An urban planning decision feasibility analysis system based on a smart city

Abstract

The application relates to a smart city-based urban planning decision feasibility analysis system, aiming to solve the problems of experience dominance, static solidification, appeal imbalance and technology isolation in traditional urban planning. The system includes four modules: data collection, algorithm processing, decision output and dynamic optimization. The data collection module dynamically collects urban basic space, dynamic monitoring, industry and livelihood data and completes preprocessing; the algorithm processing module realizes spatial structure quantization, resource allocation optimization, planning risk assessment and dynamic feasibility prediction; the decision output module generates a planning scheme containing land use layout, energy configuration and public service facility distribution parameters; the dynamic optimization module updates and threshold triggering mechanism, iteratively adjusts the planning scheme. The system is suitable for new area planning and stock area renovation scenarios, realizes the transformation of planning decision, and improves the scientificity and landing effect of planning decision.

Description

Technical Field

[0001] This application belongs to the field of smart cities, specifically involving a feasibility analysis system for urban planning decisions based on smart cities. Background Technology

[0002] Against the backdrop of rapid urbanization and the deep integration of smart city construction, traditional urban planning has long relied on the experience and subjective qualitative analysis of decision-makers. This has resulted in insufficient mining and inefficient utilization of multi-source data generated during the dynamic operation of cities. Furthermore, it lacks quantitative assessment tools for core objectives such as economic development, ecological protection, and people's livelihood, often leading to conflicting goals and hindering synergistic balance. Traditional planning is largely static, lacking effective alignment with the timeline of urban development. Moreover, cross-regional and cross-departmental planning lacks a coordinated mechanism, easily resulting in fragmented spatial layouts. Planning decisions tend to involve unbalanced resource allocation, failing to adequately consider the demands of diverse stakeholders and providing insufficient protection for public interests. Existing planning tools employ independent algorithms for spatial analysis, risk assessment, and resource allocation, lacking a closed-loop system for data flow and collaborative linkage, thus failing to achieve dynamic data feedback and efficient utilization.

[0003] Against this backdrop, there is an urgent need to build an urban planning decision-making feasibility analysis system that adapts to the development needs of smart cities, integrates multi-source data and coupled algorithms, and has dynamic optimization capabilities, in order to break through many bottlenecks of traditional planning models. Summary of the Invention

[0004] The purpose of this application is to solve the above-mentioned problems by designing a feasibility analysis system for urban planning decisions based on smart cities, characterized by comprising:

[0005] The data acquisition module is used to dynamically collect urban basic spatial data, dynamic monitoring data, and industry and people's livelihood data, and to perform preprocessing.

[0006] Algorithm processing module, the algorithm processing module includes:

[0007] The spatial syntax unit is used to output the global integration degree (GI) and local traversal potential (LTP) based on spatial syntax as the underlying algorithm, serving as the spatial constraint basis for resource allocation.

[0008] The multi-objective particle swarm optimization unit is used to optimize the allocation of urban land, energy and public service facilities resources with the global integration degree (GI) and local traversal potential (LTP) as hard constraints, aiming to maximize economic benefits, minimize carbon emissions and maximize public satisfaction, and output resource allocation schemes.

[0009] An improved Bayesian network unit is used to integrate the resource allocation scheme with real-time monitoring data from the Internet of Things, and to calculate the probability of occurrence of three core risks—floods, traffic congestion, and air pollution—using a dynamic conditional probability formula.

[0010] The spatiotemporal coupled system dynamics unit is used to integrate the global integration degree (GI) and local traversal potential (LTP) output by the spatial syntax unit, the resource allocation scheme output by the multi-objective particle swarm optimization unit, and the occurrence probability output by the improved Bayesian network unit, and to construct a spatiotemporal dimension urban development model through the spatiotemporal coupled state equation.

[0011] The decision output module is used to generate planning schemes based on the urban development model constructed by the algorithm processing module.

[0012] The dynamic optimization module includes a dual-cycle regular update unit and a threshold-triggered dynamic optimization unit, which are used to dynamically adjust and iteratively optimize the planning scheme.

[0013] Preferably, the urban planning decision-making feasibility analysis system based on smart cities relies on a smart city sensing network, which includes the Internet of Things, remote sensing, GIS, and a big data platform.

[0014] Preferably, the core function of the spatial syntax is to parse the topological relationships of urban road networks and public spaces, quantify spatial accessibility and usage potential, and output the global integration degree. and local travel potential As the spatial constraint basis for resource allocation;

[0015] Wherein, the global integration degree Through formula calculate,

[0016] Where n is the total number of spatial nodes in the study area, and its value range is... ; It is a logarithmic function with base 2; MD is the average shortest path length of the nodes, with a value ranging from 50 to 2000m; It is a logarithmic function with base 2;

[0017] Local travel potential Through formula calculate,

[0018] in For nodes The set of nodes within a 500m radius. For nodes To the node The straight-line distance, ranging from 10 to 500 meters. For nodes The average daily pedestrian flow, with a value range of [value missing]. Number of people For nodes The corresponding plot area, with a range of values, is as follows: .

[0019] Preferably, the particle position update formula of the multi-objective particle swarm optimization algorithm is as follows:

[0020]

[0021] The particle velocity update formula is:

[0022] ,

[0023] The optimization criteria of the multi-objective particle swarm optimization algorithm include:

[0024] Maximize economic benefits :

[0025]

[0026] Minimize carbon emissions :

[0027]

[0028] Maximize public satisfaction :

[0029]

[0030] in, For the first The particle in the first The position of the dimension, with a value range of 0 to 1; The particle velocity is represented by a value ranging from -0.5 to 0.5. This is the inertia weight, with a value ranging from 0.4 to 0.9; For individual learning factors; As a whole learning factor; For each individual; The result is a random number for the whole system; The optimal position for an individual particle, with a value ranging from 0 to 1; This represents the globally optimal position for the population, and its value ranges from 0 to 1. The area allocated for the kth type of resource is 0.1 to 100 km². Let k be the energy consumption coefficient for land use category k. For the first The economic output coefficient of land use category ranges from 0.5 to 1 billion yuan / km². The development intensity coefficient for the kth land use category ranges from 0.6 to 1.2. The carbon emission coefficient for land use category k ranges from 0.1 to 50,000 tons / km². For the first The average service distance for public service facilities ranges from 200 to 1500 meters. For the first The usage frequency weight of facility type ranges from 0.5 to 2;

[0031] The constraints of the multi-objective particle swarm optimization algorithm include:

[0032] Global integration constraints:

[0033]

[0034] Local travel potential constraints:

[0035]

[0036] Total resource constraints:

[0037] ,

[0038] in, This refers to the global integration degree of node i; For the local traversal potential of node i; This represents the minimum global integration level, ranging from 0.3 to 0.6. The maximum local crossing potential is defined as follows, with a range of values. The total planned land area is 10 to 1000 km².

