Water conservancy flood prevention scheduling method and system, storage medium and program product

Abstract

The invention discloses a water conservancy flood prevention scheduling method and system, a storage medium and a program product, and relates to the field of data processing systems specially suitable for the prediction purpose, and the method comprises the steps: obtaining dynamic social data; the dynamic social data comprises media data and mobile phone signaling data; identifying the media data to obtain semantic features or visual features so as to position a plurality of real-time water accumulation points; generating real-time waterlogging distribution information according to the feature information of the plurality of real-time water accumulation points; determining real-time population thermodynamic information according to the mobile phone signaling data; generating a comprehensive risk assessment result representing the safety risk levels of people in different areas in the downstream city; and optimizing the comprehensive risk assessment result by taking minimization of the personnel safety risk of the high-risk area as a decision-making target, and generating or adjusting a flood prevention scheduling instruction to control the discharge flow or flood storage capacity of the upstream water conservancy facility. By implementing the application, the scheduling decision of the upstream water conservancy facility can be optimized, and the accuracy of the flood prevention scheduling strategy is ensured.

Description

Technical Field

[0001] This application relates to the field of data processing systems specifically designed for forecasting purposes, and more particularly to a water conservancy flood control scheduling method, system, storage medium, and program product. Background Technology

[0002] Flood control and water conservancy scheduling is a crucial link in the national disaster prevention and mitigation system. It involves the scientific and coordinated regulation of water conservancy projects such as reservoirs, dams, and pumping stations to reduce flood peaks, control flood progression, and ensure the safety of downstream areas. Currently, many large-scale water conservancy projects are directly connected to densely populated and economically developed urban areas downstream. Therefore, modern flood control and water conservancy scheduling requires precise assessment of the impact of scheduling actions on downstream cities, paying particular attention to urban flooding risks and resident safety.

[0003] In related technologies, monitoring data from key nodes in downstream cities is integrated into the water conservancy flood control dispatch system. This solution pre-deploys specialized IoT sensors, such as water level gauges and flow meters, in areas prone to urban flooding, including underpasses, low-lying areas, and core drainage networks. The physical parameters collected by these sensors are transmitted in real time to the water conservancy flood control dispatch center. Based on this data uploaded from fixed points and combined with upstream water conditions, the dispatch center assesses the carrying capacity of downstream cities and uses this information as one of the reference points for making or adjusting dispatch decisions, such as adjusting the discharge flow from upstream reservoirs.

[0004] However, the relevant technologies rely entirely on a limited number of spatially fixed physical sensors to perceive risks in downstream cities. Due to the complex urban geography, fixed sensor networks have numerous monitoring blind spots and lack awareness of population safety, resulting in low accuracy of flood control and dispatch strategies. Summary of the Invention

[0005] This application provides a water conservancy flood control scheduling method, system, storage medium, and program product, which are used to optimize the scheduling decisions of upstream water conservancy facilities and ensure the accuracy of flood control scheduling strategies.

[0006] Firstly, this application provides a water conservancy flood control scheduling method, applied to a flood control scheduling system. The method includes: during a flood control scheduling response, acquiring dynamic social data from the geographical area of ​​a downstream city, containing spatiotemporal location information; this dynamic social data includes at least media data actively posted by users on social media platforms and mobile signaling data provided by mobile communication network operators; identifying the media data to obtain semantic or visual features associated with urban flooding, thereby locating multiple real-time waterlogging points; and generating real-time urban flooding distribution information characterizing the spatial distribution of urban flooding in the downstream city based on the feature information of the multiple real-time waterlogging points. Based on mobile phone signaling data, the number of signaling requests within a preset geographic grid is determined, generating real-time population heat map information representing the spatial distribution and density of the population in downstream cities. Based on real-time urban flooding distribution information and real-time population heat map information, the scope of urban flooding impact and population heat map distribution are spatially overlaid and analyzed to generate a comprehensive risk assessment result representing the safety risk level of people in different areas of downstream cities. With minimizing the safety risk of people in high-risk areas as the decision-making objective, the comprehensive risk assessment result is optimized and analyzed to generate or adjust flood control dispatch instructions applied to upstream water conservancy facilities to control the downstream flow or flood storage capacity of upstream water conservancy facilities.

[0007] In the above embodiments, the flood control dispatch system can acquire social media data and mobile phone signaling data in real time, accurately locate waterlogging points through semantic and visual analysis, and perform spatial overlay analysis combined with population thermal distribution to accurately assess the safety risks to people in different areas of the city. Based on the assessment results, the system can adjust the downstream flow of upstream water conservancy facilities or the flood storage capacity to minimize the safety risks to people in high-risk areas, achieving comprehensive perception of urban flooding and population distribution, and improving the accuracy and timeliness of flood control dispatch decisions.

[0008] In conjunction with some embodiments of the first aspect, in some embodiments, the step of acquiring dynamic social data containing spatiotemporal location information from the geographical range of downstream cities during flood control dispatch response specifically includes: establishing a social media data collection task within a preset time window; filtering social media data with geographic tags according to the geographical boundaries of downstream cities to obtain media data; acquiring base station signaling data from downstream cities; dividing the base station signaling data according to the base station location into preset geographic grids; counting the number of signaling signals in each grid; extracting the roaming status of user devices from the base station signaling data; identifying user devices moving across grids; generating dynamic displacement data representing user movement trajectories; identifying the gathering and evacuation of people based on the dynamic displacement data; updating the number of signaling signals in the geographic grids to obtain mobile phone signaling data; and integrating media data and mobile phone signaling data into dynamic social data.

[0009] In the above embodiments, the flood control dispatch system has established a complete dynamic social data collection and processing mechanism. By collecting social media data within a set time window and combining it with the gridded processing of base station signaling data, it can not only obtain static population distribution information, but also identify personnel movement trajectories by analyzing the roaming status of user devices, and grasp the situation of population gathering and evacuation in real time, enabling the system to comprehensively grasp the urban flooding and population distribution.

[0010] In conjunction with some embodiments of the first aspect, in some embodiments, the step of identifying media data to obtain semantic or visual features associated with urban flooding in order to locate multiple real-time waterlogging points specifically includes: extracting text content containing preset waterlogging keywords from the media data, and locating text candidate waterlogging points based on the geographical location description information in the text content; filtering image content containing waterlogging scenes from the media data, and locating image candidate waterlogging points based on the geographical marker information in the image content; clustering the text candidate waterlogging points and image candidate waterlogging points according to geographical location, and removing waterlogging points with inconsistent timestamps or location deviations exceeding a preset deviation threshold to obtain real-time candidate waterlogging points; calculating the confidence level of the real-time candidate waterlogging points based on distribution density and reporting time, and filtering out real-time candidate waterlogging points with confidence levels higher than a preset confidence threshold to obtain multiple real-time waterlogging points.

[0011] In the above embodiments, the flood control dispatch system has established an intelligent mechanism for identifying and verifying water accumulation points. By extracting keywords from social media text content and recognizing scenes from image content, combined with geographic location information for multi-source data cross-validation, and through spatiotemporal clustering analysis and confidence assessment, the probability of false alarms and missed alarms is effectively reduced.

[0012] In conjunction with some embodiments of the first aspect, in some embodiments, before the step of spatially overlaying and analyzing the scope of waterlogging impact and population thermal distribution based on real-time waterlogging distribution information and real-time population thermal information to generate a comprehensive risk assessment result characterizing the safety risk level of people in different areas of the downstream city, the method further includes: determining the maximum influence radius of each waterlogging point under different rainfall intensities based on historical waterlogging records to obtain the waterlogging impact area; determining the inundation depth level within the waterlogging impact area based on road slope and drainage facility distribution; identifying major traffic arteries and densely populated areas within the waterlogging impact area based on building distribution and road connectivity, and determining the difficulty of passage for people in the waterlogging impact area under different inundation depth levels; and determining the danger level of different waterlogging points based on the difficulty of passage for people.

[0013] In the above embodiments, the flood control dispatch system establishes a scientific waterlogging impact assessment system. By analyzing multiple factors such as historical flooding records, road slope, and drainage facility distribution, the system can accurately predict the impact range of waterlogging points and assess the difficulty of personnel passage by combining building distribution and road connectivity. This enables the system to comprehensively assess the safety risks of different areas, making flood control dispatch decisions more precise.

[0014] In conjunction with some embodiments of the first aspect, in some embodiments, the step of determining the maximum influence radius of each waterlogging point under different rainfall intensities based on historical flooding records to obtain the waterlogging-affected area specifically includes: extracting the geographical coordinates, corresponding rainfall intensity values, and measured influence radii of each historical waterlogging point from the historical flooding record database, and constructing a mapping model between rainfall intensity and waterlogging influence radius; calculating the surface runoff path and catchment area around each historical waterlogging point based on the digital elevation model and drainage facility distribution data, and determining the direction of waterlogging diffusion; calculating the predicted influence radius of each waterlogging point under the current rainfall intensity based on the mapping model between rainfall intensity and waterlogging influence radius, combined with the surface runoff path and catchment area; and generating the boundary coordinates of the influence area of ​​each waterlogging point based on the predicted influence radius and the direction of waterlogging diffusion, thereby obtaining the waterlogging-affected area.

