Urban updating area demarcation method based on DBSCAN-GM algorithm

Abstract

The invention provides a town updating area demarcation method based on a DBSCAN-GM algorithm. The town updating area demarcation method based on the DBSCAN-GM algorithm comprises the steps of performing preliminary clustering on data by using a Gaussian-Means algorithm, determining a clustering center, calculating key parameters EPS and MinPts of the DBSCAN algorithm based on the clustering center, and performing further clustering analysis by using the DBSCAN-GM algorithm. The method not only simplifies the parameter selection process, but also improves the stability and accuracy of the clustering result, improves the parameter selection and clustering stability by combining the advantages of the DBSCAN and Gaussian-Means algorithms, and improves the precision and reliability of town updating area delimitation, so that the advantages of big data and an intelligent algorithm can be fully utilized, and the method is suitable for large-scale popularization and application. The defects of a traditional method in the aspects of data processing and analysis are overcome, and the updating area is accurately and reasonably delimited.

Description

Technical Field

[0001] The present invention belongs to the technical field of urban and rural planning and management, and in particular relates to a method for delineating urban renewal areas based on a DBSCAN-GM algorithm. Background Art

[0002] Urban renewal is an important means to achieve sustainable urban development and improve urban functions and environmental quality. The focus of urban development has gradually shifted from large-scale incremental construction to stock renewal and structural adjustment. However, traditional urban renewal planning methods face many challenges in practical applications. Traditional urban renewal planning methods mainly rely on manual surveys and expert experience. This method has the following major problems: first, the efficiency of data collection and processing is low, and it is difficult to reflect the dynamic changes of the city in a timely manner; second, it is highly subjective and easily affected by the experience and knowledge level of planners, resulting in inconsistent planning results, lack of scientificity and objectivity; third, traditional methods are difficult to handle large-scale, multi-dimensional urban data, and cannot comprehensively and accurately evaluate the actual needs and effects of urban renewal. These problems make it difficult for traditional urban renewal planning methods to cope with the complex and changing development needs of modern cities.

[0003] In order to overcome the shortcomings of traditional urban renewal area delineation methods, in the context of today's technological changes, the application of intelligent technology has enabled the delineation of urban renewal areas to be transformed from the traditional method that relies on ground surveys and expert experience to a more dynamic, accurate and efficient process. Among them, the DBSCAN (Density-Based Spatial Clustering of Applications with Noise) algorithm is an important tool for urban renewal area delineation due to its superior performance in processing noise data and identifying clusters of arbitrary shapes. However, there are still some problems in relying solely on the DBSCAN algorithm for urban renewal area delineation. The DBSCAN algorithm is sensitive to parameter settings, especially the selection of parameters EPS (neighborhood radius) and MinPts (minimum number of points) directly affects the clustering results. Improper parameter selection will lead to unstable clustering results and affect the accurate delineation of the renewal area. In addition, when processing high-dimensional data, the DBSCAN algorithm has a high computational complexity, which easily leads to low computational efficiency. Summary of the invention

[0004] The present invention discloses a method for delineating urban renewal areas based on the DBSCAN-GM algorithm, which can fully utilize the advantages of big data and intelligent algorithms, solve the shortcomings of traditional methods in data processing and analysis, and accurately and reasonably delineate renewal areas.

[0005] The present invention discloses a method for delineating urban renewal areas based on a DBSCAN-GM algorithm, the method comprising the following steps:

[0006] Step 1: Collect urban renewal related vector data and public opinion survey results in the target study area, pre-process and standardize the collected data, construct extended fields corresponding to land class patches, and establish an urban renewal database;

[0007] Step 2: Extract features and reduce dimension of data in the urban renewal database to obtain comprehensive indicators; identify urban renewal object data points based on the comprehensive indicators, and evaluate the identification results;

[0008] Step 3: Based on the data points of urban renewal objects, the Gaussian-Means algorithm is used to perform cluster analysis on the data points, automatically dividing the data into several clusters, and determining a center point for each cluster;

[0009] Step 4: Calculate the local EPS value of each cluster, and select the minimum value among all cluster local EPS values ​​as the global EPS value. At the same time, calculate the MinPts parameter based on the cluster characteristics, and select the minimum value among all cluster local MinPts parameters as the global MinPts parameter.

[0010] Step 5: Based on the global EPS value and the global MinPts parameter, run the DBSCAN algorithm to perform the final cluster analysis to identify potential urban renewal areas;

[0011] Step 6: Distinguish potential key urban renewal areas from potential urban renewal spatial units, and repeat steps 3 to 5 within the potential key urban renewal areas to identify potential urban renewal spatial units within the potential key urban renewal areas;

[0012] Step 7: Based on the vector data collected in step 1 and the results of the public opinion survey, the regional boundaries of potential urban renewal areas, including potential urban renewal key areas and potential urban renewal spatial units, are corrected to form an urban renewal area delineation plan.