[0039] Preferably, the improved Bayesian network unit integrates the resource allocation scheme output by the multi-objective particle swarm optimization algorithm with real-time monitoring data from the Internet of Things to assess the probability of occurrence of risks such as floods, traffic congestion, and air pollution;

[0040] The dynamic conditional probability of the improved Bayesian network unit is expressed by the formula...

[0041] calculate,

[0042] in, The dynamic conditional probability value represents the probability at n input nodes. Given a given state, the probability that the target node / intermediate node Y is in a certain state; For node variables; For input node variables; For input node The dynamic calibration weights are derived from the improved Bayesian network unit dynamic weight calibration formula. It is dynamically updated according to the multi-objective particle swarm algorithm scheme; For input node The conditional probability of a node Y in a certain state comes from the improved Bayesian network unit conditional probability table; The prior probability of node Y is set by historical data and dynamically fine-tuned during iterative calibration. The chain multiplication symbol represents the product of n input nodes. Perform a series of multiplication operations; The summation symbol represents the summation of the product of the two states of node Y, used for normalization to ensure that the output probability value is within the range of 0 to 1.

[0043] Preferably, the spatiotemporal coupled system dynamics integrates spatial syntax, multi-objective particle swarm optimization algorithm, and the output results of improved Bayesian network units to construct a spatiotemporal urban development model;

[0044] The spatiotemporal coupling state equation of the dynamics of the spatiotemporal coupled system is as follows:

[0045] ,

[0046] in, The value range is continuous. Rate of change over time; This refers to the inflow rate, with the corresponding unit being annual. This is the outflow rate, with the corresponding unit being annual. The coefficient representing the influence of reachability on state variables ranges from 0.1 to 0.5. Spacetime coordinates The global integration degree at the location ranges from 0.3 to 1; This is the risk suppression coefficient for state variables, with a value ranging from 0.2 to 0.6. Spacetime coordinates The overall risk value at the location ranges from 0 to 1.

[0047] The inflow rate of the state variable includes the population migration flow rate.

[0048] ,

[0049] in, Spacetime coordinates The population is measured in units ranging from 100,000 to 1,000,000. Spacetime coordinates The number of job positions ranges from 5,000 to 500,000. This represents the natural growth rate, ranging from -5 to 15‰. The driving force of migration potential on population migration is the coefficient, with a value ranging from 0.001 to 0.01. Spacetime coordinates The range of local traversal potential values ​​at a given location is: .

[0050] Preferably, the preprocessing process of the data acquisition module includes unifying data units and removing outliers. The basic spatial data includes road network topology data and land use status data. The dynamic monitoring data includes pedestrian flow, vehicle flow, precipitation, and air quality data. The industry and livelihood data include economic output coefficient, carbon emission coefficient, and public service facility usage frequency weight.

[0051] Preferably, the planning scheme generated by the decision output module includes land use layout, energy configuration, and public service facility distribution parameters.

[0052] Preferably, the dual-cycle routine update unit is used to perform quarterly real-time monitoring updates and annual parameter calibration updates; the threshold-triggered dynamic optimization unit is used to automatically trigger the multi-objective particle swarm optimization unit to re-optimize resource allocation when the real-time monitoring data exceeds a preset threshold range; the threshold-triggered dynamic optimization unit is configured to perform the following optimization steps:

[0053] Data freeze and scene locking: Lock the current planning scheme and monitoring data snapshot at the trigger time;

[0054] Relaxing constraints: Under the premise of meeting the hard constraints of global integration (GI) and local transit potential (LTP), the total planned land area constraint is relaxed;

[0055] Multi-objective particle swarm optimization re-optimization: Using the current planning scheme as the initial population, re-run the multi-objective particle swarm optimization unit to generate alternative adjustment schemes;

[0056] Risk assessment filtering: The improved Bayesian network unit is used to calculate the comprehensive risk value of flooding, traffic congestion and air pollution of the alternative adjustment schemes, and high-risk schemes with a comprehensive risk value R>0.7 are eliminated;

[0057] Solution optimization and output: Based on the Pareto frontier, the optimal solution is selected and the parameters of land use layout, energy configuration and distribution of public service facilities are updated;

[0058] Effectiveness of the plan and monitoring in the next cycle: Output the adjusted planning plan and enter the monitoring cycle for the next quarter.

[0059] Preferably, the urban planning decision-making feasibility analysis system based on smart cities is applicable to new area planning or existing area renovation and transformation scenarios. In the existing area renovation and transformation scenario, weight coefficients are set for nodes around historical buildings during spatial syntax calculation, and constraints on the protection scope of historical buildings are added during multi-objective particle swarm optimization.

[0060] The beneficial effects of this application are as follows:

[0061] Relying on smart city sensing networks such as the Internet of Things, remote sensing, GIS, and big data platforms, the system dynamically integrates three types of data: basic spatial data, dynamic monitoring data, and data on industry and people's livelihood data. Through a preprocessing process that unifies units and eliminates outliers, it provides comprehensive quantitative basis for decision-making, completely changing the limitations of traditional planning's subjective qualitative judgment and enabling core requirements such as technological adaptability and economic rationality to be quantitatively assessed.

[0062] This innovative approach integrates four algorithms: spatial syntax, multi-objective particle swarm optimization (MPS), improved Bayesian networks, and spatiotemporally coupled system dynamics, forming a coupled logic of spatial quantification, resource optimization, risk assessment, and dynamic prediction. For example, the GI and LTP outputs of spatial syntax provide constraints for the multi-objective PMS, while the resource allocation scheme of the multi-objective PMS provides risk assessment inputs for the improved Bayesian network units. Ultimately, the improved Bayesian network units achieve multi-objective balance, solving the problem of resource bias in traditional planning.