[0015] In the above embodiments, the flood control dispatch system achieves accurate impact prediction based on historical data. By establishing a mapping model between rainfall intensity and the radius of water accumulation, and combining digital elevation models and drainage facility distribution data to analyze surface runoff characteristics, the system can accurately predict the impact range and diffusion direction of water accumulation points, improving the accuracy of the system's prediction of urban flooding development trends.

[0016] In conjunction with some embodiments of the first aspect, in some embodiments, the step of optimizing and analyzing the comprehensive risk assessment results to generate or adjust flood control dispatch instructions applied to upstream water conservancy facilities with the decision objective of minimizing the safety risks to personnel in high-risk areas includes: extracting the geographical range and population density data of each high-risk area from the comprehensive risk assessment results, and determining available evacuation channels based on the real-time traffic status of the current road network; obtaining the current water storage level, discharge flow, and gate opening parameters of upstream water conservancy facilities, and establishing an implementation impact model of the discharge flow on the downstream water level; predicting the risk of aggravated waterlogging under different discharge flows based on the implementation impact model, and generating graded dispatch time thresholds based on the traffic capacity of available evacuation channels; generating multiple flood control dispatch plans, evaluating the matching degree between the execution time of each plan and the graded dispatch time thresholds, and determining the optimal dispatch plan that minimizes the safety risks to personnel in high-risk areas; and converting the optimal dispatch plan into flood control dispatch instructions.

[0017] In the above embodiments, the flood control dispatch system establishes a dynamically optimized dispatch decision-making mechanism. By analyzing the geographical characteristics and population distribution of high-risk areas, and assessing evacuation capacity in conjunction with road network traffic conditions, the system can generate multiple flood control dispatch plans and optimize and filter them, ensuring the scientific nature and operability of flood control dispatch instructions.

[0018] In conjunction with some embodiments of the first aspect, in some embodiments, after the step of converting the optimal scheduling plan into flood control scheduling instructions, the method further includes: acquiring historical scheduling data and scheduling effect evaluation results of upstream water conservancy facilities, and establishing an effect evaluation model for the scheduling instructions; verifying the feasibility of the flood control scheduling instructions based on the effect evaluation model, and generating scheduling parameters including adjustment target values ​​and execution timing; establishing a feedback control strategy for gate opening and discharge based on the scheduling parameters, and implementing the feedback control strategy; monitoring downstream water level changes and the scope of waterlogging impact in real time, obtaining real-time monitoring results, and adjusting the control parameters of the feedback control strategy based on the real-time monitoring results.

[0019] In the above embodiments, the flood control dispatch system has established a complete dispatch effect evaluation and feedback mechanism. By analyzing historical dispatch data to build an effect evaluation model and dynamically adjusting control parameters in conjunction with real-time monitoring results, the system can continuously optimize dispatch strategies and ensure the continuous improvement of the effectiveness of flood control dispatch instructions.

[0020] In a second aspect, embodiments of this application provide a flood control dispatch system, which includes: one or more processors and a memory; the memory is coupled to the one or more processors, and the memory is used to store computer program code, which includes computer instructions, and the one or more processors call the computer instructions to cause the flood control dispatch system to perform the method described in the first aspect and any possible implementation thereof.

[0021] Thirdly, embodiments of this application provide a computer program product containing instructions that, when the computer program product is run on a flood control dispatch system, cause the flood control dispatch system to execute the method described in the first aspect and any possible implementation thereof.

[0022] Fourthly, embodiments of this application provide a computer-readable storage medium including instructions that, when executed on a flood control dispatch system, cause the flood control dispatch system to perform the method described in the first aspect and any possible implementation thereof.

[0023] Understandably, the flood control dispatch system provided in the second aspect, the computer program product provided in the third aspect, and the computer storage medium provided in the fourth aspect are all used to execute the methods provided in the embodiments of this application. Therefore, the beneficial effects they can achieve can be referred to the beneficial effects in the corresponding methods, and will not be repeated here.

[0024] One or more technical solutions provided in the embodiments of this application have at least the following technical effects or advantages: 1. By employing an intelligent flood control dispatching method based on dynamic social data, and acquiring and analyzing social media and mobile phone signaling data, the system overcomes the limitations of traditional fixed monitoring points, achieving comprehensive perception of urban flooding and population distribution. Through semantic and visual analysis of media data, combined with gridded processing of mobile phone signaling data, the system can accurately locate waterlogged areas and grasp real-time population distribution. This effectively solves the problems of numerous monitoring blind spots and insufficient perception of population safety in existing technologies. The system can accurately assess the safety risks in different areas of the city and optimize dispatching decisions accordingly.

[0025] 2. By employing a multi-layered flood impact assessment method, comprehensively analyzing multi-dimensional data such as historical flooding records, topographic features, and building distribution, the system can scientifically assess the impact range and risk level of flooding points. Through analysis of road slope and drainage facility distribution, combined with building distribution and road connectivity, the system can accurately assess the difficulty of pedestrian access in different areas. This effectively solves the problems of one-sided and unsystematic risk assessment in existing technologies, achieving accurate prediction of flood impact and scientific classification of risk levels.

[0026] 3. By employing a risk-minimizing-based scheduling decision optimization method, the system can generate and optimize flood control scheduling plans by analyzing the characteristics of high-risk areas and the status of evacuation routes. Combined with real-time parameters of upstream water conservancy facilities and downstream water level impact models, the system can accurately predict the effects of different scheduling schemes. This effectively solves the problems of experience-based scheduling decisions and insufficient optimization in existing technologies, achieving the scientific generation and dynamic optimization of scheduling plans. Attached Figure Description

[0027] Figure 1 This is a flowchart illustrating a water conservancy flood control scheduling method in an embodiment of this application; Figure 2 This is another flowchart illustrating the water conservancy flood control scheduling method in the embodiments of this application; Figure 3 This is a schematic diagram of the physical device structure of a flood control dispatch system in the embodiments of this application. Detailed Implementation

[0028] The terminology used in the following embodiments of this application is for the purpose of describing particular embodiments only and is not intended to be limiting of this application. As used in the specification of this application, the singular expressions “a,” “an,” “the,” “the,” and “this” are intended to include the plural expressions as well, unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used in this application refers to any or all possible combinations including one or more of the listed items.

[0029] Hereinafter, the terms "first" and "second" are used for descriptive purposes only and should not be construed as implying or suggesting relative importance or implicitly indicating the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature, and in the description of the embodiments of this application, unless otherwise stated, "multiple" means two or more.

[0030] To facilitate understanding, the application scenarios of the embodiments of this application are described below.

[0031] In urban flood control and disaster reduction systems, water conservancy flood control scheduling is a crucial link in reducing flood risks by controlling the downstream flow and flood storage capacity of upstream water conservancy facilities (such as reservoirs and dams). Traditional scheduling methods mainly rely on monitoring data from fixed sensors (such as water level gauges and flow meters), which suffers from numerous monitoring blind spots and limited coverage. The method proposed in this application introduces dynamic social data (including social media data and mobile phone signaling data) as a supplementary data source. It locates waterlogging points through semantic analysis and visual feature recognition technology, and generates population heat maps using signaling data to achieve comprehensive perception of urban flooding and population distribution. Simultaneously, it uses a digital elevation model to analyze terrain features, and combines the distribution of drainage facilities and historical flooding records to establish a waterlogging impact prediction model, scientifically assessing the risk levels of different areas. Based on a multi-objective optimization algorithm, the system can generate an optimal scheduling plan that ensures both flood control effectiveness and minimizes personnel safety risks. This intelligent flood control scheduling method overcomes the limitations of traditional fixed monitoring points, improves the accuracy and timeliness of scheduling decisions, and is of great significance for ensuring urban safety.

[0032] To facilitate understanding, the method provided in this implementation will be described in detail below, using the above scenario as an example. Please refer to [link / reference]. Figure 1 This is a flowchart illustrating a water conservancy flood control scheduling method in an embodiment of this application.

[0033] S101. During the flood control dispatch response, acquire dynamic social data from the geographical range of downstream cities, including spatiotemporal location information.

[0034] Among them, the flood control dispatch response period refers to the emergency response period initiated when a flood disaster or early warning occurs; dynamic social data refers to multi-source information that reflects the real-time status of the city, including social media data and mobile phone signaling data; spatiotemporal location information refers to the geographical coordinates and timestamps of the data; downstream urban geographical range refers to the urban area affected by the dispatch of upstream water conservancy facilities, which is delineated by administrative boundaries or hydrological boundaries.