[0013] Step 1 further comprises:

[0014] Step 1.1: Collect land data, population data, building data, industrial economic data, public service facility data, infrastructure data, administrative boundary data, planning boundary data, and rigid conductive boundary data of the target study area through the planning management department, and supplement the data with field exploration;

[0015] Step 1.2: pre-process the vector data collected in step 1.1, including cleaning, sorting and format conversion, and then standardize the pre-processed data to normalize it to the [0,1] interval; the standardization formula is as follows:

[0016]

[0017] Among them, x ij is the value of the i-th sample on the j-th index, min(x j ) and max(x j ) are the minimum and maximum values ​​of the j-th index respectively;

[0018] Step 1.3: Use the standardized data obtained in step 1.2 to construct extended fields corresponding to land type maps and establish a town renewal database.

[0019] As one of the preferred options, in step 1.1, the land data include the surface vector data of floor area ratio, building density, land use compliance, and land attributes; the population data include the population employment data of employed population density and the surface vector data of population density, per capita residential building area, and residential unit type; the building data include the surface vector data of the year of construction, number of building floors, building quality, and building structure; the industrial economic data include the surface vector data of economic output, business office, commercial service, and industrial matching degree; the public service facility data include the surface vector data of scientific research and innovation, culture and education, sports and fitness, medical care, social welfare, parks and green spaces, and squares; the infrastructure data include the surface vector data of transportation facilities and public utilities; the administrative boundary data include the administrative boundaries of cities, counties / districts, and towns / streets; the rigid conductive boundary data include the surface vector data of the areas overlapping the historical protection, ecological protection or other special planning and construction areas designated in the national land space master plan and relevant statutory plans; all surface vector data are based on plots.

[0020] Step 2 further includes:

[0021] Step 2.1: Directly include land that is facing development strategy adjustment, land with health and safety risks, land with weak supporting facilities that cannot be improved from the outside, land that does not meet the requirements of urban development, and land that urgently needs historical and cultural protection or historical style improvement into the urban renewal objects;

[0022] Step 2.2: Use the entropy weight method to calculate the weight of each indicator and evaluate the importance of each indicator in urban renewal. The formula for calculating the weight of each indicator is as follows:

[0023]

[0024] Among them, p ij It represents the weight of the i-th sample on the j-th indicator, and m is the total number of samples;

[0025] The formula for calculating the entropy value of each indicator is as follows:

[0026]

[0027] in, is a constant used to ensure that the entropy value ranges between [0,1];

[0028] The formula for calculating the weight of each indicator is as follows:

[0029]

[0030] Among them, w j represents the weight of the jth indicator, and n is the total number of indicators;

[0031] Step 2.3: Based on the indicator weights obtained in step 2.2, a weighted covariance matrix is ​​constructed, and the weights determined by the entropy weight method are reduced in dimension using the PCA method to extract the eigenvectors and reduce the dimension of the data. Specifically, the following sub-steps are included:

[0032] Construct the weighted covariance matrix:

[0033]

[0034] Among them, x i is the standardized data vector of the ith sample, is the sample mean vector, m is the total number of samples;

[0035] Compute the eigenvalues ​​and eigenvectors of the covariance matrix:

[0036] Cv=λv

[0037] Among them, λ is the eigenvalue and v is the corresponding eigenvector;

[0038] Select the first k largest eigenvalues ​​and their corresponding eigenvectors to form the eigenvector matrix V k ;

[0039] Using the eigenvector matrix V k Project the original data into a new lower-dimensional space:

[0040] Z=XV k

[0041] Among them, Z is the data matrix after dimension reduction, and X is the original data matrix after standardization;

[0042] Step 2.4: Multiply the principal components extracted by PCA by their corresponding weights to obtain the weighted principal components of each indicator, which are used as the final comprehensive indicator weight Y:

[0043] Y=ZW

[0044] Among them, Y is the final comprehensive indicator weight, Z is the data matrix after dimensionality reduction, and W is the weight vector calculated by the entropy weight method;

[0045] Step 2.5: Use the comprehensive index weight Y to calculate the data of each plot area to obtain the weighted comprehensive index score of each plot; based on the weighted comprehensive index score of each plot, the natural breakpoint analysis method in the geographic information system is used, combined with the expert scoring, and a comprehensive judgment is made to establish a threshold A, and the plot areas exceeding the threshold A are included in the urban renewal objects, and the plot areas below the threshold A are not included in the urban renewal objects;

[0046] Step 3 further includes:

[0047] Step 3.1, based on the surface data of the town renewal object identified in step 2, establish an envelope rectangle, and generate the geometric center point of the envelope rectangle as the data point of the town renewal object;

[0048] Step 3.2: Run the Gaussian-Means algorithm on the data points of the urban renewal objects to divide the data into the optimal number of clusters C k , each cluster has a certain center point M i , i=1…K.