[0063] A dual-cycle optimization mode is implemented, consisting of quarterly real-time monitoring updates and annual parameter calibration updates. When monitoring data exceeds a preset threshold, a multi-objective particle swarm optimization algorithm is automatically triggered to re-optimize resource allocation. This closed-loop system enables planning schemes to quickly respond to dynamic urban changes such as population influx and industrial transformation, addressing the pain points of traditional planning's disconnect and difficulty in implementation. Attached Figure Description

[0064] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments of this application will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0065] Figure 1 This is the overall system block diagram;

[0066] Figure 2 It is a graph showing the data flow relationship in the algorithm. Detailed Implementation

[0067] To enable those skilled in the art to better understand the technical solutions of this application, the application will be further described in detail below with reference to the accompanying drawings and preferred embodiments.

[0068] This application introduces a feasibility analysis system for urban planning decisions based on smart cities, including a data acquisition module, an algorithm processing module, a decision output module, and a dynamic optimization module.

[0069] The data acquisition module collects and preprocesses multi-source data, such as basic spatial data, dynamic monitoring data, and industrial and livelihood data, including unifying units and removing outliers, and outputs a dataset as the core input to the algorithm processing module. The algorithm processing module uses four coupled algorithms to sequentially output spatial quantitative indicators GI and LTP, resource allocation schemes, risk assessment results, and dynamic prediction curves, ultimately forming a set of feasible planning schemes including parameters such as land use layout and energy allocation. The decision output module does not need to generate schemes independently; it only needs to select the optimal solution based on the algorithm results to ensure quantitative support for decision-making.

[0070] The data acquisition module relies on IoT terminals, remote sensing satellites, and GIS geographic information systems to build an integrated data acquisition system. It collects basic spatial data including road network topology data (such as intersections and road segment connections), land use status data (such as plot area, use, ownership, topography, building features, and administrative divisions), and dynamic monitoring data (such as pedestrian and vehicle flow data, precipitation data, air quality data, resource and environmental monitoring data such as green space coverage and water resource status, and infrastructure operation data such as traffic lights and public facility usage status). It also collects dynamic monitoring data in real time through IoT sensors, video recognition equipment, and industry system integration (such as traffic monitoring, weather stations, and edge computing nodes). Furthermore, it integrates with government data sharing platforms, enterprise reporting systems, mobile internet big data analysis, and questionnaire surveys to collect industrial and livelihood data (such as economic output coefficients, carbon emission coefficients, public service facility usage frequency weights, and industrial development data such as employment numbers, output value, population and livelihood data, and consumption and service data). Specific implementation details, interface protocols, storage formats, and acquisition frequencies are specified as follows:

[0071] 1. Specific methods of data collection

[0072] IoT data acquisition: Deploy smart sensors such as temperature and humidity, settlement, noise, and pedestrian flow sensors in urban nodes, including around historical buildings, transportation hubs, and municipal facilities. The sensor model is industrial-grade RS-485, which supports wired and wireless dual-mode transmission. The deployment density is set according to the importance of the node, such as one sensor every 50 meters around historical buildings and one sensor every 100 meters around ordinary municipal nodes. The sensors collect on-site data in real time and actively upload it to the edge gateway.

[0073] Remote sensing data acquisition: Select Gaofen-1 or Gaofen-2 satellite remote sensing images, with a resolution of 1 meter for the core area and 2 meters for the general area. Connect to the National Remote Sensing Data Service Platform through the remote sensing data interface to acquire spatial data such as urban land use, building distribution, and vegetation cover. Use preprocessing methods such as orthorectification and image fusion to ensure that the data accuracy meets the system requirements.

[0074] GIS Big Data Acquisition: Connecting with the city's existing GIS geographic information platform, and using API interfaces to capture basic geographic data such as road networks, administrative divisions, coordinates of historical buildings, and distribution of municipal facilities; at the same time, connecting with the government data platform to obtain structured data such as population, economy, and planning approvals, thereby achieving linkage between basic data and dynamic data.

[0075] 2. Clear data collection frequency

[0076] Real-time monitoring data such as historical building settlement, pedestrian flow, and noise are collected once per minute. The edge gateway caches 10 minutes of data and uploads it synchronously to the core database to ensure that the real-time performance meets the system's dynamic monitoring requirements.

[0077] Periodic data such as remote sensing images and land use data are collected monthly in core areas and quarterly in general areas, and are simultaneously updated to the GIS platform to support model iteration.

[0078] Basic static data, such as administrative divisions and basic information on historical buildings, are collected once a year. If changes occur, such as planning adjustments or building renovations, updates are triggered manually and the data is synchronized within 24 hours.

[0079] 3. Data storage format specifications

[0080] Dynamic monitoring data, such as sensor readings and pedestrian flow data, are stored in JSON format. Fields include: collection timestamp, device ID, coordinates, data value, and data quality identifier, facilitating quick parsing and retrieval by the algorithm module.

[0081] Spatial geographic data such as remote sensing and GIS: vector data uses Shapefile format, remote sensing images use TIFF format, and the coordinate system is WGS84 coordinate system to ensure spatial positioning consistency; the compression ratio of remote sensing images is set to 1:5 to balance storage capacity and data accuracy.

[0082] Structured data such as population, economic data, and weighting parameters are stored in a MySQL relational database with UTF-8 character encoding. Numerical data is formatted to four decimal places, and dates are formatted as YYYY-MM-DDHH:MM:SS.

[0083] 4. Interface protocol and data access method

[0084] Sensor and edge gateway interface: The Modbus-RTU protocol is used, the baud rate is set to 9600bps, and the data verification method is CRC-16 to ensure data transmission stability. The edge gateway uploads data to the core database via the MQTT protocol.

[0085] Algorithm and data module interfaces: RESTful API interfaces are used, with the unified interface address being "http: / / [server IP]:8080 / data / api / v1". It supports two request methods: GET for data querying and POST for data submission, with a response time of ≤500ms, enabling the algorithm to quickly access real-time and historical data. Interface permissions are authenticated using a Token, valid for 24 hours, ensuring data security.

[0086] GIS platform and core database interface: Adopting the OGCWMS / WFS protocol, it enables online publishing, querying and overlay display of geographic data, and supports algorithm modules to call GIS spatial analysis functions such as buffer analysis and overlay analysis.

