Post-earthquake disaster influence assessment method and platform based on multi-source data fusion

Abstract

The invention discloses a post-earthquake disaster influence assessment method and platform based on multi-source data fusion, and relates to the technical field of data fusion and processing, and the method comprises the steps: carrying out the spatial interpolation processing of three graphs, and obtaining a spatial grid; performing region type marking on the space grid to obtain a region type; obtaining a regional type population coefficient in combination with the population distribution thermodynamic diagram; generating a comprehensive disaster grade distribution map; and classifying the comprehensive disaster grade distribution map to determine a final disaster grade. The technical problems that in post-earthquake emergency communication evaluation, spatial distribution of multi-source data of iron towers, base stations, population and the like is inconsistent, deviation occurs when disaster grade division is conducted through the data, and the disaster situation cannot be accurately evaluated are solved, accurate evaluation of the post-earthquake disaster situation is achieved, the accuracy of disaster grade division is improved, and the safety of post-earthquake emergency communication evaluation is improved. And a reliable basis is provided for reasonable allocation of post-earthquake emergency rescue resources and efficient rescue actions.

Description

Technical Field

[0001] The present invention relates to the technical field of data fusion and processing, and in particular to a post-earthquake disaster impact assessment method and platform based on multi-source data fusion. Background Art

[0002] In post-earthquake emergency communications assessment scenarios, accurately classifying disaster levels is crucial for the rational allocation of rescue resources and effective rescue efforts. Traditionally, simple statistical methods have been used to analyze and classify these data. However, with the increasing demand for accurate disaster assessments, this traditional approach has exposed serious problems in practical applications. Due to the inconsistent spatial distribution of acquired multi-source data, traditional methods are unable to effectively address this discrepancy, resulting in significant deviations in the classification of disaster levels. This makes it difficult to accurately reflect the actual disaster situation, providing a reliable basis for post-earthquake rescue decision-making, and seriously affecting rescue efficiency and effectiveness. Summary of the Invention

[0003] This application solves the technical problem that in the post-earthquake emergency communication assessment, the spatial distribution of multi-source data such as towers, base stations and population is inconsistent, which leads to deviations when using these data to classify disaster levels, making it impossible to accurately assess the disaster situation. This application obtains the tower power outage rate / offline rate, base station decommissioning rate and population thermal data respectively and generates corresponding distribution maps, uses a spatial interpolation algorithm to unify them into spatial grids with the same resolution, marks different dominant areas according to thresholds, calculates the population density weighted coefficient based on population thermal data, uses a multi-source data fusion algorithm to generate a comprehensive disaster level distribution map, and then determines the final disaster level according to rules, thereby solving the problem of disaster level classification deviation caused by inconsistent data spatial distribution in the post-earthquake emergency communication assessment, making the disaster level assessment results more accurate and reliable.

[0004] In response to the above technical problems, this application proposes a technical solution for a post-earthquake disaster impact assessment method and platform based on multi-source data fusion.

[0005] In the first aspect, the present application provides a post-earthquake disaster impact assessment method based on multi-source data fusion, wherein the method includes: performing spatial interpolation processing on the power outage distribution map, the equipment abnormality distribution map and the population distribution heat map to obtain a spatial grid; marking the spatial grid with regional types to obtain regional types; combining the population distribution heat map, calculating the population density according to the regional type, and obtaining the regional type population coefficient; combining the regional type with the regional type population coefficient based on a multi-source data fusion algorithm to generate a comprehensive disaster level distribution map; classifying the comprehensive disaster level distribution map according to preset disaster level classification rules to determine the final disaster level.

[0006] On the second aspect, the present application provides a post-earthquake disaster impact assessment platform for multi-source data fusion, wherein the system includes: a spatial grid acquisition unit, used to perform spatial interpolation processing on the power outage distribution map, the equipment abnormality distribution map and the population distribution heat map to obtain a spatial grid; a region type acquisition unit, used to mark the spatial grid with a region type to obtain a region type; a population coefficient acquisition unit, used to combine the population distribution heat map, calculate the population density according to the region type, and obtain the region type population coefficient; a distribution map generation unit, used to combine the region type with the region type population coefficient based on a multi-source data fusion algorithm to generate a comprehensive disaster level distribution map; a disaster level determination unit, used to classify the comprehensive disaster level distribution map according to a preset disaster level classification rule to determine the final disaster level.