[0049] Step 4 further includes:

[0050] Step 4.1: For each cluster, calculate the average distance from the cluster center to all points in the cluster as the local EPS value of the cluster:

[0051]

[0052] Among them, r j is the local EPS value of the jth cluster, N j is the number of points in the jth cluster, M j is the center point of the jth cluster, x ij is the i-th point in the j-th cluster, distance 2 (M j ,x ij ) is the center point M j To point x ij The Euclidean distance of

[0053] Step 4.2: Select the minimum value among all the local EPS values ​​of the clusters as the global EPS value:

[0054] EPS global =min(r1,r2,…,r k )

[0055] Among them, EPS global is the global EPS value, k is the total number of clusters, r1,r2,…,r k is the local EPS value of each cluster;

[0056] Step 4.3. For each cluster, calculate the MinPts parameter:

[0057]

[0058] Among them, MinPts j is the MinPts parameter of the jth cluster, r j is the local EPS value of the jth cluster, N j is the number of points in the jth cluster, TotalVolume j is the total area of ​​the jth cluster;

[0059] Step 4.4, select the minimum value among the local MinPts values ​​of all clusters as the global MinPts value:

[0060] MinPts global =min(MinPts1,MinPts2,…,MinPts k )

[0061] Among them, MinPts global is the global MinPts value, k is the total number of clusters, MinPts1, MinPts2, …, MinPts k is the local MinPts value of each cluster.

[0062] Step 5 further includes:

[0063] Step 5.1, using the EPS value and MinPts parameter calculated in step 4, execute the DBSCAN algorithm;

[0064] Step 5.2: Generate a cluster label for each data point based on the output of the DBSCAN algorithm; mark the data points whose number of neighbors is greater than or equal to the MinPts parameter as core points, mark the data points whose number of neighbors is less than the MinPts parameter but within the neighborhood of a core point as boundary points, and mark the data points that do not belong to any cluster as noise points;

[0065] Step 5.3, evaluate the clustering effect by calculating the silhouette coefficient index;

[0066] Step 5.4: By visualizing the core points and their neighborhood radius, the sets of several clusters and the ranges of their domain radius are superimposed. The superimposed area is used as the identified high-density area of ​​urban renewal objects and is defined as a potential urban renewal area.

[0067] Step 6 further comprises:

[0068] Step 6.1. Use GIS tools to calculate the area of ​​each potential urban renewal area, classify the areas, and update the classification results into the GIS database; for cities at prefecture level and above, mark areas with an area of ​​more than 1 square kilometer as potential urban renewal key areas; mark areas with an area of ​​less than or equal to 1 square kilometer as potential urban renewal spatial units; and directly mark counties, county-level cities, and townships as potential urban renewal spatial units;

[0069] Step 6.2: Repeat steps 3 to 5 within the potential urban renewal focus area to identify potential urban renewal spatial units within the potential urban renewal focus area.

[0070] Step 7 further includes:

[0071] Step 7.1, integrate the vector data collected in step 1 into a unified geographic information system;

[0072] Step 7.2: In the geographic information system, overlap the boundaries of the secondary planning zones of the national land space master plan, the administrative boundaries of the township level and above, and the boundaries of the potential key urban renewal areas, and use the spatial segmentation tool to spatially correct the boundaries of the potential key urban renewal areas;

[0073] Step 7.3: Overlap the boundaries of detailed planning units, grassroots mass autonomous organizations and potential urban renewal space units, and use the spatial segmentation tool to spatially correct the boundaries of potential urban renewal space units; overlap and correct the boundaries of potential urban renewal areas and historical and cultural block protection areas in historical and cultural cities;

[0074] Step 7.4: Organize public participation activities through questionnaire surveys, public meetings and online platforms to present preliminary boundary adjustment results to residents and stakeholders, collect their feedback and opinions and analyze them, identify residents’ and stakeholders’ concerns and suggestions on regional boundary adjustments, and make further adjustments to the boundaries of key urban renewal areas and urban renewal spatial units;

[0075] Step 7.5: Update the final revised regional boundaries into the GIS database to form a town renewal area GIS data package.

[0076] The beneficial effects of the present invention are:

[0077] By combining the advantages of DBSCAN and Gaussian-Means algorithms, the parameter selection and clustering stability issues are improved, and the accuracy and reliability of urban renewal area delineation are improved. Specifically, when performing data clustering, the DBSCAN-GM algorithm first uses the Gaussian-Means algorithm to perform preliminary clustering of the data, determines the cluster center, and then calculates the key parameters EPS and MinPts of the DBSCAN algorithm based on the cluster center for further cluster analysis. This method not only simplifies the parameter selection process, but also improves the stability and accuracy of the clustering results. The urban renewal area delineation method based on the DBSCAN-GM algorithm of the present invention can make full use of the advantages of big data and intelligent algorithms, solve the shortcomings of traditional methods in data processing and analysis, and accurately and reasonably delineate renewal areas. BRIEF DESCRIPTION OF THE DRAWINGS

[0078] Figure 1 It is a flow chart of the urban renewal area delineation method based on DBSCAN-GM algorithm of the present invention;

[0079] Figure 2 It is a schematic diagram of the execution steps of the urban renewal area delineation method based on the DBSCAN-GM algorithm of the present invention. DETAILED DESCRIPTION

[0080] The following examples will enable those skilled in the art to more fully understand the present invention, but are not intended to limit the present invention in any way.

