Urban updating decision-making method based on public participation

Abstract

The invention relates to the field of city updating, and discloses a city updating decision-making method based on public participation, which comprises the following steps: acquiring city space basic data including a vector map, remote sensing image data and historical, cultural and geographic information data of a city; on the basis of the acquired spatial data, a public participation base map is constructed, and the base map comprises information of a road network, land use types and historical and cultural areas of the city; collecting opinion data of the public on facility site selection through an interactive platform on the public participation base map; and according to the collected public opinion data, constructing a multi-subject scoring model. According to the method, a decision-making method based on public participation is adopted, space basic data and public opinions of cities are combined, facility site selection is evaluated through a multi-subject scoring model, and the opinions of different subjects can be integrated by the system by setting weight values of the public, experts and the government, so that the diversity and fairness of site selection schemes are ensured.

Description

Technical Field

[0001] The present invention relates to the field of urban renewal, and specifically to an urban renewal decision-making method based on public participation. Background Art

[0002] With the continuous acceleration of the urbanization process, urban renewal has become a key means to improve urban functions and environmental quality. Reasonable facility location and land planning are of great significance for improving the living quality and social welfare of cities. However, traditional urban planning decisions are usually dominated by experts and the government, lacking extensive public participation, resulting in planning schemes not meeting the needs of residents and easily triggering social contradictions. Therefore, how to better incorporate public opinions into the decision-making process through reasonable technical means has become an important issue in urban renewal.

[0003] Current urban renewal methods to a certain extent adopt spatial data analysis technology and combine basic indicators such as traffic and land use for facility location. Some systems use geographic information system (GIS) platforms to display the urban spatial structure to support decision-making. At the same time, some technical solutions collect public opinions through expert reviews and social surveys to assist in site selection decisions. In addition, with the application of remote sensing technology and convolutional neural networks, some urban renewal schemes can also accurately identify land use types to provide basic data for subsequent planning.

[0004] However, existing technologies often have deficiencies. First, the traditional decision-making process still mainly relies on the experience of experts and the dominance of the government, with low public participation, resulting in it being difficult for the scheme to reflect the real needs of residents. Second, existing spatial data analysis methods mostly rely on manual modeling and static data and are difficult to reflect urban changes in real time, resulting in poor timeliness of planning decisions. Third, although existing suitability analysis methods consider basic indicators such as traffic accessibility, they lack effective integration of social needs and public feedback and are difficult to provide multi-dimensional support for decision-making. For this reason, technical personnel in this field have proposed an urban renewal decision-making method based on public participation to solve the above problems. Summary of the Invention

[0005] Aiming at the deficiencies of the existing technology, the present invention provides an urban renewal decision-making method based on public participation, which solves the problems of insufficient public participation, lagging data update, and lack of multi-dimensional comprehensive consideration in facility location decision-making in the existing technology.

[0006] To achieve the above objectives, the present invention is realized through the following technical solutions: An urban renewal decision-making method based on public participation, including the following steps:

[0007] Obtain basic urban spatial data, including vector maps of the city, remote sensing image data, and historical and cultural geographical information data;

[0008] Based on the acquired spatial data, a public participation base map is constructed, and the base map includes information on the road network, land use types, and historical and cultural regions of the city;

[0009] On the public participation base map, opinion data of the public on facility siting is collected through an interactive platform, and the opinion data includes facility types, siting coordinates, and importance scores;

[0010] According to the collected public opinion data, a multi-agent scoring model is constructed, and by setting the weight values of the public, experts, and the government, the candidate facility sitings are scored;

[0011] Suitability analysis is carried out on the candidate facility sitings, and the suitability analysis includes spatial indicators such as traffic accessibility, service coverage, and land developability, and a suitability score is obtained;

[0012] Weighted calculation is performed on the obtained comprehensive scoring value and the suitability scoring value, and the final optimal siting is selected;

[0013] The optimal siting result is visually displayed through a Geographic Information System (GIS) platform and feedback is given to the public for further confirmation.

[0014] Preferably, the construction of the public participation base map includes:

[0015] Using the space syntax method to model the urban road network and extract the spatial integration degree and spatial choice degree of road nodes;

[0016] Based on the integration degree and choice degree, the urban roads are screened to construct an urban skeleton network;

[0017] Based on a convolutional neural network model, land use type recognition is performed on remote sensing images to generate a spatial layer with semantic labels;

[0018] The historical block and cultural landmark data are superimposed on the map through buffer analysis to mark the boundary area of the urban context.

[0019] Preferably, the calculation of the spatial integration degree is based on the following rules:

[0020] The integration degree of any node is equal to the function value of its average path depth to all other nodes, where the average path depth is the average minimum path length from this node to all nodes in the system.