[0007] This application proposes one or more technical solutions, which have at least the following technical effects:

[0008] This application generates corresponding distribution maps by obtaining the tower offline rate in the power outage area, the base station decommissioning rate in the communication equipment malfunction area, and the population thermal data in the densely populated area. These distribution maps are unified into spatial grids of the same resolution using a spatial interpolation algorithm, and different dominant areas are marked according to a preset threshold. The population density weighted coefficient is calculated in combination with the population thermal data, and a comprehensive disaster level distribution map is generated using a multi-source data fusion algorithm containing specific weights. Finally, the final disaster level is determined according to the preset rules, achieving an accurate assessment of the post-earthquake disaster situation, improving the accuracy of the disaster level classification, and providing a reliable basis for the rational deployment of post-earthquake emergency rescue resources and efficient rescue operations.

[0009] The above content summarizes the present application's method and platform for solving a post-earthquake disaster impact assessment method based on multi-source data fusion. The present application will describe the steps of the technical solution in detail in the following specific implementation methods to facilitate a clear and complete understanding of the present application by technical personnel. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0011] Figure 1 This is a flow chart of a post-earthquake disaster impact assessment method based on multi-source data fusion provided in an embodiment of the present application.

[0012] Figure 2This is a structural diagram of a post-earthquake disaster impact assessment platform using multi-source data fusion provided in an embodiment of the present application.

[0013] Explanation of reference numerals: spatial grid obtaining unit 1, area type obtaining unit 2, population coefficient obtaining unit 3, distribution map generating unit 4, disaster level determining unit 5. DETAILED DESCRIPTION

[0014] This application obtains the tower outage rate / offline rate, base station decommissioning rate and population thermal data and generates corresponding distribution maps, which are then unified into spatial grids with the same resolution using a spatial interpolation algorithm. Area types are marked based on thresholds, and the population thermal data is combined to calculate the weighted coefficient of population density. A multi-source data fusion algorithm is used to generate a comprehensive disaster level distribution map, and the final disaster level is determined according to the rules. Rescue resources are rationally planned based on the comprehensive disaster level distribution map, enabling the precise implementation of rescue operations in the affected areas. This achieves an accurate assessment of the post-earthquake disaster situation, improves the accuracy of the disaster level classification, and provides a reliable basis for the rational deployment of post-earthquake emergency rescue resources and efficient rescue operations.

[0015] The following will be combined with the accompanying drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only some of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0016] It should be noted that any variations of the terms "include" and "have" are intended to cover non-exclusive inclusions. For example, a process, method, system, product or server that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or modules that are not clearly listed or are inherent to these processes, methods, products or devices.

[0017] Example 1, as Figure 1 As shown, a post-earthquake disaster impact assessment method based on multi-source data fusion, wherein the method includes:

[0018] Step A100: Perform spatial interpolation processing on the power outage distribution map, the equipment abnormality distribution map, and the population distribution heat map to obtain a spatial grid.

[0019] In the embodiments of the present application, spatial interpolation is a technical method for estimating and predicting data at unmeasured points based on known data points. A spatial grid is a spatial data structure that divides geographic space into regular grid cells and is used to organize and analyze geographic data.

[0020] Specifically, the offline rate of towers in power outage areas, the decommissioning rate of base stations in areas with abnormal communication equipment, and the population thermal data and geographic coordinate information in densely populated areas are obtained, and corresponding distribution maps are generated respectively. Then, a spatial interpolation algorithm is used to obtain a spatial grid. The specific steps are described in detail in A110-A140.

[0021] Step A200: Mark the spatial grid with a region type to obtain the region type.

[0022] Optionally, based on the comparison results of the tower offline rate and the first preset threshold, and the base station decommissioning rate and the second preset threshold, the power outage dominant area, the equipment abnormality dominant area, and the dual impact area are marked respectively. These areas are combined to obtain the area type. The specific steps are described in detail in A210-A240.

[0023] Step A300: Calculate the population density according to the regional type in combination with the population distribution heat map to obtain the regional type population coefficient.

[0024] In the embodiment of the present application, the regional type population coefficient is a quantitative indicator used to measure the degree to which different regional types are affected by population factors.

[0025] In one embodiment of the present application, the population thermal data is first graded according to preset rules to obtain the population density weight, and then the hotspot area compensation factor is generated based on the functional geographic object information of the spatial grid. Finally, the two are combined to perform a population density correction calculation for the regional type to obtain the regional type population coefficient. The specific steps are described in detail in A310-A330.

[0026] Step A400: combining the regional type and the regional type population coefficient based on a multi-source data fusion algorithm to generate a comprehensive disaster level distribution map.

[0027] Specifically, a multi-source data fusion algorithm based on historical samples of labeled disaster levels and obtained by minimum error fitting is used. The weights of the tower offline rate, base station decommissioning rate, and population density coefficient are taken into account, and the regional type and regional type population coefficient are combined to generate a comprehensive disaster level distribution map. The specific steps are detailed in A410.