[0035] The flood control dispatch system initiates this step upon receiving a flood warning or when urban flooding occurs. Specifically, the system first determines the geographical boundaries of downstream cities and establishes a data collection task. For social media data, the system acquires user-posted content with location tags within a specified geographical area in real time; for mobile signaling data, the system obtains user terminal data at the base station level from mobile communication operators. The system performs timestamp verification and spatial location verification on the acquired data to ensure its timeliness and geographical accuracy, and integrates the verified data into dynamic social data in a standard format.

[0036] In some embodiments, the collection and processing of dynamic social data can be achieved in multiple ways: Optionally, the flood control dispatch system can establish a social media API interface to obtain geotagged content from platforms such as Weibo and Douyin in real time, extract key information through natural language processing, and determine the accurate location using geocoding technology; Optionally, the flood control dispatch system can establish a data sharing mechanism with telecommunications operators to periodically obtain anonymized user location data, divide geographical grids using spatial clustering algorithms, and calculate the real-time population distribution within the grids. It is understood that other data sources and processing methods can also be used to acquire and analyze dynamic social data, and this is not limited here. In practical applications, the system also needs to consider data privacy protection requirements and anonymize personal information.

[0037] S102. Identify media data to obtain semantic or visual features associated with urban flooding in order to locate multiple real-time water accumulation points.

[0038] Among them, media data refers to information content containing text, images, or videos published by users on social media platforms; semantic features represent keywords, location descriptions, and degree expressions related to water accumulation in the text content; visual features refer to feature information reflecting water accumulation scenes in images or videos, including features such as water surface, depth, and range; and real-time water accumulation points represent the verified specific geographical locations where water accumulation currently exists.

[0039] Upon acquiring media data, the flood control dispatch system immediately initiates identification and analysis. Specifically, the system first performs natural language processing on the text content, extracting content containing keywords such as "waterlogging" and "urban flooding," and identifying geographical location information and descriptions of waterlogging levels. For image content, the system uses a deep learning model to identify waterlogging scenes, extracting features such as water surface area and depth, and combining this with geographic marker information in the image. The system cross-validates the results of text and image analysis, merges reports from similar locations using spatiotemporal clustering methods, calculates the credibility of each candidate waterlogging point, and ultimately determines the location and feature information of the real-time waterlogging point.

[0040] In some embodiments, intelligent identification of media data can be achieved in multiple ways: Optionally, the flood control dispatch system can establish a multimodal feature extraction model based on deep learning, simultaneously processing text semantics and image visual information, and focusing on key features through an attention mechanism to achieve accurate identification and location of waterlogging scenes; Optionally, the flood control dispatch system can construct a knowledge graph to assist analysis, integrating prior knowledge such as historical waterlogging point information, terrain features, and drainage facility distribution to improve the accuracy of waterlogging point identification. It is understood that other artificial intelligence technologies can also be used to achieve feature extraction of media data and waterlogging point identification, which is not limited here.

[0041] The main challenge in implementing this step is identifying false information. To address this, the flood control dispatch system has established a multi-dimensional information authenticity verification mechanism. First, the system analyzes the publisher's historical credibility, assigning higher weight to users who are consistently active and publish accurate information. Second, the system performs consistency checks on multiple reports from the same location, analyzing the continuity of the reports' timeline and the similarity of their content. Finally, the system compares the identification results with surrounding terrain features and historical water accumulation records, eliminating abnormal reports that do not conform to objective laws. For example, if a large number of similar water accumulation reports suddenly appear at a location, and the terrain conditions at that location are not prone to water accumulation, the system will mark it as suspicious information and require manual verification.

[0042] S103. Based on the feature information of multiple real-time water accumulation points, generate real-time waterlogging distribution information that characterizes the spatial distribution of urban waterlogging in the downstream area.

[0043] Among them, the feature information represents attribute data such as the location coordinates of the water accumulation point, the water depth, the range and the duration of the water accumulation; the real-time urban flooding distribution information refers to the digital description of the spatial distribution of urban water accumulation, including the location of the water accumulation point, the range of impact and the severity.

[0044] After identifying multiple real-time waterlogging points, the flood control dispatch system performs spatial analysis and distribution projection. Specifically, the system first calculates the direction of water spread and the potential impact range based on the location and characteristics of each waterlogging point, combined with digital elevation models and drainage network distribution data. Then, considering surface runoff paths and catchment area characteristics, the system analyzes the mutual influence and correlation between waterlogging points. Through spatial interpolation and hydrodynamic models, the system generates continuous urban flooding distribution raster data covering the entire city, with each raster cell containing water depth and risk level information. Finally, the system analyzes the development trend of urban flooding based on time-series data and predicts short-term changes.

[0045] In some embodiments, the generation of urban flooding distribution information can be achieved in multiple ways: Optionally, the flood control dispatch system can employ GIS-based spatial analysis methods, combining spatial data such as topography, drainage, and land use to construct a two-dimensional diffusion model, simulate the dynamic changes in the water accumulation area, and generate a refined raster distribution map; Optionally, the flood control dispatch system can use deep learning methods to train a spatial prediction model based on historical urban flooding data, enabling rapid extrapolation from discrete water accumulation points to continuous distribution. It is understood that other spatial analysis and simulation prediction methods can also be used to generate urban flooding distribution information, and this is not limited here.

[0046] The main challenge in implementing this step is insufficient spatial interpolation accuracy. To address this, the flood control dispatch system employs an adaptive interpolation strategy based on multi-source data fusion. The system first dynamically adjusts interpolation parameters and the search radius according to the spatial distribution density of waterlogged points. In sparse data areas, the system incorporates historical waterlogging patterns and terrain features for constrained interpolation; in dense data areas, it uses a local weighting method to improve interpolation accuracy. Simultaneously, the system continuously corrects and optimizes the interpolation results through real-time monitoring data. For example, when a new waterlogged point is reported, the system quickly updates the distribution information of the surrounding area to ensure the timeliness and accuracy of the flood distribution map.

[0047] S104. Determine the number of signaling messages within the preset geographic grid based on mobile phone signaling data, and generate real-time population heat map information representing the spatial distribution and population density of the downstream city population.

[0048] Among them, mobile signaling data represents the location information of user terminals collected by mobile communication base stations; preset geographic grid refers to the regular grid units that divide urban areas according to unified standards; signaling quantity refers to the number of active user terminals in the grid within a specific time period; real-time population heat information represents a dynamic data layer reflecting population density and distribution, usually presented in the form of a heat map.

[0049] After receiving base station signaling data, the flood control dispatch system performs population distribution analysis. Specifically, the system first divides the urban area into equally sized geographical grids, typically 200 meters × 200 meters. For each grid, the system counts the number of active users within a specific time window (e.g., 15 minutes) and spatially allocates them based on base station coverage. Considering user mobility, the system tracks signaling roaming records to identify population flow trends. Through time-series analysis, the system calculates the population change rate for each grid and, combined with building distribution and functional attributes, corrects the population density estimation results. Finally, the system generates a heat map representing population distribution and continuously updates the dynamic changes.

[0050] In some embodiments, population thermal information can be generated in various ways: Optionally, the flood control dispatch system can adopt a multi-level grid partitioning strategy, using smaller grid cells in densely populated areas to improve accuracy and larger grid cells in sparsely populated areas to reduce computation, balancing accuracy and efficiency through adaptive algorithms; Optionally, the flood control dispatch system can combine urban functional zoning and temporal characteristics to establish a dynamic weighting model, adjusting the density calculation method according to the population activity characteristics of different time periods and regions. It is understood that other spatial statistics and population analysis methods can also be used to generate thermal information, and this is not limited here.

[0051] S105. Based on real-time urban flooding distribution information and real-time population thermal information, spatial overlay analysis is performed on the impact range of urban flooding and population thermal distribution to generate a comprehensive risk assessment result that characterizes the safety risk level of people in different areas of the downstream city.

[0052] Among them, the scope of waterlogging impact refers to the geographical area affected by the water accumulation point, including the directly waterlogged area and the indirectly affected area; the personnel safety risk level indicates the degree of threat to people in a specific area from waterlogging; the comprehensive risk assessment result refers to the comprehensive regional risk assessment obtained after considering the degree of waterlogging and population density.

[0053] After obtaining data on urban flooding distribution and population heat maps, the flood control dispatch system conducts a risk assessment. Specifically, the system first calculates the impact range of each waterlogging point, considering factors such as terrain slope, road network, and building distribution, and delineates directly affected areas and potential impact areas. Then, the system spatially overlays the impact range layer with the population heat map layer to calculate the population size and density within different impact areas. The system further analyzes the distribution of important facilities, accessibility of evacuation routes, and population vulnerability within the area, comprehensively assessing personnel safety risks. Based on multi-factor weighted calculations, the system classifies urban areas into different risk levels and dynamically updates the risk assessment results.

