends. Fig.10 Shown is a schematic diagram of the starting point and multiple end points.

[0131] For the two factors that affect the line selection, namely the slope and land use type, the spatial digital element layers are established respectively; then the influence degree of each grid point in each layer on the construction cost of the construction road path is given quantitatively, and its value is the attribute value of the grid point in the spatial digital layer (indicating the cost of the path passing through the grid point); then through spatial overlay analysis, the grid point attribute values ​​of each factor are subjected to linear algebraic operations to obtain the comprehensive cost grid layer. Combining the comprehensive cost with the Euclidean spatial distance calculation to form the cost distance, the optimal route connecting the starting and ending points can be generated by the Dijkstra algorithm, the optimal path search algorithm of GIS. Its model is expressed as:

[0132] costmin(road)=f*p5

[0133] costmin(road): optimal road cost; f: optimal road length, in meters; p5: comprehensive road cost, in yuan / m.

[0134] Furthermore, the cost calculation method of this embodiment in S12 is:

[0135] Calculate the cost distance within the site selection area. The cost distance is: a distance from the source location weighted by the cost factor (or the cumulative travel cost). It is a distance expressed in cost units, not in geographic units. The cost distance equation can express the minimum cumulative cost distance from each grid point to the nearest point. The purpose is to identify the minimum cost direction and path between the selected target point and the source point. Expressed in the form of nodes and connecting lines: Each grid point is regarded as a node, and the nodes are connected by connecting lines. Each node stores data including the cost value passing through the point and the cumulative cost value. The specific steps are:

[0136] Each grid point records a coordinate. The elevation in the coordinate is the real terrain and elevation, and also includes the road cost required to pass through the path. The road cost of the grid point is calculated by weighting the costs of different factors. The formula is as follows:

[0137] P=P1*w1+P2*w2

[0138] P: road cost required to pass through the path; P1: slope cost passing through the point; P2: land type cost passing through the point; w1, w2: cost weights of slope and land type.

[0139] The slope of each point is automatically calculated by the GIS program, and the land type is extracted by intelligent image recognition and then vectorized. After calculating the road cost of each grid point, it is also necessary to calculate the cost of walking from a point to its connected point. The traditional approach is to directly add the cost of the two points and divide by 2. That is:

[0140] Cost=factor*(Cost1+Cost2) / 2

[0141] Here Cost1 and Cost2 are the cost values ​​of point 1 and point 2 respectively, and factor is the horizontal distance adjustment coefficient for straight line and diagonal path optimization. If a straight line is taken, the coefficient is 1; if a diagonal line is taken, the coefficient is This method cannot reflect the impact of height difference on the result. Generally speaking, the routing cost of two points with height difference is greater than that of two points with the same height. Therefore, this embodiment introduces the concept of maximum pitch angle, so that in the process of road optimization, a flat path is taken as much as possible, which is closer to the actual situation. The formula is as follows:

[0142] Cost=factor*(Cost1+Cost2) / 2*pow(2,a)

[0143] a refers to the height difference between two points divided by the actual distance between the two points, with a value range of 0-1. When the pitch angle of the two points is greater than an angle (temporarily set at 15°), it means that the cost is too high, and other paths that meet the conditions are searched. If there are other paths, other paths are selected first, otherwise this path is used. This restriction takes into account the restrictions on the road slope when the machinery enters the site during mechanized construction of the road. A large slope does not meet the construction standards.

[0144] Furthermore, in S12, this embodiment involves cost factor assignment:

[0145] Set cost factors and assign values ​​to them. Comprehensive cost refers to a general measure of the impact of multiple factors on the construction cost or expense of the road route plan. A high comprehensive cost of the route plan means high construction costs and low comprehensive benefits; a low comprehensive cost of the route plan means low construction costs and high comprehensive benefits.

[0146] The cost value of each factor is determined by combining the Delphi method with the Analytic Hierarchy Process (AHP). The Delphi method and the Analytic Hierarchy Process are two methods widely used in operations research theory. The Delphi method, also known as the expert survey method, is essentially a feedback anonymous inquiry method. Its general process is: after obtaining the opinions of experts on the problem to be predicted, the opinions are sorted, summarized, and counted, and then anonymously fed back to the experts, and opinions are again solicited, concentrated, and fed back again until a consensus is reached. The Analytic Hierarchy Process treats a complex multi-objective decision-making problem as a system, decomposes the goal into multiple small goals or criteria, and then decomposes it into several levels of multiple indicators (or criteria, constraints). The hierarchical single ranking (weight) and total ranking are calculated through the fuzzy quantification method of qualitative indicators to make goal (multiple indicators) and multi-scheme optimization decisions.