[0087] The algorithm processing module includes space syntax, multi-objective particle swarm optimization, improved Bayesian network, and spatiotemporal coupled system dynamics; space syntax serves as the underlying algorithm, outputting the global integration degree. and local travel potential As the spatial constraint basis for resource allocation; the multi-objective particle swarm optimization algorithm outputs the aforementioned spatial syntax. , As a hard constraint, focusing on maximizing economic benefits, minimizing carbon emissions, and maximizing public satisfaction, the system optimizes the allocation of core resources such as urban land, energy, and public service facilities, ultimately outputting a resource allocation plan. The improved Bayesian network unit integrates the resource allocation plan output by the multi-objective particle swarm optimization algorithm with real-time monitoring data collected by the Internet of Things, calculating the probability of occurrence of three core risks—flooding, traffic congestion, and air pollution—using dynamic conditional probability formulas. Furthermore, the system improves the global integration degree of the Bayesian network unit's integrated space syntax output. and local travel potential The system employs a multi-objective particle swarm optimization algorithm to output resource allocation schemes and improved Bayesian network unit output probability. Through spatiotemporal coupled state equations, it constructs a spatiotemporal urban development model. The four algorithms are executed sequentially, with data flow and complementary functions. The output of the preceding algorithm serves as the input of the following algorithm, forming a technical closed loop from spatial constraints to resource optimization to risk screening to dynamic prediction. This ensures that the planning scheme has spatial rationality, multi-objective balance, risk controllability, and long-term sustainability.

[0088] The decision output module is based on the quantitative results of the algorithm processing module, and follows the following principles:

[0089] All solutions must meet the rigid thresholds set by the algorithm processing module to ensure compliance with the hard constraints of not breaching the planning bottom line; based on the three objective functions of the multi-objective particle swarm optimization algorithm, the solution balances economic, ecological, and livelihood demands, avoiding overall imbalance caused by the optimization of a single objective; for two major scenarios, namely new area planning and existing area renewal, the solution outputs a customized solution based on the principle of scenario adaptation and differentiation; the solution outputs a land use layout that clearly defines the area, location, and development intensity of various types of land use; an energy allocation solution that optimizes the energy supply structure by combining industrial layout and ecological constraints; and a public service facility solution that optimizes the facility layout based on the goal of maximizing people's satisfaction, providing a clear carrier for subsequent review and dynamic optimization;

[0090] The dynamic optimization module includes a dual-cycle regular update unit and a threshold-triggered dynamic optimization unit, used to dynamically adjust and iteratively optimize the planning scheme; the dual-cycle regular update unit is used to perform quarterly real-time monitoring updates and annual parameter calibration updates; the specific rules are as follows:

[0091] 1. Quarterly real-time monitoring and updates

[0092] Monitoring period: Monitoring is completed in the last week of each quarter, with data covering the entire three months of the quarter. Specific monitoring indicators:

[0093] Data level: Data collection accuracy ≥ 98%, data transmission latency ≤ 500ms, data missing rate ≤ 2%; Model level: Prediction deviation rate of urban development model of spatiotemporal coupled system dynamic unit ≤ 5%, weight coefficient application accuracy ≥ 95%; System level: Interface response time ≤ 500ms, system operation failure rate ≤ 0.5%, priority matching degree of existing renovation nodes ≥ 90%.

[0094] Update methods and procedures:

[0095] a. Automatic monitoring: The system has a built-in monitoring module that automatically captures the above indicator data at the end of each quarter and generates a "Quarterly Monitoring Report" containing the indicator compliance status, deviation data and cause analysis;

[0096] b. Manual review: Technical personnel will review the monitoring report within 3 working days to confirm the cause of the deviation, such as data acquisition failure, unreasonable weighting coefficient, etc.

[0097] c. Rapid updates: For deviations that can be automatically adjusted, such as data acquisition frequency optimization and interface parameter fine-tuning, the system will automatically complete the update; for deviations that require manual intervention, such as local adjustment of weight coefficients, the system will be triggered to complete the synchronous update within 24 hours after the manual adjustment is completed.

[0098] 2. Annual parameter calibration update

[0099] Calibration cycle: Calibration is completed in late December each year, covering all system operation data for the current year, and the calibration results are used as the system operation benchmark for the following year.

[0100] Parameter calibration object:

[0101] Algorithm parameters: spatiotemporal coupling parameters and predicted attenuation coefficient in the urban development model of spatiotemporal coupled system dynamics unit; range of values ​​for distance factor and node type factor in the weight coefficient of existing area renovation;

[0102] Data parameters: Optimized data acquisition frequency, data storage compression ratio, and interface response thresholds for different regions and node types;

[0103] Control parameters: boundary of the protection area of ​​historical buildings, priority determination threshold of existing renovation nodes.

[0104] Calibration method:

[0105] Automatic calibration: The system adopts the deviation feedback calibration method. Based on the monitoring data and actual transformation effect data of the current year, it automatically calculates the optimal values ​​of parameters, such as the parameters of the dynamic unit model of the spatiotemporal coupled system. It fits the actual data and predicted data through the least squares method and adjusts the parameters to minimize the deviation rate. Manual calibration: For parameters that still do not meet the standards after automatic calibration or involve planning adjustments, technicians will manually adjust and calibrate them in conjunction with on-site surveys and government planning documents. The calibration process is recorded in writing. Calibration verification: After calibration, the system uses data from the last month of the year for verification to ensure that the system monitoring indicators meet the standards after all parameters are calibrated, such as the model prediction deviation rate ≤5%. After verification, the system can be officially applied to the operation of the next year.

[0106] Dual-cycle linkage: The deviation data updated quarterly serves as the core basis for annual parameter calibration; the parameters after annual calibration serve as the benchmark threshold for real-time monitoring in the next quarter, forming a closed loop of monitoring-update-calibration-optimization.