[0028] Step A500: Classify the comprehensive disaster level distribution map according to preset disaster level classification rules to determine the final disaster level.

[0029] In the embodiment of the present application, the preset disaster level classification rules are a set of pre-set standard systems for classifying the comprehensive disaster level distribution map to determine the final disaster level.

[0030] Specifically, first, the range of comprehensive disaster severity values ​​must be clarified. For example, by analyzing the distribution of comprehensive disaster severity levels under different disaster scenarios in historically annotated samples (data sets generated by collecting, processing, and manually annotating multi-source data on real disaster areas during historical disaster events), the levels are divided into three levels: a comprehensive value between 0 and 0.3 is considered mild, indicating that the disaster has had minimal impact on communication infrastructure and personnel, requiring only local monitoring; a value between 0.3 and 0.7 is moderate, indicating that the disaster has caused a regional decline in communication service quality and requires the deployment of medium-scale rescue resources; and a value above 0.7 is severe, corresponding to large-scale communication outages or high-risk areas with concentrated personnel, requiring priority deployment of emergency support. The setting of these thresholds requires dynamic adjustment by technical personnel based on the different disaster types (e.g., earthquakes focus on tower offline rates, while typhoons focus on equipment outage rates). For example, in earthquake disasters, the moderate level threshold for areas with a greater impact from tower offline rates can be lowered to 0.25 to highlight the critical role of power outages.

[0031] Next, a classification operation is performed on each spatial grid unit in the comprehensive disaster level distribution map. Taking a certain city grid as an example, its comprehensive disaster level value is 0.82, which is higher than the severe disaster threshold of 0.7, and the grid belongs to a dual-impact area (tower offline rate 45%, base station decommissioning rate 35%), and the superimposed population density coefficient is 0.9 (including large hospitals and transportation hubs). According to the rules, the comprehensive value interval is matched first, and 0.82 falls into the severely disaster-stricken range, and the regional type correction is triggered at the same time, that is, the dual-impact area is automatically upgraded by one level under the same comprehensive value (if the original rule is 0.7-0.9 severe, such areas are directly judged as extremely severe, and the example is temporarily set to level three), and finally marked as severely disaster-stricken. This process requires traversing all grids row by row and column by column to ensure that the level determination of each unit takes into account both numerical calculations and regional type attributes.

[0032] During the classification process, the accuracy of the rules is calculated by cross-validating historical cases (such as 150 labeled samples from a flood disaster in 2023): the correct classification rate of lightly affected grids is 92%, moderately affected grids are 88%, and severely affected grids are 95%. The missed judgment rate (actual severe grids are not identified) is controlled within 5%. If it is found that the misjudgment rate of a certain type of area (such as an area dominated by equipment anomalies) exceeds 10%, the weight parameters or threshold boundaries corresponding to this type will be retroactively adjusted. For example, in areas dominated by equipment anomalies, because the direct impact of the base station decommissioning rate on communication interruption is more significant, a special rule of "half-level increase if the comprehensive value ≥ 0.25" can be set in the rules to make the level judgment of this type of area more in line with the actual impact.

[0033] Finally, the classification results are presented as a visual distribution map, with each grid filled with a color corresponding to the level (such as red for severe, orange for moderate, and yellow for mild), and geographical elements (roads, rivers, and residential areas) superimposed to assist in decision-making. For example, after classification, a disaster area shows that the severely affected areas are concentrated in the city center (double impact + high population density compensation), the moderately affected areas are distributed along the transmission lines (power outages dominate + medium population), and the mildly affected areas are mostly suburban farmland (low population + single equipment anomalies). This classification result can directly determine the allocation of rescue resources. Emergency power generation vehicles and temporary base stations will be prioritized in severely affected areas, technical teams will be arranged for inspections in moderately affected areas, and routine monitoring will be maintained in mildly affected areas.

[0034] Through the nesting of the above-mentioned multi-dimensional rules, a precise transformation from data fusion calculation to decision-making availability level is achieved, and the business value of disaster assessment is implemented in the specific actions of emergency response.

[0035] Furthermore, step A100 in the method provided in the embodiment of the present application includes:

[0036] A110: Obtain the tower offline rate in the power outage area, extract the corresponding geographic coordinate information, and generate a power outage distribution map.

[0037] A120: Obtain the base station decommissioning rate in areas with communication equipment malfunctions, extract the corresponding geographic coordinate information, and generate a device malfunction distribution map.

[0038] A130: Obtain population thermal data for densely populated areas, extract corresponding geographic coordinate information, and generate a population distribution heat map.

[0039] A140: Perform spatial interpolation processing on the power outage distribution map, the equipment anomaly distribution map, and the population distribution heat map based on a spatial interpolation algorithm to obtain the spatial grid.