[0147] In this embodiment, the cost values ​​of the two influencing factors are 1, 3, 5, 7, and 9, which are stored in each grid node, indicating the high or low cost of the route passing through the node. That is, the higher the value, the higher the cost of the route passing through the grid point, and the lower the value, the lower the comprehensive cost of passing through the grid point.

[0148] This embodiment takes into account the main influencing factors of temporary roads for mechanized construction of transmission lines, and takes slope and land use type as comprehensive cost considerations for optimizing the construction road path.

[0149] Slope: It affects the difficulty of building the construction road and also determines the difficulty of the construction machinery entering the site. The steeper the slope, the more difficult it is for the construction machinery to enter the site, and the higher the cost and process requirements of road construction.

[0150] Before introducing the slope cost factor assignment, it is necessary to first determine the terrain of the preferred range of the road path. The specific terrain type can be determined by combining the slope and altitude (the altitude of the starting point and the end point is compared to take the larger value). As shown in Table 5, the terrain type when the slope is 0-10°: plain, plateau, basin; the terrain type when the slope is greater than 10°: hills, mountains. Determine according to the slope of the starting point and the end point (the larger value of the two), the default value of 0-10° is used when the slope is ≤10°, and the default value of 0-60° is used when the slope is greater than 10°.

[0151] Table 5 Terrain type identification

[0152]

[0153]

[0154] For plains, plateaus, and basins, the slope range is 0-10°, the cost value for 0-2° is 1, the cost value for 2-4° is 3, and the cost value for 4-6° is 5; the cost value for 6-8° is 7, and the cost value for 8-10° is 9; for hills and mountains, the slope range is 0-60°, the cost value for 0-5° is 1, the cost value for 5-10° is 3, and the cost value for 10-20° is 5; the cost value for 20-40° is 7, and the cost value for 40-60° is 9. Landform elements: according to their different effects on ecological balance and the difficulty of construction, the cost value of ditches, rivers, wetlands, beaches, ponds, etc. is 9, the cost value of woodlands, orchards, shrub forests, etc. is 7, the cost value of irrigated land, paddy fields, dry land, etc. is 5, the cost value of grasslands, rural homesteads, etc. is 3, and the cost value of highway land, rural roads, industrial land, etc. is 1.

[0155] Five experts were invited to evaluate the relative importance of the two cost factors, and the weight values ​​of each evaluation factor were calculated, as shown in Table 6:

[0156] Table 6 Classification, value assignment and weight of terrain cost factors and feature cost factors

[0157]

[0158] S13. Use the "exhaustive search" algorithm to traverse all possible candidate solutions for substation sites, and select multiple optimal solutions that meet specific optimization criteria. Use the exhaustive search optimization algorithm to find the optimal solution process:

[0159] S1301. Using cost optimization as the objective function of the exhaustive search algorithm can be expressed as:

[0160] F=cost min (volume)+cost min (road)

[0161] Where F is the optimal solution of the objective function obtained through loop traversal calculation, which is used as the substation site with the optimal cost; cost min (volume) is the earthwork cost of the station, including the cost of filling and cutting, retaining wall cost and land acquisition cost; cost min (road): cost of the road into the station.

[0162] S1302. Define the search space: Define the space to be searched, determine the boundary of the problem and the scope of the search, and set the search span and rotation step size.

[0163] The search position combined with the step size s and the rotation angle θ can be expressed as:

[0164] P_{new}(x′,y′)=P_{current}(x,y)(s)+f(v)(θ)

[0165] Among them, P_{current} is the current position, and P_{new} is the new position after moving step s and rotating by θ.

[0166] S1303. Design state representation: According to the specific definition of the problem, calculate the state of each position and encode the state for optimization. Use conditions such as slope / prohibited area / outline to eliminate the candidate results that do not meet the conditions. For example: if the slope of a certain position is greater than 15°, it will be eliminated.

[0167] S1304. Generate candidate solutions: Generate candidate solutions for each possible state according to the principle of the algorithm.

[0168] S1305. Update the optimal solution: Each time a candidate solution is generated, it is evaluated using the objective function. If a single optimal solution is to be found, the currently found global optimal solution is iterated and recorded. If multiple optimal solutions are to be found, such as the top 20 optimal solutions, a fixed-size priority queue is maintained to store the top 20 optimal solutions. Fig.11 This is a feasible site distribution in the embodiment.