[0021] Preferably, the convolutional neural network model includes:

[0022] The input layer receives remote sensing image block data, and the image contains three color channels;

[0023] The convolutional layer and the pooling layer extract spatial texture features and output an intermediate feature map;

[0024] The fully connected layer maps the feature map to the classification labels of land use types, and the output is a predefined set of land classes;

[0025] The loss function uses the cross-entropy function for the supervised learning process.

[0026] Preferably, the collection of public opinions includes:

[0027] The public selects the proposed facility type through the interactive map interface;

[0028] The public selects the proposed facility location coordinates on the map;

[0029] The public submits feedback information, including the facility type identifier, the selected site coordinates, and the subjective importance score;

[0030] The public feedback information is written into the database in a structured manner.

[0031] Preferably, the construction of the multi-agent scoring model includes:

[0032] Set three types of agents: the public, experts, and the government, and assign a fixed weight value to each type of agent, with the sum of the three being 1;

[0033] Each type of agent assigns an independent score to each candidate location;

[0034] The scores of each type of agent are weighted and calculated according to the preset weights, and the comprehensive score value of the candidate location is output.

[0035] Preferably, the public score is derived from the statistical result of the spatial density of public opinion data, the expert score is derived from the matching of facility planning experience rules, and the government score is calculated based on the urban strategic planning control requirements.

[0036] Preferably, the facility site selection suitability analysis includes:

[0037] Set multiple spatial suitability indicators, including traffic accessibility, service radius coverage, land development feasibility, environmental friendliness, and policy restriction factors;

[0038] Normalize the minimum and maximum values of each indicator;

[0039] Assign a weight value to each indicator and calculate the comprehensive suitability score of each candidate location.

[0040] Preferably, the calculation of the most preferred site includes:

[0041] Set a regulation parameter to control the weight balance ratio between the public score and the suitability score;

[0042] For each candidate location, calculate its comprehensive score which is equal to the complement of the public rating multiplied by the adjustment factor plus the suitability rating multiplied by the adjustment factor;

[0043] Select the location with the highest score as the final site selection result.

[0044] Preferably, the visual output of the geographic information system platform includes:

[0045] The public opinion heat map layer, which reflects the concentrated areas of public feedback;

[0046] The facility suitability score layer, which reflects the spatial distribution of the suitability degree of each location;

[0047] The most optimal site selection result layer, which marks the final proposed facility layout location.

[0048] The present invention provides a decision-making method for urban renewal based on public participation. It has the following beneficial effects:

[0049] 1. The present invention adopts a decision-making method based on public participation, combines the spatial basic data of the city with public opinions, evaluates the facility site selection through a multi-agent scoring model, and by setting the weight values of the public, experts and the government, the system can comprehensively integrate the opinions of different agents, ensuring the diversity and fairness of the site selection plan. This technical solution effectively solves the problem that traditional site selection methods rely too much on the judgment of a single agent, avoids biases in the decision-making process, and improves the social recognition of the decision-making results.

[0050] 2. The present invention models the urban space and identifies land use types through the spatial integration degree and convolutional neural network technology, and then constructs a base map for public participation. This method not only improves the accuracy of urban planning, but also enhances the transparency of the planning process. Compared with the existing base map construction methods that simply rely on manual drawing or empirical rules, the present invention has significantly improved in the automation and accuracy of data processing, reduces human factor interference, and ensures the scientificity and rationality of the planning.

[0051] 3. The present invention combines suitability analysis with public feedback, obtains a comprehensive score value through weighted calculation, and selects the optimal facility site. By quantifying and assigning weights to multiple spatial suitability indicators such as traffic accessibility, service coverage, and land developability, it realizes multi-dimensional and multi-angle site selection optimization. Compared with the traditional single-index evaluation method, the present invention provides a more comprehensive site selection plan considering multiple factors, making the site selection result more practical and feasible.

[0052] 4. The present invention uses a Geographic Information System (GIS) platform for visual display, integrating the public opinion heat map, the facility suitability scoring layer, and the optimal site selection result layer to intuitively display the spatial distribution of urban renewal decisions. This technical solution effectively improves the visibility and transparency of decision-making, facilitating all stakeholders to understand and evaluate the site selection results. Compared with the complex and hard-to-understand report forms in traditional methods, the present invention provides an intuitive and easy-to-understand decision-making support method, enhancing the enthusiasm of public participation and the openness of the decision-making process. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 It is a schematic flow chart of the method of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0054] Next, in conjunction with the accompanying drawings of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0055] Please refer to the attached Figure 1 , an urban renewal decision-making method based on public participation provided by an embodiment of the present invention includes the following steps:

[0056] S1. Obtain urban spatial basic data, including the vector map of the city, remote sensing image data, and historical and cultural geographical information data;

[0057] Specifically, in an urban renewal decision-making method based on public participation of the present invention, the acquisition and collation of urban spatial basic data need to be completed first. This step not only provides comprehensive spatial information support for the construction of the subsequent public participation base map, but also is an important data source for the training of the convolutional neural network model and the automatic recognition of land use types. To ensure the accuracy and scientific nature of subsequent spatial analysis and public opinion collection, the integrity and accuracy of the spatial basic data are crucial, directly affecting the reliability of the urban renewal decision-making plan.