[0040] In this embodiment of the present application, the tower offline rate refers to the ratio of the number of towers that are unable to operate normally due to power outages within a power outage area to the total number of towers within the power outage area. The base station outage rate refers to the ratio of the number of base stations that are unable to provide communication services due to hardware failures, transmission interruptions, etc. within a communication equipment malfunction area to the total number of base stations within the area. Population thermal data is continuous data that reflects the degree of geographical concentration of people.

[0041] Specifically, first, the offline rate of towers in the power outage area and their corresponding geographic coordinates are obtained. Tower offline rates are primarily monitored by installing power sensors at the tower end to collect real-time data such as the mains input status and operating parameters of backup power sources (such as batteries and generators). This status information is transmitted to a monitoring center via the Internet of Things (IoT) or dedicated communication links. When the mains power is interrupted and the backup power source cannot continue to supply power (e.g., battery depletion or generator failure), the tower is considered offline. The monitoring center aggregates the real-time status data of all towers in a designated area and calculates the proportion of offline towers to the total number of towers in the area, which is the tower offline rate. Combining the longitude and latitude coordinates of each tower, a geographic information system (GIS) is used to correlate offline rate data with spatial location, generating a power outage distribution map with a color gradient indicating the level of offline rate, visually demonstrating the impact of power outages on communication towers.

[0042] Next, we obtain the base station decommissioning rate and corresponding geographic coordinates in areas with communication equipment dysfunction. The base station decommissioning rate is determined by collecting real-time operational status data from each base station through the local operator's communication network management system. The ratio of decommissioned base stations within each grid to the total number of base stations in that grid is calculated, generating similar decommissioning rate data ranging from 0% to 100%. Similarly, using the aforementioned GIS technology, we match decommissioning rate data with base station coordinates to generate a device anomaly distribution map, clearly illustrating the spatial distribution characteristics of communication equipment dysfunction.

[0043] Then, mobile phone signaling data (such as base station connection records), census data (such as neighborhood demographics), or real-time location data (such as GPS track points) are preprocessed to remove outliers and unify the spatiotemporal benchmarks (such as aggregating location points by hour or day to match the geographic units of the census data). Using the kernel density estimation (KDE) spatial interpolation method, the discrete population distribution data are converted into a continuous surface based on geographic coordinates. The algorithm assigns a kernel function (such as a Gaussian kernel) to each person location (such as a mobile phone signaling point or the center of a census block). The spatial influence range is defined by setting a bandwidth parameter (such as a 500-meter radius). For each target grid point, the weighted density values ​​of all surrounding data points are calculated (the weights decay with increasing distance). The final cumulative population density value (unit: people / square kilometer) is finally accumulated to obtain the thermal value of the grid point. During the process, the bandwidth and kernel function type need to be adjusted according to the data characteristics to ensure that the thermal values ​​of areas with concentrated populations (such as city centers and commercial areas) are significantly higher than those of sparse areas, forming a continuous thermal surface that is highly consistent with the actual population distribution. The thermal values ​​of each grid point are then associated with the coordinates (latitude and longitude) of its center point to generate a refined population distribution heat map.

[0044] After completing the distribution mapping of the three types of data, spatial interpolation processing is performed on the power outage distribution map, equipment anomaly distribution map, and population distribution heat map based on the inverse distance weighted interpolation algorithm. The resolution of the spatial grid is set according to the spatial area and / or population density. The data of each map is weighted averaged according to the principle of "the closer the distance, the greater the impact". The original data of different resolutions are unified into a spatial grid of the same resolution. The specific steps are detailed in A141.

[0045] Through the above steps, different types of data are ensured to be comparable and calculable at the same spatial scale, effectively solving the core problem of data space mismatch in existing technologies.

[0046] Furthermore, step A140 in the method provided in the embodiment of the present application includes:

[0047] A141: The spatial interpolation algorithm includes an inverse distance weighted interpolation algorithm, and the resolution of the spatial grid is set based on the spatial area and / or population density.

[0048] Specifically, the geographic coordinates (latitude and longitude) and attribute values ​​(such as tower offline rate, base station decommissioning rate, and population density) of each data point are first extracted from the three types of distribution maps to form a data set containing coordinates and attributes. The spatial grid resolution is then set based on the actual characteristics of the assessment area. This resolution is determined by technicians in this field by combining both spatial area and population density (either spatial area or population density alone, or both simultaneously). For example, in densely populated urban centers, where human activities are more sensitive to communication needs, a high-resolution grid of 500 meters by 500 meters is set. In suburban or rural areas, the resolution is adjusted to 1 kilometer by 1 kilometer based on the larger spatial area and sparse population, balancing computational efficiency and data accuracy.

