A layout method of an electric bicycle public charging station

Abstract

The application discloses a layout method of electric bicycle public charging stations, and relates to the field of city construction. The method firstly trains each basic site selection planning model of a mixed site selection planning model, and performs performance evaluation on the calculation results of each model; then, the comprehensive evaluation results of the mixed site selection planning model and multiple basic models are compared, the optimal decision threshold of the mixed site selection planning model is determined when the mixed site selection planning model performs best; subsequently, the grid data of a region to be planned is input into the mixed site selection planning model, and the preliminary screening result of the region to be planned is determined according to the model calculation result and the optimal decision threshold; finally, the preliminary screening result is further screened based on the convenience and safety requirements of electric bicycle riders, and the suitable site selection and layout region of the electric bicycle charging station is determined.

Description

Technical Field

[0001] This invention relates to the field of computer technology, and in particular to a layout method for public charging stations for electric bicycles. Background Technology

[0002] With the continuous advancement of urbanization, the expansion of urban areas, and the increase in various infrastructure facilities, while bringing tremendous convenience to urban residents' lives, it has also brought disadvantages such as increased commuting distances, costs, and time. To cope with the ever-changing living environment, electric bicycles have gradually become a major short-distance transportation tool for the public.

[0003] Existing research on the site selection of electric bicycle charging stations is mostly "demand-oriented," addressing the site planning problems of charging stations and battery swapping stations. Its approach to solving these problems is usually limited to alleviating charging demand, and the methods and technologies employed are relatively simplistic. This not only results in a lack of detailed and diverse site selection recommendations but also a disconnect from actual environmental conditions, making it difficult to provide effective decision support for charging station site selection.

[0004] Therefore, optimizing the existing layout methods of public charging stations for electric bicycles and constructing a reasonable site selection model to provide more effective decision support for site selection has become an urgent problem to be solved. Summary of the Invention

[0005] Therefore, it is necessary to provide a layout method for public charging stations for electric bicycles to address the aforementioned technical problems. This method can optimize the layout of public charging stations for electric bicycles.

[0006] The present invention adopts the following technical solution:

[0007] This invention provides a layout method for public charging stations for electric bicycles, comprising:

[0008] The trained basic site selection planning models were comprehensively evaluated using a test raster dataset to obtain the comprehensive evaluation results for each model. Each raster dataset includes a different number of interest points, which include name, category, and address.

[0009] The test grid dataset is input into the hybrid site selection planning model to obtain the test results for each test grid. Discrete feature values ​​are extracted from the test results and used as optional decision thresholds for the hybrid site selection planning model. The hybrid site selection planning model is comprehensively evaluated based on the optional decision thresholds and the test results to obtain the comprehensive evaluation results of the hybrid site selection planning model under different optional decision thresholds. The hybrid site selection planning model is obtained by integrating multiple basic site selection planning models.

[0010] By comparing the comprehensive evaluation results of multiple basic site selection planning models and hybrid site selection planning models under different optional decision thresholds, the optimal decision threshold for the hybrid site selection planning model is determined.

[0011] The area to be planned is rasterized to obtain raster data;

[0012] The grid data of the area to be planned is input into the hybrid site selection planning model to obtain the preliminary planning results of each grid in the area to be planned. Based on the preliminary planning results and the optimal decision threshold, the preliminary screening results of the area to be planned are determined.

[0013] Based on considerations of rider convenience and safety, the initial screening results were further filtered to obtain the second screening results for the area to be planned. The second screening results were then used to determine the layout area for public charging stations for electric bicycles within the area to be planned.

[0014] Preferably, in the calculation results of the hybrid site selection planning model, values ​​greater than or equal to the optimal decision threshold include low-confidence positive classes, medium-confidence positive classes, medium-high-confidence positive classes, and high-confidence positive classes; the preliminary screening results include setting the grid as no charging station construction, unsuitable for charging station construction, suitable for charging station construction, more suitable for charging station construction, and most suitable for charging station construction, and labeling the corresponding graphics; based on the preliminary planning results and the optimal decision threshold, the preliminary screening results are determined, specifically including:

[0015] Grids whose values ​​in the calculation results are less than or equal to the optimal decision threshold are set to not have charging stations built;

[0016] Grids whose values ​​in the calculation results are greater than the optimal decision threshold and less than or equal to the low confidence positive class are set as unsuitable for building charging stations and marked as the first setting graph;

[0017] Grids with values ​​greater than the low-confidence positive class and less than or equal to the medium-confidence positive class in the calculation results are set as suitable for building charging stations and labeled as the second setting graph;

[0018] Grids with values ​​greater than the medium confidence positive class and less than or equal to the medium-high confidence positive class in the calculation results are set as more suitable for building charging stations and are marked as the third setting graph;

[0019] Grids with values ​​greater than the medium-high confidence positive class and less than or equal to the high confidence positive class in the calculation results are set as the most suitable for building charging stations and are marked as the fourth setting graph.