[0058] In this embodiment, the urban spatial basic data obtained by the system includes the vector map of the city, remote sensing image data, and historical and cultural geographical information data. The vector map of the city usually comes from government planning departments or authoritative surveying and mapping institutions, and the data content should at least cover the urban road network, building outlines, plot boundaries, and functional zoning information. Generally, to ensure the integrity and timeliness of the spatial data, the vector map data should be georegistered using the national standard coordinate system and have a data format that can be directly imported into the GIS platform, such as the.shp or.geojson format.

[0059] As an option, the remote sensing image data can be collected from public satellite remote sensing platforms or commercial high-resolution remote sensing data service providers. Preferably, multispectral images with a resolution of not less than 0.5 meters are used to ensure that spatial details are clearly distinguishable. The remote sensing image data serves not only as the input source for the convolutional neural network model but also for identifying urban land use types and feature characteristics, providing an important basis for subsequent spatial analysis. Specifically, the remote sensing image should include at least three color channels, namely the red channel (R), the green channel (G), and the blue channel (B), for facilitating subsequent processing by the neural network model.

[0060] In a possible implementation, to strengthen the identification and analysis of historical and cultural protection areas, the system further obtains urban historical and cultural geographical information data. This data includes information such as historical block boundaries, cultural landmark locations, and cultural relic protection units. The data format can be vector data or raster data. Generally, the historical and cultural data should be recognized by the cultural competent department and marked with clear boundaries and geographical coordinates to ensure the complete expression of cultural areas in subsequent analysis.

[0061] Specifically, after the remote sensing image data is obtained, it needs to undergo preprocessing operations, including image correction, denoising, and band combination, to ensure that the image quality meets the requirements of subsequent modeling.

[0062] In addition, after the preprocessing of remote sensing data is completed, the historical and cultural geographical information data is overlaid onto the spatial data set to ensure that the influencing factors of historical and cultural protection areas can be synchronously identified and considered during the subsequent modeling process. As an option, the overlay of cultural geographical information can be completed through the GIS buffer analysis method to ensure that the plots within a reasonable range around cultural landmarks can be identified and given key consideration by the system.

[0063] S2. Based on the obtained spatial data, construct a public participation base map, where the base map includes information on the road network, land use types, and historical and cultural areas of the city;

[0064] Specifically, after the acquisition of urban spatial basic data is completed, to realize public visual participation and subsequent decision-making analysis, the present invention enters the stage of constructing the public participation base map. This step is closely connected to the aforementioned basic data processing. The core lies in integrating spatial syntax analysis, convolutional neural network recognition, and the overlay of historical and cultural information to systematically generate a public participation base map with spatial logic and attribute semantics, directly serving subsequent public opinion collection and the establishment of decision-making models.

[0065] In this embodiment, first, based on the obtained urban road vector data, the spatial syntax modeling method is used to conduct a structural analysis of the urban road network, quantify the spatial relationships between roads, and extract key spatial semantic indicators. Generally, the road network will be transformed into an axial graph model, and based on the road centerlines, a spatial network graph connected by nodes is generated.

[0066] Specifically, the system first calculates the connection value of each road, and the formula is as follows:

[0067] C i = α·K i ;

[0068] Where: C i represents the connection value of the i-th road, dimensionless; α is the correction coefficient under the special road organizational structure, dimensionless; K i represents the number of other roads directly connected to the i-th road, in units of roads.

[0069] In a possible implementation, if the road has an overpass or a layered structure, the system will automatically adjust α to reflect the true accessibility between roads, ensuring the accuracy and scientificity of the connection value calculation.

[0070] Then, the system further calculates the integration degree (IntegrationValue) based on the road connection relationship to quantify the spatial aggregation ability between the node and the overall network. The calculation of the integration degree follows the following formula:

[0071]

[0072] Where: Int i represents the integration degree of the i-th road; d ij represents the minimum topological path length from the i-th road to the j-th road, in units of step length; n is the total number of roads in the road network, in units of roads.

[0073] Generally, the higher the integration degree, the stronger the aggregation ability of the road in the overall road network, and it belongs to the core space of urban traffic and activities.