[0107] The threshold-triggered dynamic optimization unit is used to automatically trigger the multi-target particle swarm optimization unit to reconfigure resources when real-time monitoring data exceeds a preset threshold range; the preset threshold range includes:

[0108] Plot ratio deviation threshold: ±0.2, triggered when exceeding the threshold by ≥5%;

[0109] Road network saturation threshold: 0.4–0.7, triggered when >0.7 or <0.4;

[0110] Air Quality Index threshold: 0-100, triggered by >100 for 3 consecutive days;

[0111] The threshold-triggered dynamic optimization unit is configured to perform the following optimization steps:

[0112] Data freeze and scene locking: Lock the current planning scheme and monitoring data snapshot at the trigger time;

[0113] Relaxing constraints: Under the premise of meeting the hard constraints of global integration (GI) and local transit potential (LTP), the total planned land area constraint is relaxed;

[0114] Multi-objective particle swarm optimization re-optimization: Using the current planning scheme as the initial population, re-run the multi-objective particle swarm optimization unit to generate alternative adjustment schemes;

[0115] Risk assessment filtering: The improved Bayesian network unit is used to calculate the comprehensive risk value of flooding, traffic congestion and air pollution of the alternative adjustment schemes, and high-risk schemes with a comprehensive risk value R>0.7 are eliminated;

[0116] Solution optimization and output: Based on the Pareto frontier, the optimal solution is selected and the parameters of land use layout, energy configuration and distribution of public service facilities are updated;

[0117] Effectiveness of the plan and monitoring in the next cycle: Output the adjusted planning plan and enter the monitoring cycle for the next quarter.

[0118] The aforementioned smart city-based urban planning decision-making feasibility analysis system relies on a smart city sensing network, which includes the Internet of Things (IoT), remote sensing, and GIS. The IoT covers key areas such as urban road network nodes, public spaces, industrial parks, and ecological green spaces, deploying pedestrian counting sensors, vehicle flow monitoring equipment, precipitation sensors, air quality monitors, and soil moisture sensors. The remote sensing includes topographic data, land use status data, building density distribution, and green space coverage. The GIS unifies the coordinates and converts the macro-spatial data acquired by remote sensing, the point data collected by the IoT, and the road network topology data to form a structured spatial database. It supports the construction of topological relationships between road network nodes and land parcel boundaries, providing a basic topological model for spatial syntax calculation of global integration and local transit potential. It generates land use layout maps, spatial accessibility heat maps, risk distribution diagrams, etc., providing a visual result carrier for the decision output module.

[0119] The core function of the spatial syntax is to analyze the topological relationships of urban road networks and public spaces, quantify spatial accessibility and usage potential, and output the global integration degree. and local travel potential As a spatial constraint basis for resource allocation, the topological relationships include identifying the hierarchical connections between main roads, secondary roads, and branch roads, clarifying the roles of different roads (e.g., by analyzing the number and connectivity of road network nodes, identifying core channels and potential bottlenecks in regional traffic flow); sorting out the connection relationships between public spaces and surrounding road networks, residential plots, and commercial facilities, and assessing the foundation of public spaces (e.g., whether a square is directly connected to multiple roads or is adjacent to densely populated residential nodes); the core value of the spatial syntax lies in transforming abstract urban spatial relationships into calculable and constrainable quantitative indicators, providing a scientific spatial logic basis for subsequent multi-objective resource allocation such as economic, ecological, and livelihood resources.

[0120] Wherein, the global integration degree Through formula calculate,

[0121] Where n is the total number of spatial nodes in the study area, and its value range is... MD is the average shortest path length of the node, with a value ranging from 50 to 2000m;

[0122] Local travel potential Through formula calculate,

[0123] in For nodes The set of nodes within a 500m radius. For nodes To the node The straight-line distance, ranging from 10 to 500 meters. For nodes The average daily pedestrian flow, with a value range of [value missing]. Number of people For nodes The corresponding plot area, with a range of values, is as follows: .

[0124] The multi-objective particle swarm optimization algorithm outputs a space syntax. and As a constraint, the particle position update formula of the multi-objective particle swarm optimization algorithm is as follows:

[0125]

[0126] The particle velocity update formula is:

[0127] ,

[0128] The optimization criteria of the multi-objective particle swarm optimization algorithm include:

[0129] Maximize economic benefits :

[0130]

[0131] Minimize carbon emissions :

[0132]

[0133] Maximize public satisfaction :

[0134]

[0135] in, For the first The particle in the first The position of the dimension, with a value range of 0 to 1; Particle velocity, ranging from -0.5 to 0.5; This is the inertia weight, with a value ranging from 0.4 to 0.9; , The learning factor has a value range of 1.5 to 2.5. , The result is a random number, ranging from 0 to 1. The optimal position for an individual particle, with a value ranging from 0 to 1; This represents the globally optimal position for the population, and its value ranges from 0 to 1. The area allocated for the kth type of resource is 0.1 to 100 km². For the first The economic output coefficient of land use category ranges from 0.5 to 1 billion yuan / km². The development intensity coefficient for the kth land use category ranges from 0.6 to 1.2. The carbon emission coefficient for land use category k ranges from 0.1 to 50,000 tons / km². For the first The average service distance for public service facilities ranges from 200 to 1500 meters. For the first The usage frequency weight of facility type ranges from 0.5 to 2;

[0136] The constraints of the multi-objective particle swarm optimization algorithm include: , ,in The value range is 0.3 to 0.6. The range of values ​​is The total planned land area is 10 to 1000 km².

[0137] The improved Bayesian network unit integrates resource allocation schemes output by the multi-objective particle swarm optimization algorithm with real-time IoT monitoring data to assess the probability of occurrence of risks such as flooding, traffic congestion, and air pollution. The flood risk is assessed by associating parameters in resource allocation such as the proportion of green space, road network drainage density, and land elevation planning, combined with real-time precipitation and soil moisture monitoring data. The traffic congestion risk is assessed by linking land use layout, such as the proportion of industrial or residential land, and road network density, with real-time vehicle and pedestrian flow monitoring data to predict the likelihood of node congestion. The air pollution risk is assessed by associating scheme parameters such as industrial land layout and energy configuration type, combined with real-time air quality data. The improved Bayesian network unit transforms abstract planning schemes into quantifiable risk indicators, providing key support for achieving a closed loop of multi-objective optimization combined with low-risk implementation through dual-dimensional data fusion.

[0138] The dynamic conditional probability of the improved Bayesian network unit is expressed by the formula...