[0054] In some embodiments, risk assessment can be achieved in multiple ways: Optionally, the flood control dispatch system can establish a multidimensional risk assessment model based on the analytic hierarchy process (AHP), assigning different weights to factors such as water depth, population density, building type, and road accessibility according to their importance, and obtaining a comprehensive risk score through matrix operations; Optionally, the flood control dispatch system can employ machine learning methods to train a risk prediction model based on historical disaster data, enabling rapid assessment of risk levels under different scenarios. It is understood that other risk analysis methods can also be used to achieve comprehensive risk assessment, and this is not limited here.

[0055] S106. With the goal of minimizing the safety risks to personnel in high-risk areas, optimize and analyze the comprehensive risk assessment results, and generate or adjust flood control dispatch instructions applied to upstream water conservancy facilities to control the downstream flow or flood storage capacity of upstream water conservancy facilities.

[0056] Among them, high-risk areas refer to urban areas with a higher risk level in the comprehensive risk assessment results; personnel safety risk refers to the degree of threat to life and property that people in the area may suffer; flood control dispatch instructions refer to the specific operational requirements for upstream water conservancy facilities; downstream discharge flow and flood storage capacity refer to the amount of water discharged downstream by water conservancy facilities and the volume that can be used for water storage, respectively.

[0057] The flood control dispatch system immediately initiates decision optimization after completing the risk assessment. Specifically, the system first identifies all high-risk areas and analyzes their spatial distribution characteristics and the causes of risk formation. Then, the system establishes a multi-objective optimization model, taking minimizing the personnel safety risk in high-risk areas as the primary objective, while considering the dispatch constraints of water conservancy facilities, such as maximum discharge flow, minimum ecological flow, and flood control storage capacity limitations. The system uses a dynamic programming algorithm to calculate the risk change trend under different dispatch schemes, taking into account the time lag effect of scheme implementation. Finally, the system selects the optimal dispatch scheme and generates a detailed sequence of dispatch instructions, including the discharge flow or flood storage requirements for each time period.

[0058] In some embodiments, scheduling optimization can be achieved in multiple ways: Optionally, the flood control scheduling system can adopt a dynamic optimization method based on scenario simulation to generate a scheduling plan library for different rainfall scenarios, match the optimal plan with real-time rainfall data, and make dynamic adjustments based on risk assessment results; Optionally, the flood control scheduling system can establish an intelligent scheduling model based on deep reinforcement learning, master the optimal scheduling strategy through long-term training, and achieve rapid decision-making under complex hydrological conditions. It is understood that other optimization algorithms can also be used to generate scheduling decisions, and this is not limited here.

[0059] The following provides a more detailed description of the process of the method provided in this implementation. Please refer to [link / reference]. Figure 2This is another flowchart illustrating the water conservancy flood control scheduling method in this application embodiment.

[0060] S201. During the flood control dispatch response, acquire dynamic social data from downstream cities that includes spatiotemporal location information.

[0061] Referring to step S101, the flood control dispatch system will acquire dynamic social data, including media data and mobile phone signaling data.

[0062] In some embodiments, the flood control dispatch system collects and integrates multi-source data to comprehensively perceive urban waterlogging and population distribution. Specifically, the flood control dispatch system establishes a social media data collection task within a preset time window, filters social media data with geographic tags based on the geographical boundaries of downstream cities to obtain media data; acquires base station signaling data from downstream cities, divides the base station signaling data according to a preset geographic grid based on the base station location, and counts the number of signaling signals in each grid; extracts the roaming status of user devices from the base station signaling data, identifies user devices moving across grids, and generates dynamic displacement data representing user movement trajectories; identifies population gathering and evacuation trends based on the dynamic displacement data, updates the number of signaling signals in the geographic grid, and obtains mobile phone signaling data; and integrates media data and mobile phone signaling data into dynamic social data.

[0063] Among them, the preset time window represents the data collection time range set by the system, which is usually 0.5-2 hours; the geographic tag refers to the location information attached to the social media content; the preset geographic grid represents the regular grid unit that divides the urban area, with a typical size of 200 meters × 200 meters.

[0064] After initiating a flood control response, the flood control dispatch system begins data collection and processing. Specifically, the system first determines the geographical boundaries of downstream cities and establishes a data collection task. For social media data, the system acquires geotagged content within a specified range in real time; for base station signaling data, the system divides areas according to a preset grid and counts the number of signaling signals within each grid. The system further analyzes user equipment roaming records, identifies trajectories of movement across grids, and updates the personnel distribution in each grid in real time. Finally, the system integrates the processed media and signaling data into dynamic social data in a standard format. Roaming status refers to the switching records of user equipment between different base stations, and dynamic displacement data represents the movement trajectory information of user equipment.

[0065] S202. Identify media data to obtain semantic or visual features associated with urban flooding in order to locate multiple real-time water accumulation points.

[0066] Referring to step S102, the flood control dispatch system will locate urban waterlogging points through semantic and visual analysis.

[0067] In some embodiments, the flood control dispatch system identifies and verifies the location of urban waterlogging points through text and image analysis technology. Specifically, the system extracts text content containing preset waterlogging keywords from media data and locates candidate waterlogging points based on the geographic location description information in the text content; it filters image content containing waterlogging scenes from media data and locates candidate waterlogging points based on the geographic marker information in the image content; it clusters the candidate waterlogging points in text and image according to geographic location, and removes waterlogging points with inconsistent timestamps or location deviations exceeding a preset deviation threshold to obtain real-time candidate waterlogging points; it calculates the confidence level of the real-time candidate waterlogging points based on distribution density and reporting time, and filters out real-time candidate waterlogging points with confidence levels higher than a preset confidence threshold to obtain multiple real-time waterlogging points.

[0068] Among them, the preset water accumulation keywords represent a set of words used to identify water accumulation-related content, such as "water accumulation" and "waterlogging"; the geographic location description information refers to the location, road section or landmark mentioned in the text; the water accumulation scene refers to the water accumulation phenomenon shown in the image; the geotag information represents the location data attached to the image; the preset deviation threshold refers to the allowable range of location error; and the confidence level represents the credibility of the water accumulation point report.

[0069] Upon acquiring media data, the flood control dispatch system immediately begins identifying waterlogged areas. Specifically, the system first performs natural language processing on the text content to extract keywords related to waterlogging, and then converts the location descriptions into coordinates using geocoding technology. Simultaneously, the system uses computer vision models to analyze image content, identify waterlogging scene features, and determine the location by combining this with the image's geotag information. The system then performs spatiotemporal clustering analysis on the text and image recognition results, calculating cluster centers as candidate waterlogged areas. By evaluating the report's temporal continuity, spatial clustering, and source reliability, the system calculates the confidence level of each candidate point, selecting high-confidence points as real-time waterlogged areas.

[0070] In some embodiments, the identification and verification of waterlogging points can be achieved in multiple ways: Optionally, the flood control dispatch system can establish a multimodal recognition model based on deep learning, simultaneously processing text semantics and image visual features, and extracting key information through an attention mechanism to achieve accurate identification and location of waterlogging scenes; Optionally, the flood control dispatch system can construct a verification mechanism based on knowledge graphs, integrating prior knowledge such as historical waterlogging records, terrain features, and drainage facility distribution to improve the reliability of the identification results. It is understood that other artificial intelligence technologies can also be used to identify and verify waterlogging points, and this is not limited here.

[0071] The main challenge in implementing this step is the interference of false information. To address this, the flood control dispatch system has established a multi-dimensional information authenticity verification mechanism. First, the system analyzes the historical credibility of information publishers, assigning higher weight to users who frequently publish accurate information. Then, the system performs consistency checks on multiple reports from the same location, analyzing the continuity of the reports' time and the similarity of their content. Finally, the system compares the identification results with surrounding terrain features and historical water accumulation records, eliminating abnormal reports that do not conform to objective laws. For example, if a large number of similar water accumulation reports suddenly appear at a location, and the terrain conditions at that location are not prone to water accumulation, the system will mark it as suspicious information and require manual verification.

[0072] It should be noted that the confidence score is calculated using a spatiotemporal weighted fusion method. For each candidate water accumulation point, its confidence score is Score = w1D + w2T, where D is the spatial density term, T is the temporal relevance term, and w1 and w2 are weighting coefficients, with w1 + w2 = 1. The spatial density D is calculated based on the area per unit area (e.g., 1 km²). 2 The confidence score is calculated by normalizing the number of reported points within a given time period. The time recentity, T, is calculated by exponential decay based on the difference between the reported time and the current time: T = exp(-λΔt), where λ is the decay coefficient and Δt is the time difference (in hours). When multiple reports of flooding are received from citizens within 30 minutes, and the reported locations are concentrated within a 100m radius, the confidence score for that location will significantly increase.

[0073] S203. Based on the characteristic information of multiple real-time water accumulation points, generate real-time urban flooding distribution information that characterizes the spatial distribution of urban flooding in the downstream city.

[0074] Referring to step S103, the flood control dispatch system will generate real-time spatial distribution of urban flooding.