[0049] Based on the set grid framework, for each grid center point to be interpolated, the inverse distance weighted algorithm is used to calculate its attribute value through the following steps: with the center point as the center of the circle, a search radius is defined (usually 1-2 times the resolution, such as a 500-meter grid corresponds to a 1-kilometer search radius), and all original data points within the range are extracted; the Euclidean distance from each data point to the center point is calculated according to the distance formula, and the inverse of the distance is used as the weight (such as the weight of the point at a distance of 200 meters is 1 / 200, and the weight of the point at a distance of 800 meters is 1 / 800) to ensure that close data points have a greater impact on the result; through the weighted average formula Where Z(x0) is the grid point attribute value, Z(x i ) is the original data point attribute, d i is the distance, p is set to 2 to ensure the rationality of weight attenuation, and the tower offline rate, base station decommissioning rate and population thermal value of the grid point are calculated.

[0050] Taking a local area in a disaster area as an example, the original power outage distribution map contains three data points with offline rates of 40%, 60%, and 30%, respectively. The distances from the center point of the grid to be interpolated are 300 meters, 500 meters, and 800 meters, respectively. According to the inverse distance weighted calculation, the offline rate of the grid point is (40% / 0.3 2 +60% / 0.5 2 +30% / 0.8 2 ) / (1 / 0.3 2 +1 / 0.5 2 +1 / 0.8 2 )≈48.2%. By performing calculations on all grid points one by one, a unified spatial grid containing the three types of data is finally generated, with each grid cell having the same geographic range (e.g., 500 meters × 500 meters) and aligned attribute values.

[0051] Through the above steps, the problem of inconsistent spatial scales of multi-source data was solved, providing a standardized data basis for subsequent regional type labeling, population density calculation and multi-source fusion analysis.

[0052] Furthermore, step A200 in the method provided in the embodiment of the present application includes:

[0053] A210: If the tower offline rate is higher than a first preset threshold and the base station outage rate is lower than a second preset threshold, it is marked as a power outage dominant area.

[0054] A220: If the base station decommissioning rate is higher than the second preset threshold and the tower offline rate is lower than the first preset threshold, it is marked as a device abnormality dominant area.

[0055] A230: If the tower offline rate is higher than the first preset threshold and the base station decommissioning rate is higher than the second preset threshold, it is marked as a double-impact area.

[0056] A240: Combine the power outage dominant area, the equipment abnormality dominant area, and the dual impact area to obtain the area type.

[0057] In the embodiment of the present application, the first preset threshold is a critical value for determining whether the tower offline rate is significantly higher than the normal level. The second preset threshold is a critical value for determining whether the base station decommissioning rate is significantly abnormal.

[0058] Specifically, two core indicators are first extracted for each spatial grid cell: the tower offline rate (reflecting the impact of power outages on communication infrastructure) and the base station outage rate (reflecting the proportion of communication equipment malfunctioning). The first preset threshold (e.g., 30%) and the second preset threshold (e.g., 20%) are determined by technical personnel based on historical disaster data statistics and equipment operation standards. For example, by analyzing the distribution characteristics of tower offline rates during earthquake disasters in a certain region over the past five years, it was found that when the offline rate exceeds 30%, the power system is in a serious fault state; and when the base station outage rate exceeds 20%, the communication equipment abnormality has constituted a regional service interruption.

[0059] Next, we mark the regions based on the tower offline rate and base station decommissioning rate:

[0060] Step a: If the tower offline rate (e.g., 35%) of a grid is higher than the first preset threshold (30%), and the base station outage rate (e.g., 15%) is lower than the second preset threshold (20%), then the disaster impact in that area is determined to be primarily caused by power outages and is marked as a "power outage-dominated area." This means that communication failures in that area are primarily caused by power supply issues, and the equipment itself is operating relatively normally. For example, during a typhoon disaster in 2024, 10 grids in a coastal township experienced tower offline rates of 32%-40% due to damaged transmission lines, but base station outage rates were all below 18%, and were all accurately marked as such areas.

[0061] Step b: If the base station outage rate (e.g., 25%) is higher than the second preset threshold (20%), and the tower offline rate (e.g., 25%) is lower than the first preset threshold (30%), then the area is marked as an "equipment anomaly-dominated area." Communication failures in such areas are primarily due to equipment failures (e.g., base station hardware damage, transmission link interruption), rather than power supply issues. For example, in an industrial park, lightning strikes damaged the motherboards of multiple base stations, resulting in a base station outage rate of 28%, but a tower offline rate of only 15%. This area was clearly marked as an equipment anomaly-dominated area, providing precise positioning for subsequent targeted maintenance.