[0081] The present invention discloses a method for delineating urban renewal areas based on a DBSCAN-GM algorithm, the method comprising the following steps:

[0082] Step 1: Collect urban renewal related vector data and public opinion survey results in the target study area, pre-process and standardize the collected data, construct extended fields corresponding to land class patches, and establish an urban renewal database;

[0083] Step 2: Extract features and reduce dimension of data in the urban renewal database to obtain comprehensive indicators; identify urban renewal object data points based on the comprehensive indicators, and evaluate the identification results;

[0084] Step 3: Based on the data points of urban renewal objects, the Gaussian-Means algorithm is used to perform cluster analysis on the data points, automatically dividing the data into several clusters, and determining a center point for each cluster;

[0085] Step 4: Calculate the local EPS value of each cluster, and select the minimum value among all cluster local EPS values ​​as the global EPS value. At the same time, calculate the MinPts parameter based on the cluster characteristics, and select the minimum value among all cluster local MinPts parameters as the global MinPts parameter.

[0086] Step 5: Based on the global EPS value and the global MinPts parameter, run the DBSCAN algorithm to perform the final cluster analysis to identify potential urban renewal areas;

[0087] Step 6: Distinguish potential key urban renewal areas from potential urban renewal spatial units, and repeat steps 3 to 5 within the potential key urban renewal areas to identify potential urban renewal spatial units within the potential key urban renewal areas;

[0088] Step 7: Based on the vector data collected in step 1 and the results of the public opinion survey, the regional boundaries of potential urban renewal areas, including potential urban renewal key areas and potential urban renewal spatial units, are corrected to form an urban renewal area delineation plan.

[0089] See also Figure 1 and Figure 2 The urban renewal area delineation method of the present invention specifically comprises the following steps:

[0090] S1. First, data collection and urban renewal database construction are carried out.

[0091] Data was collected by searching various information departments through computer networks. Data that could not be collected were obtained through manual questionnaire surveys and python3.12 big data analysis. Then, the data of each indicator were standardized through the entropy weight standardization formula and normalized to the [0,1] interval. The urban renewal database was established by uniformly importing the geographic information platform and constructing extended fields corresponding to the land class maps.

[0092] S2. Then, urban renewal objects are identified based on comprehensive indicators and the identification results are evaluated.

[0093] The entropy weight method is used to calculate the proportion, entropy value and weight of each indicator, evaluate the importance of each indicator in urban renewal, and further construct a weighted covariance matrix based on the indicator weight. The PCA method is used to reduce the dimension of the weight determined by the entropy weight method, extract the eigenvector, and reduce the dimension of the data.

[0094] Multiply the principal components extracted by PCA with their corresponding weights to obtain the weighted principal components of each indicator, which is used as the final comprehensive indicator Y. Based on this and the classification of urban renewal objects, the corresponding types of urban renewal objects are identified and the urban renewal threshold A is evaluated to obtain the urban renewal object surface data.

[0095] S3. Perform cluster analysis based on urban renewal object data points.

[0096] The Gaussian-Means algorithm is used to perform preliminary cluster analysis on the data points, determine the cluster center point, and perform a Gaussian distribution hypothesis test to ultimately optimize the location of the cluster center point.

[0097] S4. Calculate and select the key parameters EPS and MinPts for the DBSCAN algorithm.

[0098] The local EPS value of each cluster is calculated, and the minimum value among the local EPS values ​​of all clusters is selected as the global EPS value. At the same time, the MinPts parameter is calculated based on the cluster characteristics.

[0099] S5. Run the DBSCAN algorithm and mark the clustering results.

[0100] According to the calculated EPS and MinPts parameters, the DBSCAN algorithm is run to perform cluster analysis, and the core points, boundary points, and noise points are marked to evaluate the clustering effect.

[0101] S6. Use GIS tools to classify regions and repeat cluster analysis in key areas.

[0102] The area of ​​each potential urban renewal area was calculated and classified into key areas and spatial units, and the cluster analysis was repeated within the key areas to further identify renewal spatial units.

[0103] S7. Integrate data to make boundary corrections and make further adjustments through public participation.

[0104] The collected vector data were integrated into the GIS system for spatial segmentation and boundary correction, and public feedback was collected through questionnaires, public meetings and online platforms to further adjust regional boundaries.

[0105] Examples

[0106] The technical solution of the present invention will be described in detail below by taking the urban renewal area delineation of an ancient city based on the DBSCAN-GM algorithm as an example.

[0107] Step 1: Collect data from the target study area through the planning management department and standardize it, construct extended fields corresponding to the land type patches, and establish a town renewal database, including:

[0108] Step 1.1 Collect data from various information departments through computer network searches. Data that cannot be collected are obtained through manual questionnaire surveys and python3.12 big data analysis. The land data, population data, building data, industrial economic data, public service facility data, infrastructure data, administrative boundary data, planning boundary data, and rigid transmission boundary of a certain ancient city are imported into the geographic information system.