[0020] Preferably, convenience considerations involve selecting pre-selected sites along highways, while safety considerations exclude pre-selected sites containing expressways, first-class highways, and national highways. Based on rider convenience and safety considerations, the initial screening results are further filtered to obtain a second screening result for the area to be planned, specifically including:

[0021] Based on the preliminary screening results, a pre-selected site area was chosen along the highway to obtain the first pre-selected site area;

[0022] For the first pre-selected site area, pre-selected site areas containing expressways, first-class highways and national highways are excluded to obtain the second pre-selected site area;

[0023] For the second pre-selection area, pre-selection areas containing intersections are excluded, resulting in the second screening result.

[0024] Preferably, the process of constructing the hybrid site selection planning model specifically includes:

[0025] Basic site selection planning models are constructed using decision trees, random forests, backpropagation neural networks, and long short-term memory networks, respectively.

[0026] For each basic site selection planning model, the basic site selection planning model is trained using a training grid dataset;

[0027] Calculate the evaluation metrics for multiple basic site selection planning models; the evaluation metrics include accuracy, precision, recall, and F1 score.

[0028] For each basic site selection planning model, the ratio of the accuracy of the basic site selection planning model to the sum of the accuracies of multiple basic site selection planning models is determined as the weight of the basic site selection planning model when constructing the hybrid site selection planning model.

[0029] Based on the weights of the basic site selection planning model, a hybrid site selection planning model is constructed.

[0030] Preferably, the weights of the basic site selection planning model are calculated as follows:

[0031] ;

[0032] in, For the first n The weights of basic site selection planning models, For the first n The accuracy of basic site selection planning models.

[0033] Preferably, the optimal decision threshold for the hybrid site selection planning model is determined by comparing the comprehensive evaluation results of multiple basic site selection planning models and hybrid site selection planning models under different optional decision thresholds, specifically including:

[0034] When the overall evaluation result of the hybrid site selection planning model is greater than the overall evaluation result of each basic site selection planning model, if there are multiple optional decision thresholds, the optional decision threshold corresponding to the largest overall evaluation result of the hybrid site selection planning model shall be determined as the optimal decision threshold of the hybrid site selection planning model.

[0035] When the overall evaluation result of the hybrid site selection planning model is greater than the overall evaluation result of each basic site selection planning model, if there is a unique optional decision threshold, then the unique optional decision threshold shall be determined as the optimal decision threshold of the hybrid site selection planning model.

[0036] Preferably, the calculation result for each grid cell in the area to be planned is as follows:

[0037] ;

[0038] in, N The calculation results are for each grid cell in the area to be planned. For the first n The weights of a basic site selection planning model, For the first n The basic site selection planning model calculates the results for each grid cell in the area to be planned.

[0039] Preferably, the calculation method for the comprehensive evaluation result of the hybrid site selection planning model is as follows:

[0040] ;

[0041] in, The optional decision threshold is At that time, the comprehensive evaluation results of the hybrid site selection planning model, For the first The weight of each evaluation indicator, Give optional decision thresholds to the mixture model The next The values ​​of each evaluation indicator;

[0042] The calculation method for the comprehensive evaluation result of the basic site selection planning model is as follows:

[0043] ;

[0044] in, For the first The comprehensive evaluation results of the basic site selection planning models, For the first The weight of each evaluation indicator, For the first The first basic site selection planning model The values ​​are selected under each evaluation indicator.

[0045] This invention provides a layout device for a public charging station for electric bicycles, comprising:

[0046] The first evaluation module is used to comprehensively evaluate multiple trained basic site selection planning models using a test grid dataset, and obtain the comprehensive evaluation result for each basic site selection planning model; each grid dataset includes a different number of interest points; interest points include name, category and address;

[0047] The second evaluation module is used to input the test grid dataset into the hybrid site selection planning model, obtain the test results for each test grid, extract the discretized feature values ​​from the test results as optional decision thresholds for the hybrid site selection planning model, and perform a comprehensive evaluation of the hybrid site selection planning model based on the optional decision thresholds and the test results to obtain the comprehensive evaluation results of the hybrid site selection planning model under different optional decision thresholds; the hybrid site selection planning model is obtained by integrating multiple basic site selection planning models;

[0048] The first determination module is used to determine the optimal decision threshold of the hybrid site selection planning model by comparing the comprehensive evaluation results of multiple basic site selection planning models with the hybrid site selection planning model under different optional decision thresholds.

[0049] The rasterization module is used to rasterize the area to be planned, obtaining raster data.

[0050] The second determination module is used to input the grid data of the area to be planned into the hybrid site selection planning model, obtain the preliminary planning results of each grid in the area to be planned, and determine the preliminary screening results of the area to be planned based on the preliminary planning results and the optimal decision threshold.

[0051] The third determination module is used to filter the preliminary screening results based on rider convenience and safety considerations, obtain the second screening results of the area to be planned, and determine the layout area of ​​public charging stations for electric bicycles in the area to be planned.