[0074] As an option, to further characterize the marginality characteristics of the road space, the system synchronously calculates the edge index (EdgeValue), and the formula is:

[0075] E i = 1 - Int i ;

[0076] Where: E i is the edge index of the i-th road; Int i is the integration degree described above.

[0077] The edge index reflects the edge degree of the road in the network. The larger the value, the more the road is in the urban fringe area or a relatively blocked area.

[0078] After completing the space syntax analysis, the system extracts the urban skeleton network based on the calculation results. Specifically, the system selects the main roads with high spatial integration and obvious traffic hub characteristics according to the thresholds of integration and choice, constructs the urban skeleton framework, and provides a structural basis for the public participation base map.

[0079] In this embodiment, the system further combines remote sensing image data and uses a convolutional neural network model (CNN) to automatically identify the land use type of plots and generate semantic labels. The remote sensing image patches are input into the CNN model, and the input data structure is:

[0080] W1×H1×D;

[0081] Where: W1 represents the width of the image, in pixels; H1 represents the height of the image, in pixels; D represents the number of channels of the image, usually 3 (RGB three channels).

[0082] The calculation formula for the size of the feature map of the convolutional layer is:

[0083]

[0084] Where: W2 represents the width of the feature map after the convolution operation, in pixels; P is the size of the convolutional kernel, in pixels; Q is the size of the edge padding, in pixels; L is the convolution stride, in pixels.

[0085] In some embodiments, the convolutional kernel uses a 3×3 size, the edge padding is set to 1, and the stride is 1 to keep the spatial information of the feature map complete. Through multiple layers of convolution and pooling operations, the model extracts the spatial texture features of land use and generates intermediate feature maps.

[0086] The feature map after convolution processing enters the fully connected layer to complete the mapping of land use type labels, and the output is a set of predefined land use types in the city, such as residential land, commercial land, green space, industrial land, etc., which is convenient for subsequent public participation and decision-making analysis. Generally, in the model training process, the cross-entropy loss function (Cross-Entropy Loss Function) is used to optimize the model weight parameters in a supervised manner to improve the classification accuracy.

[0087] As an option, to enhance the spatial expression of historical and cultural protection areas, the system introduces historical block and cultural landmark data into the public participation base map. Specifically, the system sets an influence radius around the cultural landmark based on the buffer analysis method and superimposes the buffer on the spatial base map to clearly identify the boundary of the urban context and ensure that the public and decision-makers fully consider historical and cultural factors during the participation process.

[0088] S3. Collecting the public's opinion data on the facility site selection on the public participation base map through an interactive platform, wherein the opinion data includes the facility type, site selection coordinates and importance score;

[0089] Specifically, following the construction of the aforementioned public participation base map, in the technical path of the present invention, step S3 focuses on the collection and quantitative modeling of public opinions. This step is not only the core embodiment of public participation in urban renewal, but also the data basis for realizing a participatory spatial decision-making mechanism. In order to ensure the breadth and accuracy of collected opinions, a structured interaction mechanism needs to be designed, and public feedback needs to be semantically annotated and spatially mapped, ultimately completing the conversion of public opinions into spatial analysis inputs.

[0090] In this embodiment, the collection of public opinions is completed through an interactive platform based on WebGIS. The public can select any plot of land on the participation base map automatically generated by the system to express their opinions. In general, public opinions include three types of information: land use preferences (such as the desire to increase green space, improve transportation, etc.), functional suggestions (such as the construction of public service facilities), and emotional attitudes (such as satisfaction, identity, etc.).

[0091] Specifically, the system uses a labeled questionnaire interactive interface to guide the public to input their opinions in a structured form. The input data structure is as follows:

[0092] O={(l i ,t i ,w i )|i=1,2,…,n};

[0093] Where: O represents the set of public opinions; l i Indicates the spatial location pointed to by the ith opinion, in the format of plot number or longitude and latitude coordinates; t i Indicates the opinion type code (e.g. 01 for green space suggestion, 02 for functional dissatisfaction, etc.); i is the importance weight of the opinion, the unit is dimensionless and the range is [0,1].

[0094] In one possible implementation, in order to enhance the semantic expression of opinions, the system introduces a natural language processing (NLP) model to perform semantic analysis on text input. The public can freely input descriptive language, and the system embeds the sentence through a word vector model (such as Word2Vec or BERT), and performs semantic clustering based on predefined categories, mapping the text to the above t i within the classification system.

[0095] As an option, the system further introduces an opinion density function to evaluate the concentration of public opinions in a specific area. The function is defined as:

[0096]

[0097] Wherein: D(x, y) is the density of public opinions at the geographical location (x, y); (x i -y i ) is the geographical location of the i-th opinion; w i is the weight of the i-th opinion; σ is the spatial scale parameter of the density distribution, with the unit of meter.