[0169] S1306. Termination condition judgment: Check whether the search is traversed. If so, jump to step 6; otherwise, return to step 4.

[0170] S1307. Output result: Output the optimal solution obtained as the substation site selection result with the best cost. Fig.12 It is the optimal site selection plan, including substation site, access channel planning, filling and excavation, demolition volume and retaining wall, etc.

[0171] S1308. The algorithm ends.

[0172] The present invention is explained from the viewpoints of purpose of use, efficiency, progress and novelty, and the practical progress it has is in line with the functional enhancement and use requirements emphasized by the Patent Law. The above description and drawings of this application are only preferred embodiments of this application, and are not intended to limit this application. Therefore, all structures, devices, features, etc. that are similar or identical to those of this application, that is, all equivalent replacements or modifications made in accordance with the scope of the patent application of this application, should fall within the scope of protection of the patent application of this application.

[0173] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

[0174] The present invention is not limited to the above optional implementations, and anyone can derive other various forms of products under the enlightenment of the present invention. The above specific implementations should not be understood as limiting the scope of protection of the present invention. The scope of protection of the present invention should be based on the definition in the claims, and the description can be used to interpret the claims.

Claims

1. A method for optimizing substation site based on remote sensing image recognition, characterized in that: The following steps are involved: S1. Obtain the site shape, size and site selection area of ​​the substation, collect remote sensing satellite images, digital elevation data and vector thematic data in the site selection area, and establish a raster network; S2. Identify the ground features of each grid point in the grid network based on the remote sensing satellite image of the grid in the selected area, and assign ground feature costs to the ground features of each grid point based on the identification results of each grid point and the preset ground feature cost factors; S3. Calculate the slope of each grid point in the grid network according to the digital elevation data of the grid in the site selection area, assign terrain cost factors to the slope of each grid point according to the slope calculation results of each grid point and the preset terrain cost factors, and make a feasibility judgment on the slope of each grid point in the grid network based on the preset safety slope and the site area in combination with the vector thematic data to obtain a feasible site selection area grid; S4. Based on the grid of the feasible site selection area, the shape and size of the substation site, a sliding search is performed at a given step length or north deflection angle to obtain a feasible site selection site, wherein the feasible site does not occupy the grid of the infeasible site selection area, and the north deflection angle direction of the site is due north; S5. Perform building group analysis on the grid network of the feasible station site and the surrounding area to obtain the building group area grid; S6. According to the substation inlet and outlet information and the building complex area grid, along the inlet and outlet direction of the feasible station site, the inlet and outlet channels are arranged at equal intervals from the middle to both sides according to the outlet interval, and the available inlet and outlet channels are calculated and checked whether they meet the substation inlet and outlet information. If the outlet of the feasible station site cannot meet the substation inlet and outlet information, repeat steps S4-S6, otherwise execute step S7; the substation inlet and outlet information includes: voltage level, number of loops, cross arm width, demolition distance, and reserved channel width; S7. According to the feasible station site selection, the site excavation and filling cost is calculated to obtain the site excavation and filling cost; S8. According to the feasible station site selection, the station access road planning and cost calculation are performed to obtain the station access road planning and station access road cost; S9. Based on the cost calculation results in steps S7 and S8, a substation site selection cost target model is constructed, and all feasible substation sites are searched through an exhaustive search algorithm to obtain the substation site with the optimal cost.

2. A method for optimizing substation site based on remote sensing image recognition according to claim 1, characterized in that: The vector thematic data at least include basic farmland vector data, primary forest land vector data, earthquake zone vector data, environmental protection zone vector data and urban and rural planning area vector data.

3. The method for optimizing substation site based on remote sensing image recognition according to claim 1, characterized in that: The step S3 comprises the following steps: S301. Collect digital elevation data of the site selection area, divide the target site selection area into grids, and obtain a grid network of the target site selection area; S302. Calculate the slope value of each grid in the grid network of the site selection area, compare it with the preset safety slope, and connect the adjacent grid nodes with a slope greater than the safety slope into several connected areas; S303. According to the digital elevation data of the site selection area, the area of ​​each connected area is calculated, and the feasibility of the connected area is judged based on the site area. The grid whose connected area is smaller than the site area is combined with the grid containing the vector thematic data, and the grid is identified as an infeasible area grid; S304: Superimpose and compare the infeasible area grid with the site selection area grid to obtain a feasible site selection area grid.