[0139] calculate,

[0140] in, The dynamic conditional probability value represents the probability at n input nodes. Given a given state, the probability that the target node / intermediate node Y is in a certain state; For node variables; For input node variables; For input node The dynamic calibration weights are derived from the improved Bayesian network unit dynamic weight calibration formula. It is dynamically updated according to the multi-objective particle swarm algorithm scheme; For input node The conditional probability of a node Y in a certain state comes from the improved Bayesian network unit conditional probability table; The prior probability of node Y is set by historical data and dynamically fine-tuned during iterative calibration. The chain multiplication symbol represents the product of n input nodes. Perform a series of multiplication operations; The summation symbol represents the summation of the product of the two states of node Y, used for normalization to ensure that the output probability value is within the range of 0 to 1.

[0141] The spatiotemporal coupled system dynamics integrates the outputs of spatial syntax, multi-objective particle swarm optimization, and improved Bayesian network units to construct a spatiotemporal urban development model. The improved Bayesian network unit inherits the global integration degree and local traversal potential output from the spatial syntax, quantifying the supporting role of spatial accessibility and transportation potential in urban development. It incorporates core parameters such as land use layout, energy allocation, and distribution of public service facilities optimized by the multi-objective particle swarm optimization algorithm, clarifying the material basis and configuration logic of urban development. It integrates the comprehensive risk value calculated by the improved Bayesian network unit, incorporating risks such as flooding, traffic congestion, and air pollution as development inhibitors into the model, ensuring the safety and sustainability of the prediction results. Through the organic integration of three-dimensional inputs, the improved Bayesian network unit can simulate the evolution trends of core indicators such as urban population, economy, and ecology over the next 10-30 years, providing a quantitative basis for the long-term feasibility of planning schemes.

[0142] A spatiotemporal coupled system dynamics urban development model is proposed, specifying output indicators, prediction time scales, and verification methods to ensure that the model output is interpretable, applicable, and verifiable. Specific details are as follows:

[0143] 1. Specific output metrics of the model

[0144] The model output metrics are divided into core metrics and auxiliary metrics. All metrics provide specific values, trends, and spatial distributions, as detailed below:

[0145] Key output metrics:

[0146] Priority of existing area renovation: For each renovation node, output a priority score and corresponding weight coefficient, and clearly classify them: high priority is 80-100 points, medium priority is 50-79 points, and low priority is 0-49 points;

[0147] Urban spatial development potential: Output development potential index according to administrative divisions to clarify the development trend of each region;

[0148] Risks in the protection of historical buildings: Output the risk level within the core protection area and the construction control zone, as well as the risk triggering factors;

[0149] Resource allocation optimization suggestions: Output the resource allocation ratio for each existing upgrade node to support resource coordination. Auxiliary output indicators:

[0150] Spatiotemporal coupling bias: Outputs model prediction bias values ​​for different time periods and regions to support model parameter calibration;

[0151] Transformation effect prediction: Predict the environmental improvement rate and appearance preservation rate of each priority node after transformation;

[0152] Long-term development early warning: Outputs early warning information on potential risks to the protection of historical buildings and the lag in the renovation of existing buildings within a certain period of time in the future, as well as the early warning trigger threshold.

[0153] 2. Model prediction time scale

[0154] Short-term forecast: 1-3 years, focusing on prioritizing the renovation of existing areas and optimizing resource allocation to support the formulation of annual renovation plans; forecast accuracy ≥95%, with forecast results updated quarterly based on real-time data.

[0155] Medium-term forecast: 5-10 years, focusing on the potential for urban spatial development and the long-term risks of historical building protection, supporting the special plan for urban stock renewal; forecast accuracy ≥90%, and the forecast results are updated once a year in combination with annual calibration parameters.

[0156] Long-term forecast: 10-20 years, focusing on the overall urban development trend and the synergy between the protection of historical buildings, supporting the connection with the overall urban plan; forecast accuracy ≥85%, and the forecast results are updated every 3 years.

[0157] Forecast timeframes: Short-term forecasts are output quarterly, medium-term forecasts are output annually, and long-term forecasts are output on a 3-year cycle; all forecast results are labeled with the forecast time, data basis, and accuracy for easy decision-making reference.

[0158] 3. Model Validation Methods

[0159] A three-dimensional verification method combining historical data verification, real-time data verification, and field survey verification is employed to ensure the accuracy and reliability of the model output. The specific process is as follows:

[0160] Historical data verification:

[0161] a. Select historical urban data from the past 5 years, such as records of existing renovation, data on the protection of historical buildings, and data on urban development indicators, as the training set, input them into the spatiotemporal coupled system dynamics model, and output the prediction results;

[0162] b. Compare the model prediction results with the actual data from the past 5 years, and calculate the prediction deviation rate using the formula: Deviation rate = |Predicted value - Actual value| / Actual value × 100%;

[0163] c. Validation criteria: The prediction deviation rate of core indicators is ≤5%, and the prediction deviation rate of auxiliary indicators is ≤8%. If the criteria are not met, return to adjust the model parameters and re-validate until the criteria are met.

[0164] Real-time data verification

[0165] a. Each quarter, real-time data, such as data collection data and existing renovation progress data, are compared with the short-term prediction results of the model to calculate the deviation rate;

[0166] b. If the deviation rate exceeds the standard, quarterly parameter fine-tuning is triggered, and the model is re-verified after adjustment to ensure that the model adapts to urban development changes in real time.

[0167] Field research and verification:

[0168] a. Each year, select no fewer than 20 existing renovation sites and 10 historical buildings to conduct on-site surveys and collect actual data such as renovation effects and protection status;

[0169] b. Compare the field survey data with the model output results such as the predicted transformation effect and the early warning of protection risks to verify the matching degree between the model output and the actual situation;

[0170] c. Validation criteria: Matching degree ≥ 90%. If the standard is not met, adjust the model parameters such as weight coefficients and spatiotemporal coupling parameters based on the survey results, and re-validate after completing the annual calibration.

[0171] Verification Records: All verification processes are recorded in both written and electronic form, forming a "Verification Report of Dynamic Model of Spatiotemporal Coupled System", which is archived once a year to ensure that the verification process is reproducible and traceable.

[0172] The spatiotemporal coupling state equation of the dynamics of the spatiotemporal coupled system is as follows:

[0173] ,

[0174] in, The value range is continuous. This refers to the inflow rate, with the corresponding unit being annual. This is the outflow rate, with the corresponding unit being annual. The coefficient representing the influence of reachability on state variables ranges from 0.1 to 0.5. Spacetime coordinates The global integration degree at the location ranges from 0.3 to 1; This is the risk suppression coefficient for state variables, with a value ranging from 0.2 to 0.6. Spacetime coordinates The overall risk value at the location ranges from 0 to 1.