[0075] S204. Determine the number of signaling messages within the preset geographic grid based on mobile phone signaling data, and generate real-time population heat map information representing the spatial distribution and population density of the downstream city population.

[0076] Referring to step S104, the flood control dispatch system will generate a real-time population distribution heat map.

[0077] S205. Based on historical flood records, determine the maximum radius of influence of each waterlogging point under different rainfall intensities to obtain the waterlogging affected area.

[0078] Among them, historical flood records represent detailed data on past waterlogging events, including location, time, rainfall intensity, and affected area; maximum impact radius refers to the maximum diffusion distance of waterlogging impact under a specific rainfall intensity; and waterlogging affected area represents the complete geographical range affected around the waterlogging point.

[0079] After obtaining the real-time locations of waterlogged areas, the flood control dispatch system analyzes the impact range. Specifically, the system first establishes a historical database, collecting and organizing complete records of past flooding events. For each waterlogged area, the system retrieves historical waterlogging records for similar locations, analyzing the changing patterns of the impact range under different rainfall intensities. Through statistical analysis, the system establishes a model corresponding to the relationship between rainfall intensity and the radius of influence. Combined with the current rainfall intensity, the system predicts the potential impact range of each waterlogged area, taking into account the correction effect of topographic features on the impact range.

[0080] In some embodiments, the impact range can be predicted in multiple ways: Optionally, the flood control dispatch system can establish an impact range prediction model based on machine learning, taking into account multi-dimensional features such as rainfall intensity, topographic features, and drainage capacity, and training it with historical data to obtain accurate impact range prediction results; Optionally, the flood control dispatch system can use a hydrological-hydrodynamic coupling model to simulate surface runoff and water diffusion processes under different rainfall conditions, and calibrate the model parameters by combining measured data. It is understood that other prediction methods can also be used to determine the impact range of water accumulation, and these are not limited here.

[0081] In some embodiments, the flood control dispatch system combines historical data and terrain features to predict the impact range of waterlogging. Specifically, the system extracts the geographical coordinates of each historical waterlogging point, the corresponding rainfall intensity value, and the measured impact radius from the historical waterlogging record database to construct a mapping model between rainfall intensity and waterlogging impact radius. Based on the digital elevation model and drainage facility distribution data, it calculates the surface runoff path and catchment area around each historical waterlogging point to determine the direction of waterlogging diffusion. Based on the mapping model between rainfall intensity and waterlogging impact radius, and combined with the surface runoff path and catchment area, it calculates the predicted impact radius of each waterlogging point under the current rainfall intensity. Based on the predicted impact radius and the direction of waterlogging diffusion, it generates the boundary coordinates of the impact area of ​​each waterlogging point to obtain the waterlogging impact area.

[0082] Among them, the historical waterlogging record database represents a dataset storing detailed information on historical waterlogging events; the measured influence radius refers to the actual observed influence range of historical waterlogging events; the digital elevation model represents the digital representation of surface elevation; the surface runoff path refers to the flow channel of rainwater on the surface; the catchment area represents the water collection area that gathers rainwater to a certain point; and the predicted influence radius refers to the waterlogging influence range calculated by the model.

[0083] After identifying real-time waterlogging points, the flood control dispatch system predicts the impact range. Specifically, the system first extracts waterlogging records from historical databases to establish a correlation between rainfall intensity and impact radius. The system then analyzes terrain features using a digital elevation model (DEM) to calculate the catchment path and catchment area for each location, determining the main direction of water flow. Based on the current rainfall intensity, the system uses the established mapping model to predict the impact radius and corrects the prediction results according to terrain features. Finally, the system combines the catchment analysis results to generate the impact area boundary for each waterlogging point, forming a complete impact range distribution map.

[0084] It should be noted that this mapping model uses a polynomial regression method to establish the functional relationship between rainfall intensity and influence radius. First, the rainfall intensity (x) and measured influence radius (y) data of historical waterlogging points are standardized. Then, the nth-order polynomial function y = a0 + a1x + a2x is fitted using the least squares method. 2 +...+anxⁿ, where the coefficients ai are determined by minimizing the sum of squared errors. The model order n is determined through cross-validation, and generally, an order of 3-5 is sufficient to obtain a good fit. For example, historical data of a waterlogging point shows that the radius of influence is approximately 100m when the rainfall intensity is 30mm / h and approximately 180m when it is 50mm / h. This model can accurately predict the radius of influence under other rainfall intensities.

[0085] S206. Based on the road slope and the distribution of drainage facilities, determine the inundation depth level within the waterlogged area.

[0086] Among them, road slope indicates the degree and direction of inclination of the road surface; drainage facility distribution indicates the spatial layout of municipal drainage facilities such as rainwater inlets and pipe networks; flooding depth level refers to the classification of water depth at different locations within the water accumulation area.

[0087] After identifying the affected area, the flood control dispatch system performs inundation depth analysis. Specifically, the system first extracts slope information from the road network based on a high-precision digital elevation model, identifying local low-lying areas and catchment nodes. Combined with spatial distribution data of drainage facilities, the system analyzes the drainage capacity of each area, including storm drain density, pipe network capacity, and drainage direction. Through a hydrodynamic model, the system simulates the flow and accumulation of water within the affected area, calculating the inundation depth at different locations. Finally, based on preset depth thresholds, the system classifies the inundated area into different levels and generates an inundation depth distribution map.

[0088] In some embodiments, the determination of inundation depth levels can be achieved in multiple ways: Optionally, the flood control dispatch system can employ a rapid assessment method based on topographic analysis, combined with the calculation of water catchment accumulation and drainage capacity assessment, to quickly estimate the distribution of inundation depth within the region; alternatively, the flood control dispatch system can establish a two-dimensional hydrodynamic model, considering surface roughness, permeability, and the operational status of drainage facilities, to accurately simulate the water accumulation evolution process. It is understood that other hydrological analysis methods can also be used to calculate inundation depth, and this is not limited here.

[0089] S207. Based on the distribution of buildings and the connectivity of roads, identify major traffic arteries and densely populated areas within the flood-affected area, and determine the difficulty of passage for people in the flood-affected area under different flood depth levels.

[0090] Among them, major traffic arteries refer to urban main roads, important travel routes and emergency evacuation routes; densely populated areas refer to areas with high traffic volume such as commercial areas, residential areas and transportation hubs; road connectivity status indicates the real-time traffic situation of the road network; and pedestrian access difficulty refers to the degree of difficulty for people to pass through the area at different flood depths.

[0091] After completing the inundation depth analysis, the flood control dispatch system conducts a traffic risk assessment. Specifically, the system first identifies the road grades and functional attributes within the affected area based on urban planning data, marking major traffic arteries and key nodes. Combining building distribution data, the system analyzes building types, density, and usage functions to identify densely populated areas. Then, the system overlays road network connectivity with inundation depth levels to assess the traffic status of different road sections. By establishing a depth-traffic difficulty mapping relationship, the system calculates the pedestrian traffic difficulty index for each area and dynamically updates it considering pedestrian flow characteristics at different times.

[0092] In some embodiments, traffic difficulty assessment can be achieved in multiple ways: Optionally, the flood control dispatch system can establish a traffic capacity assessment model based on multi-source data fusion, comprehensively considering road width, slope, paving type, and the distribution of surrounding facilities, combined with traffic restrictions under different flood depths, to quantitatively assess traffic difficulty; Optionally, the flood control dispatch system can utilize personnel passage records from historical flooding events to establish a traffic difficulty prediction model based on practical experience, improving the accuracy of the assessment. It is understood that other assessment methods can also be used to determine traffic difficulty, and this is not limited here.

[0093] The main challenge in implementing this step is the significant individual differences in people's mobility capabilities. To address this, the flood control dispatch system established a group-based mobility difficulty assessment mechanism. The system first identifies the distribution characteristics of different population groups based on building function; for example, residential areas primarily house general residents, while hospitals and nursing homes include individuals with mobility impairments. Then, the system sets differentiated mobility difficulty assessment standards for different groups. For instance, for areas surrounding schools, the system adopts a more conservative assessment standard to ensure a safety margin. Simultaneously, the system considers the spatiotemporal distribution patterns of the population, dynamically adjusting the assessment weights at different times.

[0094] S208. Determine the danger level of different water accumulation points based on the difficulty of passage for personnel.

[0095] Among them, the danger level indicates the degree of threat that the water accumulation point poses to personnel safety; the difficulty of personnel passage is a key indicator for assessing the danger level, reflecting the possibility of personnel passing through or evacuating under different circumstances.

[0096] After receiving the assessment results of passage difficulty, the flood control dispatch system classifies the hazard levels. Specifically, the system first establishes a hazard level assessment framework, incorporating factors such as the difficulty of passage, water depth, and affected area into the assessment system. Then, based on the severity of the passage difficulty and the distribution characteristics of people within the affected area, the system establishes hazard level classification standards. Through multi-factor weighted calculations, the system assigns a corresponding hazard level to each waterlogged point and dynamically updates it based on real-time monitoring data. The system also considers the duration and development trend of waterlogging, making time-series corrections to the hazard levels.