[0052] The present invention provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the above-described layout method for a public charging station for electric bicycles.

[0053] The present invention provides a computer device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it implements the above-described layout method for a public charging station for electric bicycles.

[0054] The above-mentioned at least one technical solution adopted in this invention can achieve the following beneficial effects:

[0055] Evaluation metrics were collected for various basic site selection planning models and hybrid site selection planning models under different optional decision thresholds. The comprehensive performance of each basic site selection planning model and the hybrid site selection planning model under different optional decision thresholds was evaluated and compared. If the comprehensive evaluation result of the hybrid site selection planning model was better than that of each basic site selection planning model, the optional decision threshold corresponding to the highest comprehensive evaluation result of the hybrid site selection planning model was determined as the optimal decision threshold. Using this optimal decision threshold can improve the reliability of electric bicycle charging station site selection when determining the preliminary screening results of the area to be planned based on the hybrid site selection planning model.

[0056] The grid data of the area to be planned is input into the hybrid site selection planning model to obtain preliminary planning results for each grid in the area. Based on the preliminary planning results and the optimal decision threshold, the preliminary screening results of the area to be planned are determined. Screening based on the preliminary planning results and the optimal decision threshold not only significantly improves the reliability of electric bicycle charging station site selection but also provides decision-makers with more flexible and diverse choices. Based on the convenience and safety of electric bicycles, the preliminary screening results are further screened to obtain the second screening results for the area to be planned. This not only optimizes the site layout of electric bicycle charging stations, improves user experience, and reduces riding safety risks, but also provides a scientific basis for urban planning.

[0057] This method can optimize the distribution of electric bicycle charging stations, provide scientific, efficient and diverse decision support for actual site selection, and improve the rationality and operability of site selection. Attached Figure Description

[0058] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this invention, illustrate exemplary embodiments of the invention and are used to explain the invention, but do not constitute an undue limitation of the invention. In the drawings:

[0059] Figure 1 A schematic diagram illustrating the layout method of a public charging station for electric bicycles provided by the present invention;

[0060] Figure 2 This is a schematic diagram of the training process of the hybrid site selection planning model provided by the present invention;

[0061] Figure 3 A schematic diagram illustrating the preliminary screening results of the hybrid site selection planning model provided by this invention;

[0062] Figure 4 This invention provides a road network information map of the area to be planned.

[0063] Figure 5 This invention provides a schematic diagram of expressways, first-class highways, and national highways within a planned area.

[0064] Figure 6 A schematic diagram of a large intersection within a planned area provided by the present invention;

[0065] Figure 7 A schematic diagram of the second screening results of the hybrid site selection planning model provided by the present invention;

[0066] Figure 8 A flowchart illustrating the layout method of a public charging station for electric bicycles provided by the present invention;

[0067] Figure 9 A schematic diagram of the layout device of a public charging station for electric bicycles provided by the present invention;

[0068] Figure 10 A schematic diagram of a computer device for implementing a layout method for public charging stations for electric bicycles, provided by the present invention. Detailed Implementation

[0069] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below in conjunction with specific embodiments and corresponding drawings. Obviously, the described embodiments are only a part of the embodiments of this invention, and not all of them. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0070] In existing technologies, some studies are guided by the charging demand of electric bicycles, using machine learning and big data to establish and solve mathematical models to optimize the site selection of electric bicycle charging stations. For example, researchers, based on battery swapping data from a large domestic battery swapping company, Point of Interest (POI) data, and some local food delivery demand data, used linear regression analysis and K-means clustering to study the relationship between battery swapping order demand and the density of surrounding POIs, and proposed a site selection strategy for electric bicycle battery swapping stations. Researchers also drew on relevant achievements in electric vehicle charging station layout, using POI clustering to determine the start and end points of food delivery, and simulating rider delivery routes to predict the spatiotemporal distribution of electric bicycle battery swapping demand, constructing a multi-objective site selection and capacity determination model for battery swapping stations that minimizes the operator's total cost and maximizes user satisfaction.

[0071] However, existing site selection planning for electric bicycle charging stations and battery swapping stations typically focuses on "alleviating charging demand" in a simple and isolated manner, employing relatively simplistic research methods. The site selection results neglect both the on-site environment and lack detailed and diverse recommendations, making it difficult to provide effective decision support for charging station site selection. This invention discloses a hybrid site selection planning model for electric bicycle charging stations (EBCS) based on the facility distribution in a certain urban area. By combining the usage characteristics of electric bicycles and multi-dimensional information such as road levels, it provides site selection professionals with diverse options while achieving a rational layout of charging stations within the planning area.

[0072] Devices such as desktop computers, servers, and laptops can implement the present invention. For ease of explanation, the following description will focus on servers as the implementing entity.