[0175] The inflow rate of the state variable includes the population migration flow rate.

[0176] ,

[0177] in, Spacetime coordinates The population is measured in units ranging from 100,000 to 1,000,000. Spacetime coordinates The number of job positions ranges from 5,000 to 500,000. This represents the natural growth rate, ranging from -5 to 15‰. The driving force of migration potential on population migration is the coefficient, with a value ranging from 0.001 to 0.01. Spacetime coordinates Local travel potential.

[0178] The aforementioned smart city-based urban planning decision-making feasibility analysis system is applicable to new regional planning or existing regional renewal and renovation scenarios. In the existing regional renewal and renovation scenario, weight coefficients are set for nodes surrounding historical buildings during spatial syntax calculation, and constraints on the protection scope of historical buildings are added during multi-objective particle swarm optimization.

[0179] 1. Boundary of the protected area for historical buildings

[0180] Core protection area: A circular area with a radius of 20 meters extending outward from the outline of the historical building, or a rectangular area extending 15 meters × 15 meters. Irregular buildings shall be fitted with polygons. The core protection area shall be ≥500㎡. No destructive alterations are allowed within this area, and only necessary repair works are permitted.

[0181] Construction Control Zone: An area with a radius of 50 meters extending outward from the boundary of the core protection area serves as a buffer zone for the core protection area; any renovation projects within this area must comply with the requirements for the management of the appearance of historical buildings and must not damage the visual integrity of the historical buildings.

[0182] Boundary confirmation method: The precise coordinates of historical buildings, accurate to six decimal places, are imported through the GIS platform to automatically generate the boundary lines of the core protection area and construction control zone. These boundaries are then manually verified in conjunction with on-site surveys and confirmed before being entered into the system as the basis for setting weight coefficients.

[0183] 2. Weighting coefficient quantification standard

[0184] The weighting coefficient ranges from [0,1]. The larger the value, the higher the priority and the stricter the control requirements for the node in the existing infrastructure upgrade and renovation. The weighting coefficient is calculated by weighting the "distance factor + node type factor", and the specific rules are as follows:

[0185] Key influencing factor: Distance factor quantification

[0186] Nodes within the core protection area: distance factor = 1.0;

[0187] For nodes within the control zone: distance factor = 0.7-0.9; the closer to the boundary of the core protection area, the larger the value; for example, take 0.9 within 10 meters of the boundary, and 0.7 within 50 meters of the boundary;

[0188] Nodes within 100 meters of historical buildings outside the construction control zone: distance factor = 0.4-0.6;

[0189] Nodes within 100 meters of historical buildings: distance factor = 0.1-0.3;

[0190] Ordinary nodes without the influence of historical buildings: distance factor = 0.0, meaning they are not included in the weighting considerations related to historical buildings.

[0191] Correction Factor: Node Type Factor Quantification

[0192] Municipal infrastructure nodes such as water supply, power supply, and drainage: Node type factor = 1.1;

[0193] Traffic nodes such as intersections and pedestrian entrances / exits: Node type factor = 1.0;

[0194] Green spaces and recreational areas: Node type factor = 0.9;

[0195] Temporary building node: Node type factor = 0.5.

[0196] Final weighting coefficient calculation: Weighting coefficient = distance factor × node type factor, the result is rounded to two decimal places; for example: municipal infrastructure nodes within the core protection area, weighting coefficient = 1.0 × 1.1 = 1.1; traffic nodes within the construction control zone and 20 meters from the boundary, weighting coefficient = 0.8 × 1.0 = 0.8.

[0197] Weighting coefficient update: The distance factor and node type factor are calibrated once a year based on the renovation of historical buildings and planning adjustments to ensure that the quantitative standards match the actual renovation needs.

[0198] The above description is only a preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A feasibility analysis system for urban planning decisions based on smart cities, characterized in that, include: The data acquisition module is used to dynamically collect urban basic spatial data, dynamic monitoring data, and industry and people's livelihood data, and to perform preprocessing. Algorithm processing module, the algorithm processing module includes: The spatial syntax unit is used to output the global integration degree (GI) and local traversal potential (LTP) based on spatial syntax as the underlying algorithm, serving as the spatial constraint basis for resource allocation. The core function of the spatial syntax is to parse the topological relationships of urban road networks and public spaces, quantify spatial accessibility and usage potential, and output the global integration degree. and local travel potential As the spatial constraint basis for resource allocation; global integration Through formula calculate, Where n is the total number of spatial nodes in the study area, and its value ranges from 1 to 2. ; It is a logarithmic function with base 2; MD is the average shortest path length of the nodes, with a value ranging from 50 to 2000m; Local travel potential Through formula calculate, in, For nodes The set of nodes within a 500m radius; For nodes To the node The straight-line distance ranges from 10 to 500 m. For nodes The average daily pedestrian flow, with a value range of [value missing]. Number of people; For nodes The corresponding plot area, with a range of values, is as follows: ; The multi-objective particle swarm optimization unit is used to optimize the allocation of urban land, energy and public service facilities resources with the global integration degree (GI) and local traversal potential (LTP) as hard constraints, aiming to maximize economic benefits, minimize carbon emissions and maximize public satisfaction, and output resource allocation schemes. An improved Bayesian network unit is used to integrate the resource allocation scheme with real-time monitoring data from the Internet of Things, and to calculate the probability of occurrence of three core risks—floods, traffic congestion, and air pollution—using a dynamic conditional probability formula. The spatiotemporal coupled system dynamics unit is used to integrate the global integration degree (GI) and local traversal potential (LTP) output by the spatial syntax unit, the resource allocation scheme output by the multi-objective particle swarm optimization unit, and the occurrence probability output by the improved Bayesian network unit, and to construct a spatiotemporal dimension urban development model through the spatiotemporal coupled state equation. The decision output module is used to generate planning schemes based on the urban development model constructed by the algorithm processing module. The dynamic optimization module includes a dual-cycle regular update unit and a threshold-triggered dynamic optimization unit, which are used to dynamically adjust and iteratively optimize the planning scheme.