[0097] It should be noted that the traffic difficulty index D is calculated using a multi-factor weighted average: D = Σ(wi × fi), where wi is the weight of each factor, and fi is the normalized factor score. The main factors considered include water depth h (m), water flow velocity v (m / s), road slope θ (°), and nighttime lighting conditions L (lx). The scoring function for each factor is in piecewise linear or exponential form, such as the water depth score fh = exp(αh), where α is the attenuation coefficient. When the water depth exceeds 0.5m and the water flow velocity reaches 0.3m / s, even on a flat road, a high traffic difficulty index will result, indicating that the area is no longer suitable for pedestrian traffic.

[0098] In some embodiments, the determination of hazard levels can be achieved in multiple ways: Optionally, the flood control dispatch system can establish a hazard level assessment model based on fuzzy comprehensive evaluation, standardize each assessment factor, and then perform weighted fusion to obtain a comprehensive hazard index; Optionally, the flood control dispatch system can use machine learning methods to train a hazard level classification model based on historical disaster data to achieve rapid and accurate level determination. It is understood that other classification methods can also be used to determine the hazard level of waterlogging points, and this is not limited here.

[0099] The main challenge in implementing this step is the insufficient stability of hazard level assessment results. To address this, the flood control dispatch system has established a hazard level smoothing mechanism. The system employs a time window moving average method to avoid frequent changes in hazard levels due to short-term fluctuations. Simultaneously, the system sets hazard level transition thresholds; adjustments are only triggered when changes in assessment indicators exceed preset thresholds. For example, if the passability of a waterlogged area fluctuates, the system will comprehensively consider assessment results from previous and subsequent periods to ensure the stability of the hazard level determination. The system also uses weather forecast information to predict hazard level trends and make necessary adjustments in advance.

[0100] S209. Based on real-time urban flooding distribution information and real-time population thermal information, spatial overlay analysis is performed on the impact range of urban flooding and population thermal distribution to generate a comprehensive risk assessment result that characterizes the safety risk level of people in different areas of the downstream city.

[0101] Referring to step S105, the flood control dispatch system will assess the safety risk level of personnel in different areas.

[0102] S210. Extract the geographical scope and population density data of each high-risk area from the comprehensive risk assessment results, and determine the available evacuation routes based on the real-time traffic status of the current road network.

[0103] Among them, high-risk areas refer to areas with a high level of personnel safety risk determined after comprehensive assessment; geographical scope refers to the spatial boundaries and area of ​​high-risk areas; available evacuation routes refer to road routes that can be used for the safe evacuation of personnel.

[0104] After obtaining the comprehensive risk assessment results, the flood control dispatch system conducts evacuation analysis. Specifically, the system first extracts the spatial scope of high-risk areas from the assessment results and obtains real-time population distribution data within these areas. Then, based on real-time traffic information, the system constructs a road network traffic status map, identifying road segments that are not affected by flooding and have sufficient traffic capacity. Through network analysis algorithms, the system plans multiple evacuation routes for each high-risk area and calculates the traffic capacity and evacuation time for each route. The system also considers the capacity of surrounding safe areas to ensure the feasibility of the evacuation routes.

[0105] In some embodiments, evacuation routes can be determined in multiple ways: Optionally, the flood control dispatch system can employ a multi-objective path planning algorithm, simultaneously considering factors such as path length, traffic capacity, and safety factor, to generate an optimal set of evacuation routes; alternatively, the flood control dispatch system can establish a crowd evacuation model based on cellular automata to simulate the crowd flow process under different evacuation schemes and evaluate the feasibility of the schemes. It is understood that other path planning methods can also be used to plan evacuation routes, and this is not limited here.

[0106] S211. Obtain the current water level, discharge flow, and gate opening parameters of the upstream water conservancy facilities, and establish a model to assess the impact of the discharge flow on the downstream water level.

[0107] Among them, the water storage level represents the current water level status of the water conservancy facility; the downstream discharge flow refers to the amount of water discharged downstream by the water conservancy facility; the gate opening parameter represents the degree of opening of the flood discharge gate of the water conservancy facility; and the implementation impact model refers to the mathematical model describing the impact of upstream discharge on downstream water level.

[0108] Before initiating dispatching decisions, the flood control dispatching system acquires the operational status of water conservancy facilities. Specifically, the system first collects key operational parameters of the facilities in real time, including real-time water level, current discharge flow, and gate opening data. Then, the system combines monitoring data from hydrological stations to analyze the propagation patterns and attenuation characteristics of the discharge flow in the river channel. By establishing a hydrodynamic model, the system simulates the water flow evolution process under different discharge conditions and calculates water level changes at key downstream sections. The system also considers the influence of tributary inflows and regional rainfall to construct a complete water level impact response model.

[0109] In some embodiments, the impact model can be established in several ways: Optionally, the flood control dispatch system can adopt a one-dimensional hydrodynamic model, based on the Saint-Venant equations to describe the river flow motion, and calibrate the model parameters using measured data; Optionally, the flood control dispatch system can establish a data-driven machine learning model, using historical dispatch records to train the mapping relationship between the discharge flow and the downstream water level. It is understood that other modeling methods can also be used to simulate the impact process, and this is not limited here.

[0110] S212. Based on the implementation impact model, predict the risk of worsening urban flooding under different discharge flows, and generate hierarchical scheduling time thresholds based on the traffic capacity of available evacuation channels.

[0111] Among them, the risk of worsening waterlogging indicates the possibility that upstream discharge will lead to increased water accumulation downstream; the time threshold for tiered dispatching refers to the maximum allowable execution time for different levels of dispatching schemes; and the traffic capacity refers to the number of people that can pass through evacuation channels per unit time.

[0112] After establishing an impact model, the flood control dispatch system conducts a dispatch risk analysis. Specifically, the system first simulates water level changes under different discharge scenarios based on the implementation impact model. Combined with the current waterlogging situation, the system assesses the cumulative effect of additional discharge on existing flooding. Then, the system calculates the actual capacity of each evacuation route, considering factors such as road width, slope, and safety distance, to determine the maximum number of people that can be evacuated per unit time. By comparing evacuation demand with capacity, the system calculates the shortest time required to complete evacuation and sets corresponding dispatch time thresholds.

[0113] It should be noted that the time threshold was calculated using a crowd evacuation dynamics model. This model considers the width W (m) of the evacuation route and the crowd density ρ (people / m²). 2 Given the average moving speed v (m / s), the maximum evacuation flow rate F = W × ρ × v per unit time is calculated. Combining this with the total number of people N in the high-risk area, the minimum time T required to complete the evacuation is obtained as T = N / F. Considering a safety margin, the actual scheduling time threshold is set to 80% of T. For example, when the crowd density in an 8-meter-wide evacuation corridor is 3 people / m²... 2 When the average moving speed is 1m / s, the evacuation capacity of the passage is 24 people / s. If 1,000 people need to be evacuated in the area, the minimum evacuation time is about 42 seconds, and the corresponding scheduling time threshold should be set to 34 seconds.

[0114] In some embodiments, the time threshold can be determined in several ways: Optionally, the flood control dispatch system can establish a simulation model based on the dynamics of crowd evacuation to simulate the evacuation process under different crowd densities and calculate the safe evacuation time; Optionally, the flood control dispatch system can adopt queuing theory to treat the evacuation process as a service system and determine a reasonable evacuation time based on the arrival rate and service rate. It is understood that other analytical methods can also be used to calculate the dispatch time threshold, which is not limited here.

[0115] During the implementation of this step, the main challenge is the mismatch between the time threshold and the actual evacuation capacity. To address this issue, the flood control dispatch system has established a dynamic time threshold adjustment mechanism. The system calculates the current evacuation efficiency by monitoring the actual traffic conditions of evacuation routes in real time. When the actual evacuation speed is found to be lower than expected, the system will promptly extend the time threshold to ensure evacuation safety. For example, when a local congestion occurs on an evacuation route, the system will recalculate the required time based on the real-time traffic capacity and adjust the corresponding dispatch time threshold.

[0116] S213. Generate multiple flood control dispatch plans, evaluate the matching degree between the execution time of each plan and the hierarchical dispatch time threshold, and determine the optimal dispatch plan that minimizes the safety risks to personnel in high-risk areas.

[0117] Among them, the flood control dispatch plan represents a complete operation plan for water conservancy facilities; the execution time refers to the time required from the start of the plan to the production of its effects; the matching degree assessment refers to the degree of fit between the plan's execution time and the allowable time threshold; and the optimal dispatch plan refers to the plan that minimizes risks to the greatest extent while ensuring safety.