[0109] Land data include surface vector data of floor area ratio, building density, land use compliance, land attributes, etc. Population data include population employment data of employed population density and surface vector data of population density, per capita residential building area, residential unit type, etc.

[0110] Architectural data include surface vector data of construction year, number of floors, building quality, building structure, etc. Industrial economic data include surface vector data of economic output, business office, commercial service, industrial matching degree, etc.

[0111] Public service facilities data include surface vector data of scientific research and innovation, culture and education, sports and fitness, medical care and health, social welfare, parks and green spaces, squares, etc.

[0112] Infrastructure data includes surface vector data of transportation facilities, public facilities, ecological and environmental impacts, etc.

[0113] Administrative boundary data include township-level administrative boundaries and grassroots mass autonomous organizations (community) boundaries.

[0114] Planning boundary data include ecological protection red line, permanent basic farmland, and urban development boundary vector data.

[0115] The rigidly transmitted boundaries include the regional surface vector data that overlaps with the historical protection, ecological protection or other special planning and construction areas designated in the national land space master plan and relevant statutory plans.

[0116] The collected vector data will be cleaned, organized and formatted to ensure data accuracy and consistency.

[0117] Step 1.2: Use the entropy weight method to standardize the data of each indicator to the interval [0,1]. The standardization formula is as follows:

[0118]

[0119] Among them, x ij is the value of the i-th sample on the j-th index, min(x j ) and max(x j ) are the minimum and maximum values ​​of the j-th indicator respectively.

[0120] Step 1.3: Import the obtained statistical data, vector information, interview data and other information into the geographic information platform of the CGS-WGS-1984 coordinate system, construct extended fields corresponding to land class maps, and establish a town renewal database. The spatial elements in the database meet the requirements of TD / T1057-2020 and are included in the national land space basic data platform.

[0121] Step 2: Use the direct screening method to select urban renewal objects. For the remaining areas, use the entropy weight method to calculate the weights of each indicator and construct a weighted covariance matrix. Use the PCA method to reduce the dimension of the weights determined by the entropy weight method, extract the feature vector, and reduce the dimension of the data to obtain a comprehensive indicator. Identify urban renewal objects based on comprehensive indicators and evaluate the identification results. Specifically include:

[0122] Step 2.1: Directly include land that is facing development strategy adjustment, land with health and safety hazards, land with weak supporting facilities that cannot be improved from the outside, land whose use does not meet the requirements of urban development, and land with urgent need for historical and cultural protection or historical style improvement into the urban renewal objects.

[0123] Step 2.2: Use the entropy weight method to calculate the weight of each indicator and evaluate the importance of each indicator in urban renewal. The formula for calculating the weight of each indicator is as follows:

[0124]

[0125] Among them, p ij It represents the weight of the i-th sample on the j-th indicator, and m is the total number of samples.

[0126] The formula for calculating the entropy value of each indicator is as follows:

[0127]

[0128] in, is a constant that ensures that the entropy value is in the range [0,1].

[0129] The formula for calculating the weight of each indicator is as follows:

[0130]

[0131] Among them, w j represents the weight of the jth indicator, and n is the total number of indicators.

[0132] Step 2.3: Construct a weighted covariance matrix based on the indicator weights obtained in step 2.2, use the PCA method to reduce the dimension of the weights determined by the entropy weight method, extract the eigenvectors, and reduce the dimension of the data.

[0133] Construct the weighted covariance matrix:

[0134]

[0135] Among them, x i is the standardized data vector of the ith sample, is the sample mean vector, and m is the total number of samples.

[0136] Compute the eigenvalues ​​and eigenvectors of the covariance matrix by solving the eigenvalue decomposition problem:

[0137] Cv=λv

[0138] Among them, λ is the eigenvalue and v is the corresponding eigenvector.

[0139] Select principal components: Select the first k largest eigenvalues ​​and their corresponding eigenvectors to form the eigenvector matrix V k .

[0140] Data dimensionality reduction: Using the eigenvector matrix V k Project the original data into a new lower-dimensional space:

[0141] Z=XV k

[0142] Among them, Z is the data matrix after dimensionality reduction, and X is the original data matrix after standardization.

[0143] Step 2.4: Multiply the principal components extracted by PCA by their corresponding weights to obtain the weighted principal components of each indicator, which is used as the final comprehensive indicator. The formula is as follows:

[0144] Y=ZW

[0145] Among them, Y is the final comprehensive indicator weight, Z is the data matrix after dimensionality reduction, and W is the weight vector calculated by the entropy weight method.

[0146] Step 2.5, use the comprehensive index weight Y to calculate the data of each area and obtain the weighted comprehensive index score of each area. Based on the score, the natural breakpoint analysis method in the geographic information system (GIS) is used, combined with expert judgment, to establish a judgment threshold A. Areas exceeding this threshold are included in the urban renewal objects, and areas below this threshold are not included in the urban renewal objects.

[0147] Step 2.6: Evaluate the recognition results, including the evaluation of indicators such as accuracy and stability, to verify the effectiveness and reliability of the weight recognition system.

[0148] Step 3: Determine the final number of clusters and centers through preliminary clustering, Gaussian distribution hypothesis test, description length criterion calculation and optimization of cluster centers.