[0073] The technical solutions provided by the various embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0074] Figure 1 This is a schematic diagram of a layout method for a public charging station for electric bicycles according to the present invention, which specifically includes the following steps:

[0075] S101: The multiple trained basic site selection planning models are comprehensively evaluated using a test raster dataset to obtain the comprehensive evaluation result of each basic site selection planning model; each raster dataset includes a different number of interest points; interest points include name, category, and address.

[0076] The site selection planning of EBCS is influenced and constrained by the distribution of various surrounding facilities. POI data records key information about each important facility and is suitable for environmental feature analysis. The POI data used in this invention all originate from the Navigation Open Platform. The point of interest data includes name, category, and address. Categories include major and minor categories, and addresses include longitude, latitude, and detailed address. This invention uses rasterized POI data as raster data, with each raster containing a different number of points of interest.

[0077] Four basic site selection planning models were constructed using Decision Tree (DT), Random Forest (RF), BP neural network (BP), and Long Short-Term Memory (LSTM) networks, respectively; the four basic planning models were trained using a training grid dataset.

[0078] Specifically, in this invention, the basic site selection planning model includes DT, RF, BP, and LSTM. The test grid dataset is input into DT, RF, BP, and LSTM respectively to obtain the test results of the DT, RF, BP, and LSTM models. Based on the test results and various evaluation indicators, DT, RF, BP, and LSTM are comprehensively evaluated. The calculation method for the comprehensive evaluation result of the basic site selection planning model is shown in formula (1):

[0079] (1);

[0080] in, It is the first The comprehensive evaluation results of the basic site selection planning models, These are the weights of different evaluation indicators. The values ​​are for different evaluation indicators.

[0081] S102: Input the test grid dataset into the hybrid site selection planning model to obtain the test results of each test grid. Extract the discretized feature values ​​from the test results as optional decision thresholds for the hybrid site selection planning model. Based on the optional decision thresholds and the test results, comprehensively evaluate the hybrid site selection planning model to obtain the comprehensive evaluation results of the hybrid site selection planning model under different optional decision thresholds. The hybrid site selection planning model is obtained by integrating multiple basic site selection planning models.

[0082] Specifically, firstly, each basic site selection planning model in the hybrid site selection planning model is trained using a training grid dataset to obtain multiple trained basic site selection planning models, and the weights of the corresponding basic site selection planning models are determined based on the evaluation metrics of the basic site selection planning models; then, a hybrid site selection planning model is constructed based on the multiple trained basic site selection planning models and their weights; next, the test grid dataset is input into the hybrid site selection planning model to obtain the test results of the hybrid site selection planning model, and the discretized feature values ​​in the test results are extracted as optional decision thresholds; finally, based on the optional decision thresholds and the test results of the hybrid site selection planning model, the comprehensive evaluation results of the hybrid site selection planning model under different optional decision thresholds are determined.

[0083] In an exemplary embodiment, the process of constructing a hybrid site selection planning model specifically includes: constructing basic site selection planning models using decision trees, random forests, backpropagation neural networks, and long short-term memory networks, respectively; training each basic site selection planning model using a training grid dataset; calculating evaluation metrics for multiple basic site selection planning models, including accuracy, precision, recall, and F1 score; determining the weight of each basic site selection planning model as the ratio of its accuracy to the sum of the accuracies of multiple basic site selection planning models when constructing the hybrid site selection planning model; and constructing the hybrid site selection planning model by weighted summation and normalization of the calculation results of the four basic site selection planning models based on their weights.

[0084] Specifically, such as Figure 2 The diagram illustrates the training process of the hybrid site selection planning model. First, the data is preprocessed, dividing the preprocessed raster data into training and testing raster datasets. Second, multiple basic site selection planning models are trained. Then, the performance of these basic models is evaluated. Based on the evaluation results, the weights of each basic model are calculated, and weighted summation and normalization are performed to construct the hybrid site selection planning model. Next, the testing raster dataset is input into the hybrid site selection planning model, and optional decision thresholds are determined based on the discretized feature values ​​in the test results. The comprehensive evaluation results of each basic site selection planning model and the hybrid site selection planning model under different optional decision thresholds are calculated. If the comprehensive evaluation result of the hybrid model at a certain decision threshold is better than all basic models, the optional decision threshold corresponding to the highest comprehensive evaluation result of the hybrid site selection planning model is determined as the optimal decision threshold. If no optimal threshold is found after comprehensive evaluation of the hybrid site selection planning model, the hybrid site selection planning model is retrained.

[0085] Specifically, the evaluation metrics for the four basic site selection planning models are calculated. Common binary evaluation metrics include: accuracy. ), precision ), recall F1 score ) and AUC (Area Under Curve, ) etc., the present invention selects , , and As an evaluation metric, accuracy The calculation method is shown in formula (2):

[0086] (2);

[0087] in, For accuracy, This is a genuine positive sample. For true negative samples, This refers to samples that were actually negative but were output as positive. These are samples that are actually positive but output as negative.