[0098] Generally, according to the spatial distribution of the density of public opinions, the system identifies the hotspots of public concern in the city and adjusts the identification of key areas in the urban renewal strategy accordingly.

[0099] In this embodiment, to ensure the spatial correspondence of opinion collection, the system automatically binds each public opinion to the specific plot unit in the public participation base map. Specifically, the spatial overlay method is used to identify the plot number l i where the opinion point falls, and it is used as the only spatial positioning basis for opinion records to ensure the traceability and quantifiability of public opinions in subsequent spatial analysis.

[0100] In some embodiments, to improve the quality of real feedback of public participation, the platform sets an opinion weight correction mechanism. This mechanism considers factors such as user portraits, historical participation behaviors, and identity authentication levels, and comprehensively corrects the initial value of w i . The correction function can be expressed as:

[0101] w′ i = w i ·β(u i );

[0102] Wherein: w′ i is the corrected opinion weight; w i is the original opinion weight; β(u i ) is the weight correction factor of user u i , and the range is [0.5, 1.5].

[0103] Specifically, users with a government certification background or a professional identity in urban planning have a relatively high correction factor to enhance the reference value of professional opinions; while the weight of anonymous users is correspondingly reduced.

[0104] S4. According to the collected public opinion data, construct a multi-agent scoring model, and score the candidate facility locations by setting the weight values of the public, experts, and the government;

[0105] Specifically, following the collection and semantic quantification of the aforementioned public opinions, step S4 mainly focuses on the fusion processing of public opinion data and spatial elements, establishing a public opinion response model, and realizing the impact modeling of public preferences on the urban renewal plan generation process. This step is not only a key node for the deep coupling of public participation and spatial decision-making mechanisms but also an important implementation link for the construction of multi-source heterogeneous data fusion and response mechanisms. Through this step, the inductive expression of public preferences at the spatial level can be achieved, laying a foundation for subsequent optimization models.

[0106] In this embodiment, it is first necessary to fuse the public opinion data with the urban land spatial units. Generally, the system maps the public opinion set O = {(l i , t i , w i )} to the corresponding plot units in the public participation base map through a spatial matching mechanism. Specifically, the system marks each opinion to the plot where it is located according to the geographical coordinates or plot numbers l i to achieve spatial association.

[0107] As an option, the system defines the public preference vector V j of each plot p j to represent the comprehensive opinion distribution of the plot in public expressions. The calculation formula of this vector is as follows:

[0108]

[0109] where: V j represents the public preference vector of the jth plot; δ(l i = p j ) is an indicator function, which takes the value of 1 if the location of the ith opinion is plot p j , otherwise 0; w i is the weight of the ith opinion; T i is the type vector of the ith opinion, and the dimension is the same as the number of opinion classification types, usually in the form of one-hot encoding.

[0110] In a possible implementation manner, the system normalizes the above vector to ensure the comparability of opinion intensities between different plots. The normalized vector is expressed as follows:

[0111]

[0112] where: is the normalized public preference vector of plot p j ; ∥V j ∥1 is the L1 norm of vector V j , that is, the sum of the absolute values of each element.

[0113] Generally, the system can identify the dominant preference types of the public on each plot through this vector, such as a greater preference for increased green spaces, traffic optimization, or the addition of public facilities.

[0114] In this embodiment, to enable public opinions to participate in the spatial model analysis in the form of response factors, the system further constructs a public opinion response matrix R, which is used to quantify the matching degree of different land use types to the public preference vector. The response matrix is defined as follows:

[0115]

[0116] Where: R j,k represents the public response value between the j-th plot and the k-th type of renewal land use plan; is the public preference vector of the j-th plot; U k is the weight vector of the k-th type of land use in the dimension of public preference, with the unit being dimensionless.

[0117] Specifically, the weight vector U k is usually set by expert experience or through clustering learning based on historical participation data. For example, for "community park" land use, the proportions of "increased greening rate" and "increased leisure space" are higher in its public preference dimension, and the system obtains the corresponding weight vector through sample learning.

[0118] In some embodiments, the system introduces a public response intensity index S j , which is used to quantify the response activity degree of each plot in the overall public participation, and is defined as follows:

[0119]

[0120] Where: S j represents the public response intensity of the j-th plot; m is the total number of land use type categories; R j,k is the element value in the above response matrix.

[0121] As an option, to avoid information loss caused by too low opinion responses of certain plots, the system sets a response intensity threshold θ. When S j < θ, this plot will not participate in the optimization calculation as a dominant factor of public opinion in the subsequent model.