[0149] Step 3.1: Based on the surface data of a certain ancient city urban renewal object identified in step 2, an envelope rectangle is established, and the geometric center point of the envelope rectangle is generated for subsequent analysis based on such data points.

[0150] Step 3.2: Perform a preliminary cluster analysis using the Gaussian-Means (G-Means) algorithm to determine the optimal number and structure of clusters: Input the data set, run the Gaussian-Means algorithm, and divide the data into the optimal number of clusters C. k , each cluster has a certain center point M i (where i=1...K), the obtained clusters and center points will be used to determine the parameters EPS and Minpts of the DBSCAN algorithm.

[0151] Step 4: Calculate and select the key parameters EPS and MinPts for the DBSCAN algorithm.

[0152] 4.1 For each cluster, calculate the average distance from the cluster center to all points in the cluster as the local EPS value of the cluster:

[0153]

[0154] Step 4.2: Select the minimum value among all the local EPS values ​​of the clusters as the global EPS value:

[0155] EPS global =min(r1,r2,…,r k )

[0156] Step 4.3, for each cluster, calculate the MinPts parameter:

[0157]

[0158] Step 4.4, select the minimum value among the local MinPts values ​​of all clusters as the global MinPts value:

[0159] MinPts global =min(MinPts1,MinPts2,…,MinPts k )

[0160] Step 5: Run the DBSCAN algorithm using the calculated EPS and MinPts parameters, mark the clustering results, and evaluate the clustering effect through visualization.

[0161] Step 5.1: Use the EPS value and MinPts parameter calculated in step 4 to run the DBSCAN algorithm for cluster analysis.

[0162] Step 5.2: According to the output results of the DBSCAN algorithm, the data points are marked as follows: core points, boundary points, and noise points, and a cluster label is generated for each data point to identify which cluster the data point belongs to, or to identify it as a noise point.

[0163] Step 5.3: Evaluate the clustering effect by calculating the silhouette coefficient. The silhouette coefficient can reflect the compactness and separation of clusters and ensure the quality of clustering results.

[0164] Step 5.4: By visualizing the core points and their neighborhood radius, the sets of several clusters and the range of their domain radius are superimposed. These areas are the high-density areas of urban renewal objects in a certain ancient city identified by the DBSCAN algorithm, that is, potential urban renewal areas.

[0165] Step 6: Use GIS tools to calculate and classify the areas of potential urban renewal areas, and repeat cluster analysis within key areas to identify renewal spatial units.

[0166] Step 6.1: Use GIS tools to calculate the area of ​​each potential urban renewal area and classify the areas:

[0167] Areas with an area larger than 1 square kilometer are marked as potential key areas for urban renewal;

[0168] Areas with an area less than or equal to 1 square kilometer are marked as potential urban renewal spatial units.

[0169] Step 6.2: Repeat steps 3 to 5 within the potential urban renewal focus area to identify potential urban renewal spatial units within the potential urban renewal focus area.

[0170] Step 7: Integrate vector data for spatial segmentation and boundary correction, and further adjust the urban renewal area boundaries of an ancient city through public participation.

[0171] Step 7.1: Integrate the vector data collected in step 1, including administrative boundary data, planning boundary data, rigid conductive boundary data, etc., into a unified geographic information system (GIS).

[0172] Step 7.2: In the geographic information system, overlap the boundaries of the secondary planning zones of the national land space master plan, the administrative boundaries at the township level and above, and the boundaries of potential key urban renewal areas, and use the spatial segmentation tool to make spatial corrections to the boundaries of potential key urban renewal areas.

[0173] Step 7.3: Overlap the boundaries of the detailed planning units, the boundaries of the grassroots mass autonomous organizations (communities) and the boundaries of the potential urban renewal space units, and use the spatial segmentation tool to spatially correct the boundaries of the potential urban renewal space units.

[0174] Since a certain ancient city is a historical and cultural city, it is also necessary to overlap and correct the boundaries of the potential urban renewal area and the historical and cultural blocks.

[0175] Step 7.4: Organize public participation activities through questionnaires, public meetings and online platforms to present preliminary boundary adjustment results to residents and stakeholders, collect their feedback and opinions and analyze them. Further adjust the boundaries of key urban renewal areas and urban renewal space units by identifying the main concerns and suggestions of residents and stakeholders on regional boundary adjustments.

[0176] The above are only preferred embodiments of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions under the concept of the present invention belong to the protection scope of the present invention. It should be pointed out that for ordinary technicians in this technical field, some improvements and modifications without departing from the principle of the present invention should be regarded as the protection scope of the present invention.