[0088] Specifically, accuracy The calculation method is shown in formula (3):

[0089] (3);

[0090] in, For accuracy, This is a genuine positive sample. This refers to a sample that was actually negative but was output as positive.

[0091] Specifically, recall rate The calculation method is shown in formula (4):

[0092] (4);

[0093] in, For recall rate, This is a genuine positive sample. These are samples that are actually positive but output as negative.

[0094] Specifically, The calculation method is shown in formula (5):

[0095] (5);

[0096] in, for Fraction, For accuracy, This refers to the recall rate.

[0097] Specifically, the accuracy of each basic site selection planning model is divided by the sum of the accuracies of the four models, and this is used as the weight of that model. The calculation method for the weight of the basic site selection planning model is shown in formula (6):

[0098] (6);

[0099] in, For the first n The weights of basic site selection planning models, For the first n The accuracy of basic site selection planning models.

[0100] Specifically, the range of the discretized feature value k is all the test results of the test grid. A certain discretized feature value k in the test results of the hybrid site selection planning model is used as an optional decision threshold to adjust the side policy conditions of the hybrid site selection planning model, thereby determining the overall performance of the model. The side policy condition is to set the grid to the critical value of not building a charging station. When the calculation result of the grid obtained by the hybrid site selection planning model is less than the critical value, the grid is set to not build a charging station. This invention introduces evaluation indicators such as Acc and Prec, and uses a weighted summation and normalization method to determine the comprehensive evaluation result of the hybrid site selection planning model. The calculation method of the comprehensive evaluation result of the hybrid site selection planning model is shown in formula (7):

[0101] (7);

[0102] in, The optional decision threshold is The comprehensive evaluation results of the hybrid site selection planning model These are the weights of different evaluation indicators. The hybrid model provides a given decision threshold. The values ​​of different evaluation indicators.

[0103] S103: By comparing the comprehensive evaluation results of multiple basic site selection planning models and hybrid site selection planning models under different optional decision thresholds, determine the optimal decision threshold for the hybrid site selection planning model.

[0104] In an exemplary embodiment, the optimal decision threshold for the hybrid site selection planning model is determined by comparing the comprehensive evaluation results of multiple basic site selection planning models with those of a hybrid site selection planning model under different optional decision thresholds. Specifically, this includes: when the comprehensive evaluation result of the hybrid site selection planning model is greater than the comprehensive evaluation result of each basic site selection planning model, if there are multiple optional decision thresholds, the optional decision threshold corresponding to the largest comprehensive evaluation result of the hybrid site selection planning model is determined as the optimal decision threshold for the hybrid site selection planning model; when the comprehensive evaluation result of the hybrid site selection planning model is greater than the comprehensive evaluation result of each basic site selection planning model, if there is a unique optional decision threshold, the unique optional decision threshold is determined as the optimal decision threshold for the hybrid site selection planning model.

[0105] Specifically, for example, if the overall evaluation result of the hybrid site selection planning model is better than that of the four basic site selection planning models, and its performance is the best relative to other thresholds (0.5, 0.7, 0.8, 0.9, 1), then the decision threshold is determined to be 0.6.

[0106] S104: Rasterize the area to be planned to obtain raster data.

[0107] In this invention, the area to be planned is rasterized to obtain raster data. Electric bicycle charging stations are then laid out on the raster grids selected by this invention for which charging stations should be located. The POI data used in this invention comes from a navigation open platform, and the data content includes name, category, subcategory, longitude, latitude, and detailed address.

[0108] S105: Input the grid data of the area to be planned into the hybrid site selection planning model to obtain the preliminary planning results of each grid in the area to be planned. Based on the preliminary planning results and the optimal decision threshold, determine the preliminary screening results of the area to be planned.

[0109] In an exemplary embodiment, the calculation result for each grid cell in the area to be planned is shown in formula (8):

[0110] (8);

[0111] in, N The calculation results are for each grid cell in the area to be planned. For the first n The weights of a basic site selection planning model, For the first n The basic site selection planning model calculates the results for each grid cell in the area to be planned.

[0112] Specifically, based on the calculation results and weights of the nth basic site selection planning model, the calculation results of each grid in the area to be planned are determined after weighted summation and normalization.

[0113] In an exemplary embodiment, in the calculation results of the hybrid site selection planning model, values ​​greater than or equal to the optimal decision threshold include low-confidence positive classes, medium-confidence positive classes, medium-high-confidence positive classes, and high-confidence positive classes; the preliminary screening results include setting grids as "no charging station construction," "unsuitable for charging station construction," "suitable for charging station construction," "more suitable for charging station construction," and "most suitable for charging station construction," and labeling the corresponding graphics; based on the preliminary planning results and the optimal decision threshold, the preliminary screening results are determined, specifically including: setting grids with values ​​less than or equal to the optimal decision threshold in the calculation results as "no charging station construction," and setting grids with values ​​greater than or equal to the optimal decision threshold in the calculation results as "no charging station construction." Grids with a decision threshold less than or equal to the low-confidence positive class are designated as unsuitable for charging station construction and marked as the first setting graph. Grids with a calculation result greater than the low-confidence positive class and less than or equal to the medium-confidence positive class are designated as suitable for charging station construction and marked as the second setting graph. Grids with a calculation result greater than the medium-confidence positive class and less than or equal to the medium-high-confidence positive class are designated as more suitable for charging station construction and marked as the third setting graph. Grids with a calculation result greater than the medium-high-confidence positive class and less than or equal to the high-confidence positive class are designated as the most suitable for charging station construction and marked as the fourth setting graph.