[0122] S5. Conduct a suitability analysis on the candidate facility locations. The suitability analysis includes spatial indicators such as traffic accessibility, service coverage, and land developability to obtain a suitability score;

[0123] Specifically, after completing the spatial mapping and response modeling of public opinions, the present invention further enters the optimization and generation stage of the urban renewal plan, that is, step S5. Based on the public opinion response matrix and the public response intensity index constructed in step S4, combined with the spatial constraint conditions and functional objective requirements, a multi-objective optimization model is constructed, and an intelligent optimization algorithm is used to generate an urban renewal land use allocation plan that conforms to the public preference orientation. The technical core of this step lies in introducing public participation data into the objective function and constraint conditions of the multi-objective planning model to achieve the response and coordination of urban renewal decisions to the demands of multiple parties.

[0124] In this embodiment, first, a land use allocation matrix X for the entire urban renewal area is defined. This matrix is used to represent the land use types allocated to each spatial unit. Suppose the urban renewal area is divided into n plots, and each plot can select one of m land use types. Then:

[0125] X = {x ij}, i = 1, 2, …, n; j = 1, 2, …, m;

[0126] Where: x ij ∈ {0, 1}. If the i-th plot is allocated the j-th land use type, then x ij = 1; otherwise, it is 0.

[0127] Each plot is only allowed to be allocated one land use type. Therefore, it should satisfy:

[0128]

[0129] Specifically, the optimization objective function includes multiple objectives such as maximizing public satisfaction, improving spatial integration, and enhancing land use efficiency.

[0130] The public satisfaction objective function f1(X) is defined based on the public response matrix R as follows:

[0131]

[0132] Where: R ij is the public response value of the i-th plot to the j-th land use type; x ij is the land use allocation variable.

[0133] This function is used to evaluate the consistency between the overall land use allocation and public preferences. The larger the value, the higher the public satisfaction.

[0134] As an option, the system further constructs a spatial integration objective function f2(X). This function introduces the integration value Int i obtained from the aforementioned spatial syntax analysis and is defined as follows:

[0135]

[0136] where: Int i is the integration degree value of the i-th plot; represents the set of land use types with higher requirements for integration, such as commerce, public services, etc.

[0137] Generally, by allocating high-integration spaces to high-aggregation functions, the urban space efficiency and traffic convenience can be improved.

[0138] In a possible implementation, to control the total proportion of various types of land, the system sets the following constraint conditions:

[0139]

[0140] where: A j is the maximum total number allowed for the j-th type of land, in units of plots or area units.

[0141] At the same time, the system needs to satisfy the physical space constraints of the plots, the restriction conditions that the land use attributes in the historical and cultural protection areas cannot be changed, the policy constraints that the existing functions cannot be replaced randomly, etc. These restriction conditions will be added as hard constraint items in the model to ensure the feasibility of the optimization results.

[0142] In this embodiment, to solve the above multi-objective optimization problem, the system uses the non-dominated sorting genetic algorithm (NSGA-II) for multi-objective optimization. Generally, this algorithm can quickly converge to the Pareto optimal solution set while maintaining the population diversity. During the optimization iteration process, each individual in the population is a feasible land use configuration matrix X, and its fitness function simultaneously considers objective functions such as f1(X) and f2(X).

[0143] In some embodiments, the system conducts multiple rounds of sampling and re-evaluation on the optimization results to enhance the stability and acceptability of the solutions. Finally, the land use configuration plan with the highest public satisfaction score among the Pareto front solutions is used as the preferred land use configuration plan, and this plan is submitted to the visualization platform for further discussion and modification by the public and planners.

[0144] S6. Perform weighted calculation on the obtained comprehensive score value and suitability score value, and select the final optimal site;

[0145] Specifically, after generating the urban renewal land use allocation plan oriented to public preferences, the present invention enters step S6, that is, the feedback correction and iterative optimization stage of the renewal plan. The main objective of this step is to guide the public and decision-making entities to review and provide feedback on the preliminarily generated urban renewal plan, and based on the feedback information, carry out a new round of model response and optimization iteration to achieve the logical closed-loop of "co-construction and consensus" in the urban renewal process. Step S6 forms a closed loop with the aforementioned public opinion response mechanism and the optimization model solving logic to ensure that the urban renewal plan has the ability of continuous optimization and dynamic response.

[0146] In this embodiment, the system visualizes the optimal land use allocation plan generated in step S5 and displays it on the public participation platform. The platform supports interactive feedback operations based on the map unit level. The public can provide feedback on each land use unit p i including structured options such as "dissatisfied", "suggest modification", "prefer other land use types", etc.