Claims

1. A method for delineating urban renewal areas based on the DBSCAN-GM algorithm, characterized in that: The method comprises the following steps: Step 1: Collect urban renewal related vector data and public opinion survey results in the target study area, pre-process and standardize the collected data, construct extended fields corresponding to land class patches, and establish an urban renewal database; Step 2: Extract features and reduce dimension of data in the urban renewal database to obtain comprehensive indicators; identify urban renewal object data points based on the comprehensive indicators, and evaluate the identification results; Step 3: Based on the data points of urban renewal objects, the Gaussian-Means algorithm is used to perform cluster analysis on the data points, automatically dividing the data into several clusters, and determining a center point for each cluster; Step 4: Calculate the local EPS value of each cluster, and select the minimum value among all cluster local EPS values ​​as the global EPS value. At the same time, calculate the MinPts parameter based on the cluster characteristics, and select the minimum value among all cluster local MinPts parameters as the global MinPts parameter. Step 5: Based on the global EPS value and the global MinPts parameter, run the DBSCAN algorithm to perform the final cluster analysis to identify potential urban renewal areas; Step 6: Distinguish potential key urban renewal areas from potential urban renewal spatial units, and repeat steps 3 to 5 within the potential key urban renewal areas to identify potential urban renewal spatial units within the potential key urban renewal areas; Step 7: Based on the vector data collected in step 1 and the results of the public opinion survey, the regional boundaries of potential urban renewal areas, including potential urban renewal key areas and potential urban renewal spatial units, are corrected to form an urban renewal area delineation plan.

2. The urban renewal area delineation method based on the DBSCAN-GM algorithm according to claim 1 is characterized in that: Step 1 further comprises: Step 1.1: Collect land data, population data, building data, industrial economic data, public service facility data, infrastructure data, administrative boundary data, planning boundary data, and rigid conductive boundary data of the target study area through the planning management department, and supplement the data with field exploration; Step 1.2: pre-process the vector data collected in step 1.1, including cleaning, sorting and format conversion, and then standardize the pre-processed data to normalize it to the [0,1] interval; the standardization formula is as follows: Among them, x ij is the value of the i-th sample on the j-th index, min(x j ) and max(x j ) are the minimum and maximum values ​​of the j-th index respectively; Step 1.3: Use the standardized data obtained in step 1.2 to construct extended fields corresponding to land type maps and establish a town renewal database.

3. The urban renewal area delineation method based on DBSCAN-GM algorithm according to claim 2 is characterized in that: In step 1.1, land data include surface vector data of plot ratio, building density, land use compliance, and land attributes; population data include population employment data of employed population density and surface vector data of population density, per capita residential building area, and residential unit type; building data include surface vector data of construction year, number of building floors, building quality, and building structure; industrial economic data include surface vector data of economic output, business office, commercial service, and industrial matching degree; public service facility data include surface vector data of scientific research and innovation, culture and education, sports and fitness, medical care, social welfare, parks and green spaces, and squares; infrastructure data include surface vector data of transportation facilities and public utilities; administrative boundary data include administrative boundaries of cities, counties / districts, and towns / streets; rigid conductive boundary data include surface vector data of areas overlapping with historical protection, ecological protection or other special planning and construction areas designated in the national land space master plan and relevant statutory plans; all surface vector data are based on plots.

4. The urban renewal area delineation method based on the DBSCAN-GM algorithm according to claim 1 is characterized in that: Step 2 further includes: Step 2.1: Directly include land that is facing development strategy adjustment, land with health and safety risks, land with weak supporting facilities that cannot be improved from the outside, land that does not meet the requirements of urban development, and land that urgently needs historical and cultural protection or historical style improvement into the urban renewal objects; Step 2.2: Use the entropy weight method to calculate the weight of each indicator and evaluate the importance of each indicator in urban renewal. The formula for calculating the weight of each indicator is as follows: Among them, p ij It represents the weight of the i-th sample on the j-th indicator, and m is the total number of samples; The formula for calculating the entropy value of each indicator is as follows: in, is a constant used to ensure that the entropy value ranges between [0,1]; The formula for calculating the weight of each indicator is as follows: Among them, w j represents the weight of the jth indicator, and n is the total number of indicators; Step 2.3: Based on the indicator weights obtained in step 2.2, a weighted covariance matrix is ​​constructed, and the weights determined by the entropy weight method are reduced in dimension using the PCA method to extract the eigenvectors and reduce the dimension of the data. Specifically, the following sub-steps are included: Construct the weighted covariance matrix: Among them, x i is the standardized data vector of the ith sample, is the sample mean vector, m is the total number of samples; Compute the eigenvalues ​​and eigenvectors of the covariance matrix: Cv=λv Among them, λ is the eigenvalue and v is the corresponding eigenvector; Select the first k largest eigenvalues ​​and their corresponding eigenvectors to form the eigenvector matrix V k ; Using the eigenvector matrix V k Project the original data into a new lower-dimensional space: Z=XV k Among them, Z is the data matrix after dimension reduction, and X is the original data matrix after standardization; Step 2.4: Multiply the principal components extracted by PCA by their corresponding weights to obtain the weighted principal components of each indicator, which are used as the final comprehensive indicator weight Y: Y=ZW Among them, Y is the final comprehensive indicator weight, Z is the data matrix after dimensionality reduction, and W is the weight vector calculated by the entropy weight method; Step 2.5: Use the comprehensive index weight Y to calculate the data of each plot area to obtain the weighted comprehensive index score of each plot; based on the weighted comprehensive index score of each plot, the natural breakpoint analysis method in the geographic information system is used, combined with the expert scoring, and a comprehensive judgment is made to establish a threshold A. Plot areas exceeding the threshold A are included in the urban renewal objects, and plot areas below the threshold A are not included in the urban renewal objects.