[0114] Specifically, based on the comparison between the calculation results of each grid within the area to be planned and the optimal decision threshold, low-confidence positive class, medium-confidence positive class, high-confidence positive class, and high-confidence positive class, the grid is determined to be labeled as most suitable for building a charging station, more suitable for building a charging station, suitable for building a charging station, unsuitable for building a charging station, or not suitable for building a charging station. In this invention, the grid is labeled with different graphics according to different labeling types. This invention does not limit this; for example, the grid can also be labeled with different colors according to different labeling types. The preliminary screening results of the area to be planned are determined based on the labeling results.

[0115] In this invention, the first shape is a triangle, the second shape is a rhombus, the third shape is a circle, and the fourth shape is a square. Figure 3 The image shows the preliminary screening results of the hybrid site selection planning model. Grids marked with squares are the most suitable for building charging stations, grids marked with circles are more suitable, grids marked with diamonds are suitable for building charging stations, and grids marked with triangles are not suitable for building charging stations.

[0116] S106: Based on considerations of rider convenience and safety, the initial screening results are further filtered to obtain the second screening results for the area to be planned, and the second screening results are determined as the layout area for public charging stations for electric bicycles in the area to be planned.

[0117] In one exemplary embodiment, convenience considerations involve selecting pre-selected sites along highways, while safety considerations involve excluding pre-selected sites containing expressways, first-class highways, and national highways. Based on considerations of convenience and safety for electric bicycle riders, the initial screening results are further filtered to obtain a second screening result for the area to be planned. Specifically, this includes: selecting pre-selected sites along highways based on the initial screening results to obtain a first pre-selected site; excluding pre-selected sites containing expressways, first-class highways, and national highways from the first pre-selected site to obtain a second pre-selected site; and excluding pre-selected sites containing intersections from the second pre-selected site to obtain the second screening result. This second screening result is then used to determine the layout area for public charging stations for electric bicycles within the area to be planned.

[0118] Specifically, EBCS is a charging station for electric bicycles and also a distribution center for a large number of electric bicycles, so convenience and safety must be the top priorities.

[0119] like Figure 4 As shown, this is a road network information map of the area to be planned provided by the present invention. The location of the EBCS is set on both sides of the road to obtain the first pre-selected area, which not only increases the frequency of use of EBCS, but also makes it more convenient for electric bicycle riders to use.

[0120] like Figure 5 As shown, the proposed plan area includes expressways, first-class highways, and national highways. Research results indicate a clear relationship between collision speed and injury severity for electric bicycle riders. When the collision speed between a vehicle and an electric bicycle is 30 km / h, the risk of death for the electric bicycle rider is 2.9%; at 50 km / h, the risk is 23%; at 60 km / h, it is 50%; and at 80 km / h, it is 90%. According to highway design specifications, the design speed for first-class highways used for distribution is 80 km / h, while the design speed for second-class highways is 60 km / h. When limited by terrain and geological conditions, the speeds are adjusted to 60 km / h and 40 km / h respectively. Therefore, EBCSs located near second-class and lower-level highways are safer. National highways, as national trunk highways, connect important political and economic centers and transportation hubs throughout the country. The area has a large transport capacity and a large number of vehicles passing through it. The areas on both sides of the national highway and the surrounding areas are not suitable for planning and construction of EBCS. Based on the first pre-selected site area, the EBCS pre-selected site areas containing expressways, first-class highways and national highways are excluded, resulting in the second pre-selected site area.

[0121] like Figure 6As shown, this represents a large intersection within the planned area. Large intersections, as convergence points for various vehicles, represent one of the most complex road conditions. Intersections formed by secondary or higher-level highways, due to their higher design speeds and greater mix of vehicles, are excluded from the second pre-selection area due to grid areas containing large intersections. This results in the second screening result, as shown below. Figure 7 As shown, squares indicate the most suitable locations for building charging stations, circles indicate more suitable locations, rhombuses indicate suitable locations, and triangles indicate unsuitable locations. The results of the second screening are determined as the appropriate locations for public charging stations for electric bicycles in the area to be planned.