[0062] Step c: If both indicators exceed their corresponding thresholds (e.g., a tower offline rate of 35% and a base station outage rate of 25%), the area is marked as a "double-impacted area," indicating that the area suffered a double blow from both power outages and equipment failures, resulting in a more severe disaster impact. For example, a power outage in a city center could cause widespread offline towers, while a dense network of base stations could experience a cascading outage due to cooling system failures. This double-labeling of such areas provides key guidance for prioritizing the allocation of subsequent rescue resources.

[0063] Finally, after completing the labeling of a single grid, the labeling results of all grids are spatially combined to form a complete distribution map containing the three types of areas. Each grid cell is uniquely labeled as one of the three types, forming a structured regional type data layer. For example, in a 100×100 spatial grid map, through cell-by-cell judgment, a regional type distribution map filled with different colors is finally generated, clearly showing the spatial distribution characteristics of "power outage-dominated areas are concentrated along suburban transmission lines," "equipment anomaly-dominated areas are concentrated in industrial areas," and "double-impact areas are concentrated in the city center."

[0064] The refined labeling method described above can more accurately reveal the impact of disasters on communication infrastructure, laying a key foundation for subsequently calculating population density and generating comprehensive disaster level distribution maps based on population thermal data, ensuring that rescue decisions can formulate differentiated plans for the dominant issues in different regions.

[0065] Furthermore, step A150 in the method provided in the embodiment of the present application includes:

[0066] A151: If the power outage dominant area and the equipment abnormality dominant area are staggered, a spatial smoothing algorithm is used to smooth the power outage dominant area and the equipment abnormality dominant area to obtain an optimized power outage dominant area and an optimized equipment abnormality dominant area.

[0067] A152: Add the optimized power outage dominant area and the optimized equipment abnormality dominant area to the power outage dominant area and the equipment abnormality dominant area respectively.

[0068] Optionally, first, after completing the area type marking (marking the power outage dominant area, equipment abnormality dominant area and dual impact area based on threshold comparison) (the specific steps are described in detail in A210-A240), the spatial grid is scanned row by row and column by column to detect whether there is an interlaced distribution of the two types of dominant areas. The judgment standard is: if the number of grids in the four adjacent neighborhoods (upper, lower, left and right) of a certain grid unit that are the power outage dominant area and the equipment abnormality dominant area exceeds a preset ratio (such as 30%), then the area is determined to be an interlaced distribution area. For example, in a rectangular area consisting of 100 grids, if 25 grids are surrounded by the neighborhoods of the two dominant areas at the same time, the spatial smoothing process is triggered.

[0069] Next, a spatial smoothing algorithm (such as a median filter) is used to process the detected interleaved areas. A 3×3 sliding window is used to traverse the interleaved areas, and the region type of the central grid in the window is replaced with the region type that appears most frequently within the window. For example, if the central grid is currently marked as a power outage-dominated area, and there are five equipment anomaly-dominated areas, three power outage-dominated areas, and one dual-impact area within the window, the central grid is relabeled as an equipment anomaly-dominated area, eliminating isolated, fragmented areas. The algorithm effectively suppresses data noise through a neighborhood voting mechanism, making region boundaries more continuous and smooth.

[0070] After processing, optimized power outage-dominated areas and equipment anomaly-dominated areas are generated and added back to the original spatial grid data, replacing the interlaced distribution areas before processing. For example, in a disaster area, the initially marked power outage-dominated areas contained five "islands" of equipment anomaly-dominated areas with an area smaller than two grid cells. After median filtering, these islands were merged into adjacent equipment anomaly-dominated areas, improving the boundary clarity of the two types of areas (which can be verified by calculating the change in the ratio of regional perimeter to area). The smoothed regional data can more realistically reflect the actual impact of the disaster on infrastructure and avoid assessment bias caused by data fragmentation.

[0071] Through the above steps, the spatial distribution of areas dominated by power outages and equipment anomalies is made more consistent with the continuity characteristics of actual disaster impacts. The optimized regional data provides a more reliable spatial analysis basis for the subsequent calculation of population density and generation of comprehensive disaster level distribution maps based on population thermal data, ultimately improving the overall reliability of disaster level assessments.

[0072] Furthermore, step A300 in the method provided in the embodiment of the present application includes:

[0073] A310: Classify the population thermal data of the population distribution heat map according to a preset population density classification rule to obtain a population density weight.

[0074] A320: Generate a hotspot area compensation factor based on the functional geographic object information of the spatial grid.

[0075] A330: Based on the population density weight and the hotspot area compensation factor, a population density correction calculation is performed on the area type to obtain the population coefficient of the area type.