[0118] After determining the time thresholds, the flood control dispatch system optimizes the contingency plans. Specifically, the system first generates multiple feasible dispatch plans based on the operating parameter ranges of the water conservancy facilities. Each plan includes a detailed sequence of gate opening adjustments and the process of downstream flow changes. Then, based on an implementation impact model, the system simulates the water level changes after each plan is implemented and calculates the execution time. By comparing the execution time with the tiered dispatch time thresholds, the system evaluates the time feasibility of the plans. Finally, the system selects the plan that minimizes the safety risks to personnel in high-risk areas from the time-feasible plans as the optimal plan.

[0119] In some embodiments, contingency plan optimization can be achieved in multiple ways: Optionally, the flood control dispatch system can use a genetic algorithm to optimize the dispatch plan, taking risk minimization and time constraints as fitness functions, and obtaining the optimal solution through iterative evolution; Optionally, the flood control dispatch system can establish a dispatch decision model based on reinforcement learning, learning the optimal dispatch strategy through long-term training to achieve rapid plan generation. It is understood that other optimization algorithms can also be used to generate and screen dispatch plans, which are not limited here.

[0120] It should be noted that this implementation impact model is constructed based on a set of one-dimensional unsteady flow dynamic equations. The model uses the Saint-Venant equations to describe the river flow motion, including the continuity equation ∂A / ∂t + ∂Q / ∂x = q and the momentum equation ∂Q / ∂t + ∂(Q / ∂t + ∂(Q / ∂t) = q). 2 The equations are: ∂Q / ∂x + gA∂h / ∂x = gA(S0 - Sf), where A is the cross-sectional area, Q is the flow rate, h is the water level, S0 is the riverbed slope, and Sf is the friction slope. The water level process at any cross-section can be obtained by numerically solving the equations using the method of characteristics. The model parameters are calibrated using measured hydrological data. When the discharge flow from an upstream reservoir increases by 100 m³ / s... 3 When the water level changes at each downstream control section are measured, the process can be predicted.

[0121] During the implementation of this step, the main challenge is the insufficient operability of the optimal contingency plan. To address this issue, the flood control dispatch system has established a contingency plan feasibility verification mechanism. The system first verifies whether the plan meets the physical constraints and operational procedures of the water conservancy facilities, eliminating unfeasible schemes. Then, the system assesses the operational complexity during plan execution, prioritizing schemes that are simple to operate and have good stability. For example, when multiple contingency plans have similar effects, the system will select the scheme with fewer gate adjustments and more gradual changes, improving the reliability of execution.

[0122] S214. Convert the optimal scheduling plan into flood control scheduling instructions.

[0123] Among them, flood control dispatch instructions refer to the specific operational requirements of water conservancy facilities, including detailed parameters such as dispatch time, gate opening degree, and discharge flow.

[0124] After determining the optimal contingency plan, the flood control dispatch system performs command conversion. Specifically, the system first breaks down the dispatch process in the plan into time-series operation steps, clarifying the specific operation requirements at each time point. Then, according to the control system interface specifications of the water conservancy facilities, the system converts the operation requirements into standard format control commands. Simultaneously, the system generates the necessary user interface and prompts for human-machine interaction, ensuring that dispatchers can accurately understand and execute the commands.

[0125] In some embodiments, command conversion can be achieved in multiple ways: Optionally, the flood control dispatch system can establish a graphical command display interface, intuitively presenting the dispatch process through dynamic curves and numerical displays; alternatively, the flood control dispatch system can generate step-by-step operation guidance documents, including detailed operation instructions and precautions. It is understood that other methods can also be used to convert plans into commands, which are not limited here. During the implementation of this step, the main challenge is the untimely feedback of command execution. To address this issue, the flood control dispatch system has established a command execution monitoring mechanism. The system collects the operating parameters of water conservancy facilities in real time to verify the command execution effect. When an execution deviation is detected, the system will promptly generate a correction command to ensure the dispatch process proceeds as expected. For example, when the gate opening is not adjusted correctly, the system will quickly calculate the supplementary adjustment amount and generate a correction command.

[0126] In some embodiments, the flood control dispatching system optimizes the execution effect of dispatching instructions based on historical experience and real-time monitoring. Specifically, the system acquires historical dispatching data and dispatching effect evaluation results of upstream water conservancy facilities to establish an effect evaluation model for dispatching instructions; verifies the feasibility of flood control dispatching instructions based on the effect evaluation model, and generates dispatching parameters that include adjustment target values ​​and execution timing; establishes a feedback control strategy for gate opening and discharge based on the dispatching parameters, and implements the feedback control strategy; monitors downstream water level changes and the scope of waterlogging impact in real time, obtains real-time monitoring results, and adjusts the control parameters of the feedback control strategy based on the real-time monitoring results.

[0127] Among them, historical dispatch data represents detailed records of past flood control dispatches, including dispatch instructions and execution parameters; dispatch effect evaluation results refer to the actual effect evaluation of historical dispatch actions; effect evaluation model represents a mathematical model for predicting the effect of dispatch instruction execution; adjustment target value refers to the expected control effect; feedback control strategy represents a method for dynamically adjusting control parameters based on real-time monitoring results.

[0128] The flood control dispatch system continuously optimizes and adjusts itself after generating dispatch instructions. Specifically, the system first analyzes historical dispatch data, extracts the correspondence between dispatch instructions and their effects, and establishes an effect evaluation model. The system uses this model to verify the feasibility of newly generated dispatch instructions, ensuring that the instruction parameters are within a reasonable range. Then, the system formulates a feedback-based control strategy, including gate opening adjustment sequences and flow control targets. The system continuously monitors downstream water levels and flooding conditions, compares actual effects with expected targets, and dynamically optimizes control parameters. When actual effects deviate from expectations, the system promptly adjusts the control strategy to ensure effective dispatch.

[0129] In some embodiments, the evaluation and optimization of scheduling effects can be achieved in multiple ways: Optionally, the flood control scheduling system can establish an optimization system based on model predictive control, dynamically planning the optimal control sequence through multi-step prediction in the rolling time domain to achieve precise control of the scheduling process; Optionally, the flood control scheduling system can adopt reinforcement learning methods, treating the scheduling process as a decision sequence, and optimizing the scheduling strategy through repeated training to achieve continuous improvement of scheduling effects. It is understood that other control optimization methods can also be used to evaluate and improve scheduling effects, which are not limited here. In the implementation of this step, the main challenge is the problem of delayed response of scheduling effects. To solve this problem, the flood control scheduling system establishes a hierarchical and progressive control optimization mechanism. The system divides the control process into a rapid response layer and an optimization adjustment layer. The rapid response layer is responsible for immediate adjustments to sudden changes, while the optimization adjustment layer is responsible for continuous improvement of long-term effects. Through real-time monitoring data, the system can quickly detect deviations in control effects and adjust parameters in a timely manner. For example, when the system detects that the rate of water level rise in a certain area exceeds expectations, it will immediately adjust the opening of the upstream gate and update the control model parameters based on the new monitoring data to optimize subsequent scheduling strategies. The system also reserves a certain amount of adjustment margin to ensure that there is still sufficient control capability in extreme situations.

[0130] In this embodiment, an intelligent flood control dispatching method based on dynamic social data and multi-source information fusion is adopted. By acquiring and analyzing social media data and mobile phone signaling data in real time, and combining semantic analysis and visual recognition technology, the method accurately locates waterlogging points. Simultaneously, it utilizes spatial overlay analysis based on population thermal distribution. Therefore, it can comprehensively perceive the urban flooding situation and population distribution, achieving accurate assessment of the safety risks to people in different areas. This effectively solves the problems of numerous monitoring blind spots, insufficient perception of personnel safety risks, and experience-based dispatching decisions in traditional flood control dispatching, thus realizing intelligent and precise flood control dispatching. By establishing a historical data-driven impact prediction model and a multi-objective optimized dispatching decision-making mechanism, the system can scientifically assess the impact range of waterlogging, optimize and generate dispatching plans, improve the timeliness and accuracy of flood control dispatching, and maximize the protection of urban residents' lives and property, providing an innovative solution for smart city flood control and disaster reduction.

[0131] The flood control dispatch system in the embodiments of this invention is described below from the perspective of hardware processing. Please refer to [link / reference]. Figure 3 This is a schematic diagram of the physical device structure of a flood control dispatch system in this application embodiment.

[0132] It should be noted that, Figure 3 The structure of the flood control dispatch system shown is merely an example and should not impose any limitations on the functionality and scope of use of the embodiments of the present invention.

[0133] like Figure 3As shown, the flood control dispatch system includes a CPU 301, which can perform various appropriate actions and processes according to a program stored in ROM 302 or a program loaded from storage section 308 into RAM 303, such as executing the methods described in the above embodiments. RAM 303 also stores various programs and data required for system operation. The CPU 301, ROM 302, and RAM 303 are interconnected via bus 304. I / O interface 305 is also connected to bus 304.