5. The urban renewal area delineation method based on DBSCAN-GM algorithm according to claim 1 is characterized in that: Step 3 further includes: Step 3.1, based on the surface data of the town renewal object identified in step 2, establish an envelope rectangle, and generate the geometric center point of the envelope rectangle as the data point of the town renewal object; Step 3.2: Run the Gaussian-Means algorithm on the data points of the urban renewal objects to divide the data into the optimal number of clusters C k , each cluster has a certain center point M i , i=1…K.

6. The urban renewal area delineation method based on DBSCAN-GM algorithm according to claim 1 is characterized in that: Step 4 further includes: Step 4.1: For each cluster, calculate the average distance from the cluster center to all points in the cluster as the local EPS value of the cluster: Among them, r j is the local EPS value of the jth cluster, N j is the number of points in the jth cluster, M j is the center point of the jth cluster, x ij is the i-th point in the j-th cluster, distance 2 (M j ,x ij ) is the center point M j To point x ij The Euclidean distance of Step 4.2: Select the minimum value among all the local EPS values ​​of the clusters as the global EPS value: EPS global =min(r1,r2,…,r k ) Among them, EPS global is the global EPS value, k is the total number of clusters, r1,r2,…,r k is the local EPS value of each cluster; Step 4.

3. For each cluster, calculate the MinPts parameter: Among them, MinPts j is the MinPts parameter of the jth cluster, r j is the local EPS value of the jth cluster, N j is the number of points in the jth cluster, TotalVolume j is the total area of ​​the jth cluster; Step 4.4, select the minimum value among the local MinPts values ​​of all clusters as the global MinPts value: MinPts global =min(MinPts1,MinPts2,…,MinPts k ) Among them, MinPts global is the global MinPts value, k is the total number of clusters, MinPts1, MinPts2, …, MinPts k is the local MinPts value of each cluster.

7. The urban renewal area delineation method based on DBSCAN-GM algorithm according to claim 1 is characterized in that: Step 5 further includes: Step 5.1, using the EPS value and MinPts parameter calculated in step 4, execute the DBSCAN algorithm; Step 5.2: Generate a cluster label for each data point based on the output of the DBSCAN algorithm; mark the data points whose number of neighbors is greater than or equal to the MinPts parameter as core points, mark the data points whose number of neighbors is less than the MinPts parameter but within the neighborhood of a core point as boundary points, and mark the data points that do not belong to any cluster as noise points; Step 5.3, evaluate the clustering effect by calculating the silhouette coefficient index; Step 5.4: By visualizing the core points and their neighborhood radius, the sets of several clusters and the ranges of their domain radius are superimposed. The superimposed area is used as the identified high-density area of ​​urban renewal objects and is defined as a potential urban renewal area.

8. The urban renewal area delineation method based on DBSCAN-GM algorithm according to claim 1 is characterized in that: Step 6 further comprises: Step 6.

1. Use GIS tools to calculate the area of ​​each potential urban renewal area, classify the areas, and update the classification results into the GIS database; for cities at prefecture level and above, mark areas with an area of ​​more than 1 square kilometer as potential urban renewal key areas; mark areas with an area of ​​less than or equal to 1 square kilometer as potential urban renewal spatial units; and directly mark counties, county-level cities, and townships as potential urban renewal spatial units; Step 6.2: Repeat steps 3 to 5 within the potential urban renewal focus area to identify potential urban renewal spatial units within the potential urban renewal focus area.

9. The urban renewal area delineation method based on DBSCAN-GM algorithm according to claim 1 is characterized in that: Step 7 further includes: Step 7.1, integrate the vector data collected in step 1 into a unified geographic information system; Step 7.2: In the geographic information system, overlap the boundaries of the secondary planning zones of the national land space master plan, the administrative boundaries of the township level and above, and the boundaries of the potential key urban renewal areas, and use the spatial segmentation tool to spatially correct the boundaries of the potential key urban renewal areas; Step 7.3: Overlap the boundaries of detailed planning units, grassroots mass autonomous organizations and potential urban renewal space units, and use the spatial segmentation tool to spatially correct the boundaries of potential urban renewal space units; overlap and correct the boundaries of potential urban renewal areas and historical and cultural block protection areas in historical and cultural cities; Step 7.4: Organize public participation activities through questionnaire surveys, public meetings and online platforms to present preliminary boundary adjustment results to residents and stakeholders, collect their feedback and opinions and analyze them, identify residents’ and stakeholders’ concerns and suggestions on regional boundary adjustments, and make further adjustments to the boundaries of key urban renewal areas and urban renewal spatial units; Step 7.5: Update the final revised regional boundaries into the GIS database to form a town renewal area GIS data package.

Images