[0147] Generally, after the system collects the feedback data, it immediately updates and corrects the original public response matrix R = {R ij}. The correction process calculates the public resistance factor for each type of land use based on the difference between the feedback opinion and the original allocation plan.

[0148] Specifically, assume that the original allocation of the i-th plot is the j-th type of land use. If F ij negative feedbacks are received, then the correction coefficient γ ij is constructed, and the correction coefficient is defined as follows:

[0149]

[0150] where: γ ij is the public response correction coefficient of the i-th plot under the land use type j; F ij is the number of dissatisfied feedbacks for this land use allocation; N i is the total number of feedbacks received by the i-th plot.

[0151] In a possible implementation manner, the system applies the correction coefficient to the original response matrix, and the updated public response value is:

[0152] R ij ′ = R ij ·γ ij ;

[0153] The above correction mechanism ensures that the feedback opinion can be quantitatively incorporated into the response model to provide a basis for the next round of optimization. At the same time, to control the feedback interference intensity, the system sets a minimum feedback threshold φ. When the system believes that this allocation has relative public recognition and does not make adjustments.

[0154] In this embodiment, the updated public response matrix R' = {R' ij} will be re - input into the multi - objective optimization model described in step S5, and jointly participate in the generation of a new round of Pareto solutions with the original spatial integration objective and land use ratio constraints.

[0155] In some embodiments, to evaluate the degree of public response difference between the new solution and the old solution, the system defines a scheme similarity index θ, and the formula is as follows:

[0156]

[0157] Where: represents the land use configuration in the current iteration round t; represents the configuration result of the previous round.

[0158] Generally, if the value of θ approaches 0, it indicates that the impact of public feedback is small and the scheme structure tends to be stable; otherwise, it means that public opinions have a strong corrective effect on the scheme, and the model needs to be continuously iterated to reach a consensus.

[0159] As an option, to control the volatility in the model convergence process, the system introduces a stability factor λ to adjust the intensity of the correction weight. The update formula can be further extended to:

[0160]

[0161] Where: λ ∈ [0, 1] represents the sensitivity of the system to accepting feedback, and the larger the value, the stronger the feedback effect.

[0162] S7. Visualize the most optimal site - selection result through the Geographic Information System (GIS) platform and feedback it to the public for further confirmation.

[0163] Specifically, after public opinion collection, land use configuration optimization, and feedback iteration and correction, the present invention finally enters step S7, that is, the stage of urban renewal decision - making output based on public participation data. This step not only undertakes the spatial land use plan generated by the foregoing optimization, but also is a key link for the transformation of public participation results into policy recommendations, spatial planning texts, and drawings. This step emphasizes the structured embedding of public opinions into decision - making content, and forms an executable, interpretable, and traceable urban renewal output system through spatial data integration, model analysis result archiving, and scheme result translation.

[0164] In this embodiment, the system first uses the finally optimized land use configuration plan Perform spatial data fusion with the public response matrix R″ and output a visualized land use configuration map. Generally, the system uses a standard geographic information mapping template to layer the configuration map. Different land use types are encoded with standardized colors and accompanied by a response index layer to represent the density or satisfaction distribution of public opinions.

[0165] Specifically, for each plot p i the public response intensity S i can be represented in the layer in the form of heat. The system calculates according to the formula:

[0166]

[0167] to calculate the comprehensive public response value of each plot.

[0168] In a possible implementation, to support policymakers' intuitive understanding of the effectiveness of public participation in the plan, the system automatically generates a public participation contribution report. This report quantifies the degree of influence of public opinions on the optimization result, and the calculation method is:

[0169]

[0170] where: I represents the public participation influence index; f1(X * ) is the public satisfaction score of the final plan; f1(X base ) is the satisfaction score of the model default plan under the condition of no public participation.

[0171] Generally, the higher this index, the more substantial the guiding role of public participation in the update plan.

[0172] As an option, the system further constructs a decision-making priority recommendation table based on the public orientation. This table determines the implementation sequence of urban renewal according to the comprehensive scores U i such as the public response intensity S i , plot integration degree Int i , cultural sensitivity C i . The formula is as follows:

[0173] U i = ω1·S i + ω2·Int i + ω3·C i ;

[0174] where: ω1, ω2, ω3 are the weight coefficients of public response, spatial accessibility, and cultural sensitivity respectively; C i is the quantitative representation of whether the area where the plot is located is a historical and cultural protection area; all parameter units are dimensionless, and the sum of the weights satisfies ω1 + ω2 + ω3 = 1.

[0175] In this embodiment, the system preferentially includes the plots with high scores in the update start list and can automatically generate a phased implementation suggestion form for urban renewal, supporting project ranking and resource scheduling in the actual policy formulation process.