[0076] In this embodiment, the population density weight is a quantitative indicator assigned after classifying population distribution thermal data. It is used to represent the basic weight of the impact of different levels of population concentration on disasters. The hotspot compensation factor is a correction coefficient generated based on the functional geographic object information within the spatial grid, which is used to strengthen the impact assessment of disaster-sensitive areas.

[0077] Specifically, first, the population thermal data is quantified according to the preset population density classification rules. The classification rules usually use the natural break point method (Jenks) or the equal interval method to divide the continuous population thermal value (such as the number of people per square kilometer) into several levels. For example, the population density is divided into "low (<500 people / km 2 ), medium (500-3000 people / km 2 ), high (>3000 people / km 2 )”, with population density weights of 0.3, 0.6, and 0.9 respectively. This weight reflects the amplifying effect of the basic population concentration on the impact of disasters - the higher the population density, the larger the number of people affected under the same disaster scenario. Taking the core area of ​​a city as an example, its population thermal value is 8,000 people / km 2 , corresponding to the highest level weight of 0.9, while the suburban rural area is only 200 people / km 2 , with a weight of 0.3.

[0078] Next, the hotspot compensation factor is generated based on the information of functional geographic objects within the spatial grid (including schools, hospitals, shopping malls, transportation hubs and other places that are highly sensitive to disaster response):

[0079] Step d: Use a geographic information system (GIS) to load high-precision vector data of functional geographic objects (e.g., coordinate points or surface areas of schools, hospitals, shopping malls, and transportation hubs) and unify the coordinate system to a geographic coordinate system consistent with the spatial grid (e.g., WGS84, 1984 World Geodetic System). For each type of object, set a differentiated influence radius (e.g., an 800-meter buffer radius for hospitals, a 500-meter buffer radius for schools, a 600-meter buffer radius for large shopping malls, and a 1000-meter buffer radius for transportation hubs). Use a buffer analysis tool to generate a buffer zone for each object, and calculate the proportion of the intersection area of ​​each spatial grid with the buffer zone to the total area of ​​the grid (e.g., within a grid, the hospital buffer zone covers 30% of the area, and the shopping mall buffer zone covers 20%).

[0080] Step e: For functional areas with a planar distribution (e.g., commercial and residential areas), the kernel density estimation (KDE) algorithm was used to calculate object density. A continuous density surface was generated by setting bandwidth parameters (e.g., a 1-kilometer bandwidth for hospital density analysis and an 800-meter bandwidth for shopping malls). Density values ​​at the grid center were extracted and normalized to the range 0–1 (e.g., a normalized value of 0.8 for high-density hospital areas and 0.2 for low-density areas). A multi-factor weighting was then performed, assigning impact weights to different object types based on disaster response priorities (e.g., 0.4 for hospitals, 0.3 for transportation hubs, 0.2 for shopping malls, and 0.1 for schools). The buffer ratio or normalized density value of each object type within the grid was multiplied by its weight and then added together to obtain an initial compensation factor (e.g., a grid containing one hospital [buffer ratio 25% × 0.4 = 0.1] and one subway station [buffer ratio 15% × 0.3 = 0.045] would have an initial factor of 0.145).

[0081] Step f: Through threshold truncation and normalization, the compensation factor is limited to a range of 0-1: the compensation factor for grids without any functional objects is set to 0; grids containing a single object with a weak impact have a factor of 0.1-0.4; grids containing multiple types of objects or highly sensitive facilities (such as tertiary hospitals and international airports) have a factor of up to 0.5-1. For example, a grid in the center of a city contains two hospitals (a total buffer of 40% × 0.4 = 0.16), a high-speed rail station (a buffer of 30% × 0.3 = 0.09), and three large shopping malls (density normalized 0.6 × 0.2 = 0.12). The cumulative factor is 0.37, which after normalization is 0.74, significantly increasing the weight of this area in the population density assessment.

[0082] Finally, the population density weight is combined with the hotspot compensation factor to calculate a correction for population density. The correction formula is: Region Type Population Coefficient = Population Density Weight × (1 + Hotspot Compensation Factor). This formula amplifies the weight of population influence in key areas through the compensation factor, avoiding the bias of traditional "pure population-based" assessment methods. For example, a grid with a medium population density (weight 0.6) but containing two schools and a subway station has a compensation factor of 0.5. The final coefficient is 0.6 × (1 + 0.5) = 0.9, which is comparable to the coefficient for a high-density area without a hotspot (0.9 × 1 = 0.9), demonstrating the equivalent effect of functional attributes on population density.