4. The method for optimizing substation site based on remote sensing image recognition according to claim 1, characterized in that: The building group analysis in step S5 adopts a building group analysis method based on an adjacency operator, which specifically includes the following steps: S501. Establish a pane network in the feasible site selection area, calculate the building coefficient of each pane, and compare it with the predetermined building coefficient threshold to identify all building panes in the feasible site selection area. The specific building coefficient comparison model is: Where t is the building coefficient, v is the number of building grids in the building pane; u is the total number of grids in the building pane; a is the building coefficient threshold; S502. According to the identified building panes, determine a set of building panes to be checked, initialize the detection coefficient of each building pane to be checked, randomly select a plurality of the building panes to be checked as search starting points, and establish an initial building group set. The detection coefficient is used to represent the search status of the building pane. If the detection coefficient is 0, it means that the current building pane has not been detected. If the detection coefficient is 1, it means that the current building pane has been detected, and the detection coefficient is initialized to 0; S503. Starting from the starting point, use the adjacency operator to diffuse the search in all directions, add the adjacent building panes to the corresponding building group set, merge the panes with a detection coefficient of 1, and continue to diffuse until no new panes are added; The adjacency operator is used to determine whether adjacent panes are all building panes, and its calculation formula can be expressed as: e ij =t ij ·t ij+1 Where e is the building window pane connectivity coefficient. If it is 1, the adjacent building window panes are connected, otherwise they are not connected. ij is the building coefficient of the i-th horizontal and j-th vertical window pane; S504. Calculate the connectivity of all building panes, determine the building group according to the building group threshold, and finally obtain the building group area grid within the site selection area.

5. The method for optimizing substation site based on remote sensing image recognition according to claim 1, characterized in that: The calculation of the site excavation and filling volume in step S7 includes the following steps: S701. Based on the cut-and-fill balance principle, the cut-and-fill volume of the feasible station site area is calculated using GIS digital elevation data; S702. Perform slope determination and calculation based on fixed step length and slope angle within and outside the feasible site selection area to obtain the retaining wall volume of the feasible site selection area; S703. Perform leveling cost modeling on the feasible site selection area, and calculate the filling and cutting costs of the feasible site selection area. The leveling cost model is expressed as: cost min (volume)=f*p1+(c-f)*p2+d*p4+s1*p5+s2**p6,c>f; cost min (volume)=f*p1+(f-c)*p3+d*p4+s1*p5+s2*p6,c<f; where cost min (volume) is the filling and excavation cost of the feasible site selection area; c is the excavation volume of land leveling, in m 3 ; f is the filling volume of land leveling, in m 3 ; d is the volume of retaining wall, in m 3 ; s1 is the occupied area of the site, in m; s2 is the demolished area of houses within the site; p1 is the balance of filling and excavation, that is, when c = f, the comprehensive unit price of filling and excavation, in yuan / m 3 ; p2 is the unit price of transporting surplus soil out when the excavation volume is greater than the filling volume, that is, when c > f, in yuan / m 3 ; p3 is the unit price of purchasing external soil when the excavation volume is less than the filling volume, that is, when c < f, in yuan / m 3 ; p4 is the unit price of retaining wall, in yuan / m 3 ; p5 is the unit price of land acquisition, in yuan / m 2 ; p6 is the unit price of house demolition, in yuan / m 2 .

6. A method for optimizing substation site based on remote sensing image recognition according to claim 5, characterized in that: The slope determination and calculation in step S702 is performed by excavating at a certain step length d and a set slope angle α or filling at a slope angle β; if after excavation and filling, the slope surface under the step length does not enter the prohibited construction area and the flood inundation line, then the slope surface is judged to be feasible and excavation and filling can be performed; If the slope enters the prohibited construction area and the flood level inundation line, the slope is judged to be unfeasible and a retaining wall is used instead. The retaining wall is a gravity reinforced concrete retaining wall.

7. The method for optimizing substation site based on remote sensing image recognition according to claim 1, characterized in that: The access road calculation in step S8 is performed based on the station location and the distribution of surrounding roads, combined with the cost-based shortest path algorithm to calculate the access road cost. The calculation model is: cost min (road)=f’*P4; Among them, cost min (road) is the optimal road cost; f' is the optimal road length; P4 is the comprehensive road cost.

8. The method for optimizing substation site based on remote sensing image recognition according to claim 1, characterized in that: The cost target model in step S9 is constructed based on an exhaustive search algorithm, and the model can be expressed as: F=cost min (volume)+cost min (road); Where F is the optimal substation site, cost min (volume) is the cost of cut and fill in the feasible site selection area, cost min (road) is the optimal road cost.

Citation Information

Cited By

  • Automatic site selection method and system for mountain fan hoisting platform

    CN121882654A