2. The urban planning decision-making feasibility analysis system based on smart cities according to claim 1, characterized in that, The aforementioned feasibility analysis system for urban planning decisions based on smart cities relies on a smart city sensing network, which includes the Internet of Things, remote sensing, GIS, and a big data platform.

3. The urban planning decision-making feasibility analysis system based on smart cities according to claim 1, characterized in that, The particle position update formula for the multi-objective particle swarm optimization algorithm is: The particle velocity update formula is: ， The optimization criteria for the multi-objective particle swarm optimization algorithm include: Maximize economic benefits : Minimize carbon emissions : Maximize public satisfaction : in, For the first The particle in the first The position of the dimension, with a value range of 0 to 1; The particle velocity is represented by a value ranging from -0.5 to 0.

5. This is the inertia weight, with a value ranging from 0.4 to 0.9; For individual learning factors; As a whole learning factor; For each individual; The result is a random number for the whole system; The optimal position for an individual particle, with a value ranging from 0 to 1; This represents the globally optimal position for the population, and its value ranges from 0 to 1. The area allocated for the kth type of resource is 0.1 to 100 km². Let k be the energy consumption coefficient for land use category k. For the first The economic output coefficient of land use category ranges from 0.5 to 1 billion yuan / km². The development intensity coefficient for the kth land use category ranges from 0.6 to 1.

2. The carbon emission coefficient for land use category k ranges from 0.1 to 50,000 tons / km². For the first The average service distance for public service facilities ranges from 200 to 1500 meters. For the first The usage frequency weight of facility type ranges from 0.5 to 2; The constraints of the multi-objective particle swarm optimization algorithm include: Global integration constraints: 、 Local travel potential constraints: Total resource constraints: ， in, This refers to the global integration degree of node i; For the local traversal potential of node i; This represents the minimum global integration level, ranging from 0.3 to 0.

6. This represents the maximum local travel potential, with a range of values. It is the total planned land area, ranging from 10 to 1000 km².

4. The urban planning decision-making feasibility analysis system based on smart cities according to claim 1, characterized in that, The dynamic conditional probability of the improved Bayesian network unit is expressed by the formula... calculate, in, The dynamic conditional probability value represents the probability at n input nodes. Given a given state, the probability that the target node / intermediate node Y is in a certain state; For node variables; For input node variables; For input node The dynamic calibration weights are derived from the improved Bayesian network unit dynamic weight calibration formula. It is dynamically updated according to the multi-objective particle swarm algorithm scheme; For input node The conditional probability of a node Y in a certain state comes from the improved Bayesian network unit conditional probability table; The prior probability of node Y is set by historical data and dynamically fine-tuned during iterative calibration. The chain multiplication symbol represents the product of n input nodes. Perform a series of multiplication operations; The summation symbol represents the summation of the product of the two states of node Y. It is used for normalization to ensure that the output probability value is within the range of 0 to 1.

5. The system according to claim 1, characterized in that, The spatiotemporal coupling state equation of the dynamics of the spatiotemporal coupled system is: in, ; This refers to the inflow rate, with the corresponding unit being annual. Rate of change over time; This is the outflow rate, with the corresponding unit being annual. The coefficient representing the influence of reachability on state variables ranges from 0.1 to 0.

5. Spacetime coordinates The global integration degree at the location ranges from 0.3 to 1; This is the risk suppression coefficient for state variables, with a value ranging from 0.2 to 0.

6. Spacetime coordinates The comprehensive risk value at the location ranges from 0 to 1; the inflow rate of the state variable includes the population migration flow rate. : ， in, Spacetime coordinates The population is measured in units ranging from 100,000 to 1,000,000. Spacetime coordinates The number of job positions ranges from 5,000 to 500,000. This represents the natural growth rate, ranging from -5 to 15‰. The driving force of migration potential on population migration is the coefficient, with a value ranging from 0.001 to 0.

01. Spacetime coordinates Local travel potential.

6. The urban planning decision-making feasibility analysis system based on smart cities according to claim 1, characterized in that, The urban basic spatial data of the data acquisition module includes road network topology data, land use status data, topography, building features and administrative division data; the dynamic monitoring data includes pedestrian flow, vehicle flow, precipitation and air quality data; the industry and people's livelihood data includes economic output coefficient, carbon emission coefficient and public service facility usage frequency weight; the preprocessing includes unifying data units and removing outliers.

7. The urban planning decision-making feasibility analysis system based on smart cities according to claim 1, characterized in that, The planning scheme generated by the decision output module includes land use layout, energy configuration, and public service facility distribution parameters.

8. The urban planning decision-making feasibility analysis system based on smart cities according to claim 1, characterized in that, The dual-cycle routine update unit is used to perform quarterly real-time monitoring updates and annual parameter calibration updates; the threshold-triggered dynamic optimization unit is used to automatically trigger the multi-objective particle swarm optimization unit to reconfigure resources when the real-time monitoring data exceeds a preset threshold range; the threshold-triggered dynamic optimization unit is configured to perform the following optimization steps: Data freeze and scene locking: Lock the current planning scheme and monitoring data snapshot at the trigger time; Relaxing constraints: Under the premise of meeting the hard constraints of global integration (GI) and local transit potential (LTP), the total planned land area constraint is relaxed; Multi-objective particle swarm optimization re-optimization: Using the current planning scheme as the initial population, re-run the multi-objective particle swarm optimization unit to generate alternative adjustment schemes; Risk assessment filtering: The improved Bayesian network unit is used to calculate the comprehensive risk value of flooding, traffic congestion and air pollution of the alternative adjustment schemes, and high-risk schemes with a comprehensive risk value R>0.7 are eliminated; Solution optimization and output: Based on the Pareto frontier, the optimal solution is selected and the parameters of land use layout, energy configuration and distribution of public service facilities are updated; Effectiveness of the plan and monitoring in the next cycle: Output the adjusted planning plan and enter the monitoring cycle for the next quarter.

9. The urban planning decision-making feasibility analysis system based on smart cities according to any one of claims 1 to 8, characterized in that, The aforementioned smart city-based urban planning decision-making feasibility analysis system is applicable to new regional planning or existing regional renewal and renovation scenarios. In the existing regional renewal and renovation scenario, weight coefficients are set for nodes surrounding historical buildings during spatial syntax calculation, and constraints on the protection scope of historical buildings are added during multi-objective particle swarm optimization.

Images