[0134] The following components are connected to I / O interface 305: input section 306 including audio input devices, push-button switches, etc.; output section 307 including liquid crystal display (LCD) and audio output devices, indicator lights, etc.; storage section 308 including hard disks, etc.; and communication section 309 including network interface cards such as LAN (Local Area Network) cards, modems, etc. Communication section 309 performs communication processing via a network such as the Internet. Drive 310 is also connected to I / O interface 305 as needed. Removable media 311, such as disks, optical disks, magneto-optical disks, semiconductor memories, etc., are installed on drive 310 as needed so that computer programs read from them can be installed into storage section 308 as needed.

[0135] In particular, according to embodiments of the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of the present invention include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing computer programs for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication section 309, and / or installed from removable medium 311. When the computer program is executed by CPU 301, it performs the various functions defined in the present invention.

[0136] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of the present invention. Each block in a flowchart or block diagram may represent a module, program segment, or portion of code, which contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those shown in the drawings.

[0137] Specifically, the flood control dispatch system of this embodiment includes a processor and a memory. The memory stores a computer program. When the computer program is executed by the processor, it implements the water conservancy flood control dispatch method provided in the above embodiment.

[0138] In another aspect, the present invention also provides a computer-readable storage medium, which may be included in the flood control dispatching system described in the above embodiments; or it may exist independently and not be assembled into the flood control dispatching system. The storage medium carries one or more computer programs, which, when executed by a processor of the flood control dispatching system, enable the flood control dispatching system to implement the water conservancy flood control dispatching method provided in the above embodiments.

[0139] The above-described embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit it. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

[0140] As used in the above embodiments, depending on the context, the term "when..." can be interpreted as meaning "if...", "after...", "in response to determining...", or "in response to detecting...". Similarly, depending on the context, the phrase "when determining..." or "if (the stated condition or event) is interpreted as meaning "if determining...", "in response to determining...", "when (the stated condition or event) is detected", or "in response to detecting (the stated condition or event)".

Claims

1. A method for water conservancy flood control scheduling, characterized in that, The method, applied to a flood control dispatch system, includes: During flood control and dispatch response, dynamic social data containing spatiotemporal location information from the geographical range of downstream cities is acquired; the dynamic social data includes at least media data actively published by users on social media platforms and mobile signaling data provided by mobile communication network operators; Identify the media data to obtain semantic or visual features associated with urban flooding, in order to locate multiple real-time water accumulation points; Based on the feature information of the multiple real-time water accumulation points, real-time urban flooding distribution information representing the spatial distribution of urban flooding in the downstream city is generated. Based on the mobile phone signaling data, the number of signaling messages within the preset geographic grid is determined, and real-time population heat map information representing the spatial distribution and population density of the downstream city population is generated. Based on the real-time urban flooding distribution information and the real-time population thermal information, the spatial overlay analysis of the urban flooding impact range and population thermal distribution is performed to generate a comprehensive risk assessment result that characterizes the safety risk level of people in different areas of the downstream city. With the goal of minimizing the safety risks to personnel in high-risk areas, the comprehensive risk assessment results are optimized and analyzed to generate or adjust flood control dispatch instructions applied to upstream water conservancy facilities in order to control the downstream flow or flood storage capacity of the upstream water conservancy facilities.

2. The method according to claim 1, characterized in that, The step of acquiring dynamic social data containing spatiotemporal location information from the geographical area of ​​downstream cities during flood control dispatch response specifically includes: Establish a social media data collection task within a preset time window, and filter social media data with geographical tags according to the geographical boundaries of downstream cities to obtain the media data; Acquire base station signaling data from the downstream city, divide the base station signaling data into a preset geographical grid according to the base station location, and count the number of signaling messages in each grid. The roaming status of user equipment is extracted from the base station signaling data, user equipment moving across grids is identified, and dynamic displacement data representing the user's movement trajectory is generated. Based on the dynamic displacement data, the population gathering and evacuation patterns are identified, the signaling quantity within the geographic grid is updated, and the mobile phone signaling data is obtained. The media data and the mobile signaling data are integrated to form dynamic social data.

3. The method according to claim 1, characterized in that, The step of identifying the media data to obtain semantic or visual features associated with urban flooding in order to locate multiple real-time waterlogging points specifically includes: Extract text content containing preset water accumulation keywords from the media data, and locate candidate water accumulation points based on the geographical location description information in the text content; Image content containing water accumulation scenes is filtered from the media data, and candidate water accumulation points are located based on the geotag information in the image content; The candidate water accumulation points in the text and the candidate water accumulation points in the image are clustered according to geographical location, and water accumulation points with inconsistent timestamps or location deviations exceeding a preset deviation threshold are removed to obtain real-time candidate water accumulation points. The confidence level of the real-time candidate water accumulation points is calculated based on the distribution density and reporting time. Real-time candidate water accumulation points with confidence levels higher than a preset confidence threshold are selected to obtain multiple real-time water accumulation points.

4. The method according to claim 1, characterized in that, Before the step of performing spatial overlay analysis on the flood impact range and population thermal distribution based on the real-time flood distribution information and the real-time population thermal information to generate a comprehensive risk assessment result characterizing the safety risk level of people in different areas of the downstream city, the method further includes: Based on historical flooding records, the maximum radius of influence of each waterlogging point under different rainfall intensities is determined, thus obtaining the area affected by waterlogging. Based on the road slope and the distribution of drainage facilities, the inundation depth level within the waterlogged area is determined; Based on the distribution of buildings and the connectivity of roads, identify major traffic arteries and densely populated areas within the flood-affected area, and determine the difficulty of passage for people in the flood-affected area under different flood depth levels; The danger level of different waterlogging points is determined based on the difficulty of passage for personnel.

5. The method according to claim 4, characterized in that, The step of determining the maximum influence radius of each waterlogging point under different rainfall intensities based on historical flooding records to obtain the waterlogging-affected area specifically includes: The geographical coordinates, corresponding rainfall intensity values, and measured influence radius of each historical waterlogging point are extracted from the historical waterlogging record database to construct a mapping model between rainfall intensity and waterlogging influence radius; Based on the digital elevation model and drainage facility distribution data, the surface water catchment path and catchment area around each historical water accumulation point are calculated to determine the direction of water spread. Based on the mapping model between rainfall intensity and water accumulation radius, and combined with the surface water runoff path and the water catchment area, the predicted value of the influence radius of each water accumulation point under the current rainfall intensity is calculated. Based on the predicted radius of influence and the direction of water spread, the boundary coordinates of the influence area of ​​each water accumulation point are generated to obtain the water accumulation influence area.

6. The method according to claim 1, characterized in that, The steps of optimizing and analyzing the comprehensive risk assessment results to generate or adjust flood control dispatch instructions applied to upstream water conservancy facilities, with the decision-making objective of minimizing personnel safety risks in high-risk areas, specifically include: The geographical scope and population density data of each high-risk area are extracted from the comprehensive risk assessment results, and available evacuation routes are determined based on the real-time traffic status of the current road network. Obtain the current water level, discharge flow, and gate opening parameters of upstream water conservancy facilities, and establish a model to assess the impact of discharge flow on downstream water level. Based on the implementation impact model, the risk of worsening urban flooding under different discharge flows is predicted, and a graded scheduling time threshold is generated based on the traffic capacity of the available evacuation channels. Multiple flood control dispatch plans are generated, and the matching degree between the execution time of each plan and the hierarchical dispatch time threshold is evaluated to determine the optimal dispatch plan that minimizes the safety risk to personnel in high-risk areas. The optimal scheduling plan is then transformed into flood control scheduling instructions.

7. The method according to claim 6, characterized in that, After the step of converting the optimal scheduling plan into flood control scheduling instructions, the method further includes: Obtain historical scheduling data and scheduling effect evaluation results of the upstream water conservancy facilities, and establish an effect evaluation model for scheduling instructions; The feasibility of the flood control dispatch instructions is verified based on the effect evaluation model, and dispatch parameters including adjustment target values ​​and execution timing are generated. A feedback control strategy for gate opening and discharge flow is established based on the scheduling parameters, and the feedback control strategy is implemented. The system monitors downstream water level changes and the extent of flooding in real time, obtains real-time monitoring results, and adjusts the control parameters of the feedback control strategy based on these results.

8. A flood control dispatch system, characterized in that, The flood control dispatch system includes: one or more processors and a memory; the memory is coupled to the one or more processors, the memory is used to store computer program code, the computer program code includes computer instructions, and the one or more processors call the computer instructions to cause the flood control dispatch system to perform the method as described in any one of claims 1-7.

9. A computer-readable storage medium comprising instructions, characterized in that, When the instruction is executed on the flood control dispatch system, the flood control dispatch system performs the method as described in any one of claims 1-7.

10. A computer program product, characterized in that, When the computer program product is run on the flood control dispatch system, it causes the flood control dispatch system to perform the method as described in any one of claims 1-7.

Images