[0176] In some embodiments, to ensure the traceability and transparency of the update plan, the system archives the entire process of all public opinions, model parameters, and optimization processes, encapsulates them in the form of JSON structured files, and stores them in the urban renewal database. The archived content includes the original data of public opinions, the changes in each stage of the response matrix, the parameter settings and output results of each round of optimization models, ensuring that the decision-making basis can be restored and audited in future urban renewal implementations.

[0177] Although the embodiments of the present invention have been shown and described, those of ordinary skill in the art can understand that various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention. The scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An urban renewal decision-making method based on public participation, characterized in that, It includes the following steps: Obtain basic urban spatial data, including vector maps of the city, remote sensing image data, and historical and cultural geographical information data; Based on the obtained spatial data, construct a base map for public participation, and the base map includes information on the road network, land use types, and historical and cultural regions of the city; On the base map for public participation, collect opinion data of the public on the facility location through an interactive platform, and the opinion data includes facility types, location coordinates, and importance scores; According to the collected public opinion data, construct a multi-agent scoring model, and by setting the weight values of the public, experts, and the government, score the candidate facility locations; Conduct a suitability analysis on the candidate facility locations, and the suitability analysis includes spatial indicators such as traffic accessibility, service coverage, and land developability to obtain a suitability score; Perform a weighted calculation on the obtained comprehensive score value and the suitability score value to select the final optimal location; Visualize the optimal location result through a Geographic Information System (GIS) platform and feedback it to the public for further confirmation.

2. The urban renewal decision-making method based on public participation according to claim 1, wherein, The construction of the base map for public participation includes: Use the space syntax method to model the urban road network and extract the spatial integration degree and spatial choice degree of road nodes; Based on the integration degree and choice degree, screen the urban roads and construct the urban skeleton network; Based on a convolutional neural network model, identify the land use types of remote sensing images and generate a spatial layer with semantic labels; Overlay the historical block and cultural landmark data on the map through buffer analysis and mark the boundary area of the urban context.

3. The urban renewal decision-making method based on public participation according to claim 2, characterized in that, The calculation of the spatial integration degree is based on the following rules: The integration degree of any node is equal to the function value of its average path depth to all other nodes, where the average path depth is the average minimum path length from this node to all nodes in the system.

4. The urban renewal decision-making method based on public participation according to claim 2, characterized in that, The convolutional neural network model includes: The input layer receives remote sensing image patch data, and the image contains three color channels; The convolutional layer and the pooling layer extract spatial texture features and output an intermediate feature map; The fully connected layer maps the feature map to the classification labels of land use types, and the output is a predefined set of land classes; The loss function uses the cross-entropy function for the supervised learning process.

5. The urban renewal decision-making method based on public participation according to claim 1, wherein The collection of public opinions includes: The public selects the proposed facility type through an interactive map interface; The public selects the proposed facility location coordinates on the map; The public submits feedback information, including facility type identification, location coordinates, and subjective importance scores; The public feedback information is written into the database in a structured manner.

6. The urban renewal decision-making method based on public participation according to claim 1, wherein The construction of the multi-agent scoring model includes: Set three types of agents: the public, experts, and the government, and assign a fixed weight value to each type of agent, and the sum of the three is 1; Each type of agent assigns an independent score to each candidate location; Weight the scores of each type of agent according to the preset weights and output the comprehensive score value of the candidate location.

7. The urban renewal decision-making method based on public participation according to claim 6, characterized in that, The public scores are derived from the statistical results of the spatial density of public opinion data, the expert scores are derived from the matching of facility planning experience rules, and the government scores are calculated based on the requirements of urban strategic planning control.

8. The urban renewal decision-making method based on public participation according to claim 1, wherein, The suitability analysis of the facility location includes: Set multiple spatial suitability indicators, including traffic accessibility, service radius coverage, land development feasibility, environmental friendliness, and policy restriction factors; Perform normalization processing on the minimum and maximum values of each indicator; Assign weight values to each indicator and calculate the comprehensive suitability scores of each candidate location.

9. A method for urban renewal decision-making based on public participation according to claim 1, characterized in that The calculation of the most optimal site selection includes: Set adjustment parameters to control the weight balance ratio between public scores and suitability scores; For each candidate location, calculate its comprehensive score equal to the public score multiplied by the complement of the adjustment factor plus the suitability score multiplied by the adjustment factor; Select the location with the highest score as the final site selection result.

10. A method for urban renewal decision-making based on public participation according to claim 1, characterized in that, The visual output of the geographic information system platform includes: A public opinion heat map layer, reflecting the concentrated areas of public feedback; A facility suitability score layer, reflecting the spatial distribution of the suitability degree of each location; A most optimal site selection result layer, indicating the final recommended facility layout location.

Images