[0122] In one exemplary embodiment, the present invention provides as follows Figure 8 The flowchart shown is a layout method for public charging stations for electric bicycles. First, point of interest (POI) data is obtained from a geographic information system and rasterized to obtain POI raster data. Second, the POI raster data is processed through a hybrid site selection planning model to obtain preliminary screening results for the planning area. Then, the preliminary screening results are processed through multi-dimensional information such as the usage characteristics of electric bicycles and road levels to obtain a second screening result.

[0123] When applying the layout method for a public charging station for electric bicycles provided by this invention, it is not necessary to consider... Figure 1 The steps shown are executed in sequence. The specific execution order of each step can be determined as needed, and this invention does not impose any restrictions on it.

[0124] The above describes a layout method for a public charging station for electric bicycles, provided by one or more embodiments of the present invention. Based on the same idea, the present invention also provides a corresponding layout device for a public charging station for electric bicycles, such as... Figure 9 As shown.

[0125] Figure 9 A schematic diagram of the layout device for a public charging station for electric bicycles provided by the present invention includes:

[0126] The first evaluation module 901 is used to comprehensively evaluate multiple trained basic site selection planning models using a test raster dataset, and obtain a comprehensive evaluation result for each basic site selection planning model. Each raster dataset includes a different number of points of interest. Points of interest include name, category, subcategory, longitude, latitude, and detailed address.

[0127] The second evaluation module 902 is used to input the test grid dataset into the hybrid site selection planning model and obtain the test results for each test grid. Discretized feature values ​​are extracted from the results and used as optional decision thresholds for the hybrid site selection planning model. Based on the optional decision thresholds and the test results, the hybrid site selection planning model is comprehensively evaluated to obtain comprehensive evaluation results for the hybrid site selection planning model under different optional decision thresholds; the hybrid site selection planning model is obtained by integrating multiple basic site selection planning models.

[0128] The first determining module 903 is used to determine the optimal decision threshold of the hybrid site selection planning model by comparing the comprehensive evaluation results of multiple basic site selection planning models with the hybrid site selection planning model under different optional decision thresholds.

[0129] The rasterization module 904 is used to rasterize the area to be planned to obtain raster data.

[0130] The second determining module 905 is used to input the grid data of the area to be planned into the hybrid site selection planning model to obtain the preliminary planning results of each grid in the area to be planned, and to determine the preliminary screening results of the area to be planned based on the preliminary planning results and the optimal decision threshold.

[0131] The third determining module 906 is used to filter the preliminary screening results based on rider convenience and safety considerations, obtain the second screening results of the area to be planned, and determine the layout area of ​​public charging stations for electric bicycles in the area to be planned.

[0132] Specific limitations regarding the layout device of a public charging station for electric bicycles can be found in the above description of the layout method for a public charging station for electric bicycles, and will not be repeated here. Each module in the aforementioned layout device for a public charging station for electric bicycles can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device in hardware form, or stored in the memory of a computer device in software form, so that the processor can call and execute the operations corresponding to each module.

[0133] The present invention also provides a computer-readable storage medium storing a computer program that can be used to execute the above-described... Figure 1 A layout method for public charging stations for electric bicycles is provided.

[0134] The present invention also provides Figure 10 The schematic diagram of the computer device shown is as follows: Figure 10As shown, at the hardware level, this computer device includes a processor, internal bus, network interface, memory, and non-volatile memory, and may also include other hardware required for business operations. The processor reads the corresponding computer program from the non-volatile memory into memory and then executes it to achieve the above. Figure 1 A layout method for public charging stations for electric bicycles is provided.

[0135] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments of the methods described above. Any references to memory, storage, databases, or other media used in the embodiments provided by this invention can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, or optical storage, etc. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.

[0136] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this invention.

Claims

1. A layout method for public charging stations for electric bicycles, characterized in that, include: The trained basic site selection planning models were comprehensively evaluated using a test raster dataset to obtain the comprehensive evaluation results for each model. Each raster dataset includes a different number of interest points, which include name, category, and address. The test grid dataset is input into the hybrid site selection planning model to obtain the test results for each test grid. Discrete feature values ​​are extracted from the test results and used as optional decision thresholds for the hybrid site selection planning model. The hybrid site selection planning model is comprehensively evaluated based on the optional decision thresholds and the test results to obtain the comprehensive evaluation results of the hybrid site selection planning model under different optional decision thresholds. The hybrid site selection planning model is obtained by integrating multiple basic site selection planning models. By comparing the comprehensive evaluation results of multiple basic site selection planning models and hybrid site selection planning models under different optional decision thresholds, the optimal decision threshold for the hybrid site selection planning model is determined. The area to be planned is rasterized to obtain raster data; The grid data of the area to be planned is input into the hybrid site selection planning model to obtain the calculation results of each grid in the area to be planned. Based on the calculation results and the optimal decision threshold, the preliminary screening results of the area to be planned are determined. Based on considerations of rider convenience and safety, the initial screening results were further filtered to obtain the second screening results for the area to be planned. The second screening results were then used to determine the layout area for public charging stations for electric bicycles in the area to be planned. The grid data of the area to be planned is input into the hybrid site selection planning model to obtain the calculation results for each grid in the area. Based on the calculation results and the optimal decision threshold, the preliminary screening results of the area to be planned are determined. Specifically, this includes: in the calculation results of the hybrid site selection planning model, values ​​greater than or equal to the optimal decision threshold include low confidence positive class, medium confidence positive class, medium-high confidence positive class, and high confidence positive class; the preliminary screening results include setting the grids as not building charging stations, unsuitable for building charging stations, suitable for building charging stations, more suitable for building charging stations, and most suitable for building charging stations, and labeling the corresponding graphics; based on the calculation results and the optimal decision threshold, the preliminary screening results are determined, specifically including: values ​​in the calculation results less than or equal to the optimal decision threshold... The grid with the optimal decision threshold is set as "no charging station to be built". Grids with values ​​greater than the optimal decision threshold and less than or equal to the low-confidence positive class are set as "unsuitable for building charging stations" and labeled as the first setting graph. Grids with values ​​greater than the low-confidence positive class and less than or equal to the medium-confidence positive class are set as "suitable for building charging stations" and labeled as the second setting graph. Grids with values ​​greater than the medium-confidence positive class and less than or equal to the medium-high-confidence positive class are set as "more suitable for building charging stations" and labeled as the third setting graph. Grids with values ​​greater than the medium-high-confidence positive class and less than or equal to the high-confidence positive class are set as "most suitable for building charging stations" and labeled as the fourth setting graph.

2. The method as described in claim 1, characterized in that, Convenience considerations led to the selection of pre-selected sites along highways, while safety considerations excluded pre-selected sites containing expressways, first-class highways, and national highways. Based on rider convenience and safety considerations, the initial screening results were further refined to obtain a second screening result for the areas to be planned, specifically including: Based on the preliminary screening results, a pre-selected site area was chosen along the highway to obtain the first pre-selected site area; For the first pre-selected site area, pre-selected site areas containing expressways, first-class highways and national highways are excluded to obtain the second pre-selected site area; For the second pre-selection area, pre-selection areas containing intersections are excluded, resulting in the second screening result.

3. The method as described in claim 1, characterized in that, The process of constructing a hybrid site selection planning model specifically includes: Basic site selection planning models are constructed using decision trees, random forests, backpropagation neural networks, and long short-term memory networks, respectively. For each basic site selection planning model, the basic site selection planning model is trained using a training grid dataset; Calculate the evaluation metrics for multiple basic site selection planning models; the evaluation metrics include accuracy, precision, recall, and F1 score. For each basic site selection planning model, the ratio of the accuracy of the basic site selection planning model to the sum of the accuracies of multiple basic site selection planning models is determined as the weight of the basic site selection planning model when constructing the hybrid site selection planning model. Based on the weights of the basic site selection planning models, the calculation results of the four basic site selection planning models are weighted, summed, and normalized to construct a hybrid site selection planning model.

4. The method as described in claim 3, characterized in that, The weights of the basic site selection planning model are calculated as follows: ； in, For the first n The weights of basic site selection planning models, For the first n The accuracy of basic site selection planning models.

5. The method as described in claim 1, characterized in that, By comparing the comprehensive evaluation results of multiple basic site selection planning models and hybrid site selection planning models under different optional decision thresholds, the optimal decision threshold for the hybrid site selection planning model is determined, specifically including: When the overall evaluation result of the hybrid site selection planning model is greater than the overall evaluation result of each basic site selection planning model, if there are multiple optional decision thresholds, the optional decision threshold corresponding to the largest overall evaluation result of the hybrid site selection planning model shall be determined as the optimal decision threshold of the hybrid site selection planning model. When the overall evaluation result of the hybrid site selection planning model is greater than the overall evaluation result of each basic site selection planning model, if there is a unique optional decision threshold, then the unique optional decision threshold shall be determined as the optimal decision threshold of the hybrid site selection planning model.

6. The method as described in claim 4, characterized in that, The calculation results for each grid cell in the area to be planned are as follows: ； in, N The calculation results are for each grid cell in the area to be planned. For the first n The weights of basic site selection planning models, For the first n The basic site selection planning model calculates the results for each grid cell in the area to be planned.

7. The method as described in claim 4, characterized in that, The calculation method for the comprehensive evaluation result of the hybrid site selection planning model is as follows: ； in, The optional decision threshold is At that time, the comprehensive evaluation results of the hybrid site selection planning model, For the hybrid site selection planning model i The weight of each evaluation indicator, To provide an optional decision threshold for the hybrid site selection planning model The later i The values ​​of each evaluation indicator; The calculation method for the comprehensive evaluation result of the basic site selection planning model is as follows: ； in, For the first n The comprehensive evaluation results of the basic site selection planning models, For the first i The weight of each evaluation indicator, For the first n The first basic site selection planning model i The values ​​are selected under each evaluation indicator.

Images