[0083] Through these steps, the regional population coefficient not only quantifies the underlying population distribution but also strengthens the impact assessment of disaster-sensitive areas through a hotspot compensation mechanism. This method resolves the issue of confusion between the assessment of "high-density areas with low functional importance" and "medium-density areas with high emergency demand." The generated regional population coefficient provides more accurate data on the demographic impact dimension for subsequent multi-source data fusion algorithms, ensuring that comprehensive disaster impact assessments take into account the multi-dimensional relationship between "people, facilities, and functions."

[0084] Furthermore, step A400 in the method provided in the embodiment of the present application includes:

[0085] A410: The multi-source data fusion algorithm includes tower offline rate weights, base station decommissioning rate weights, and population density coefficient weights, and is obtained by minimum error fitting based on historical samples of labeled disaster levels.

[0086] In one embodiment, the multi-source data fusion algorithm first includes weights for tower offline rates, base station decommissioning rates, and population density coefficients. These weights are determined based on historical samples of labeled disaster severity levels, using a minimum error fitting method. Specifically, a large amount of historical disaster data is collected, including tower offline rates, base station decommissioning rates, population density coefficients, and the corresponding actual disaster severity levels. For example, data on disasters such as earthquakes and floods that occurred in different regions over the past decade is collected. This data is recorded for each disaster, and technicians in this field label the disaster severity level, such as mild, moderate, or severe, based on the actual situation.

[0087] Then, with the goal of minimizing the error, a mathematical model is used to find the weight combination that minimizes the error between the predicted disaster severity level and the actual annotated disaster severity level. Assuming a linear weighted model, the formula for calculating the comprehensive disaster severity level can be expressed as: Comprehensive disaster severity level = Tower offline rate × Tower offline rate weight + Base station decommissioning rate × Base station decommissioning rate weight + Population density coefficient × Population density coefficient weight. By repeatedly adjusting the weights and calculating the error between the predicted and actual disaster severity levels, the weight values ​​that minimize the error are ultimately found. For example, after multiple iterative calculations, the error is minimized when the weight for Tower offline rate is 0.3, the weight for Base station decommissioning rate is 0.4, and the weight for Population density coefficient is 0.3.

[0088] Next, a comprehensive calculation is performed based on the identified regional types (such as power outage-dominated areas, equipment anomaly-dominated areas, and dual-impact areas) and regional population coefficients, combined with the previously determined weights. For each spatial grid, its tower offline rate, base station decommissioning rate, and population density coefficient are obtained. For example, if a grid has a tower offline rate of 40%, a base station decommissioning rate of 30%, and a population density coefficient of 0.8, then according to the above weights, the grid's comprehensive disaster severity level = 0.4 × 0.3 + 0.3 × 0.4 + 0.8 × 0.3 = 0.48.

[0089] Finally, based on the calculated comprehensive disaster severity level, each spatial grid is classified. Different thresholds can be pre-set, such as a comprehensive disaster severity level between 0-0.3 for mild damage, 0.3-0.7 for moderate damage, and 0.7 or above for severe damage. Based on this standard, all spatial grids are classified and a comprehensive disaster severity level distribution map is generated. Different disaster severity levels are represented by different colors or symbols in the map, allowing for a visual overview of the damage distribution across the affected area and providing a strong basis for subsequent rescue decisions and resource allocation.

[0090] By determining the weights through the above-mentioned minimum error fitting based on the historical samples of labeled disaster levels, and then integrating the regional types and regional type population coefficients, the final generated comprehensive disaster level distribution map can more accurately reflect the actual impact of the disaster and provide more scientific support for disaster response.

[0091] In summary, the post-earthquake disaster impact assessment method based on multi-source data fusion provided in the embodiments of the present application has the following technical effects:

[0092] This application pre-processes multi-source data through the inverse distance weighted interpolation algorithm and the spatial smoothing algorithm, and converts the power outage distribution map, equipment abnormality distribution map and population distribution heat map into a standardized spatial grid. The spatial grid is marked with regional types through threshold comparison, and the population density is calculated by combining the population density weight and the hot area compensation factor to obtain the regional type population coefficient. Based on the historical samples of the marked disaster level, the minimum error fitting is used to determine the tower offline rate weight, the base station decommissioning rate weight and the population density coefficient weight. The multi-source data fusion algorithm is used to combine the regional type and the regional type population coefficient to generate a comprehensive disaster level distribution map. Finally, the distribution map is classified according to the preset disaster level classification rules to determine the final disaster level, realize the accurate assessment and hierarchical management of the disaster impact, achieve the accurate assessment of the post-earthquake disaster situation, improve the accuracy of the disaster level classification, and provide a reliable basis for the rational allocation of emergency rescue resources and efficient rescue operations after the earthquake.

[0093] Example 2, as Figure 2As shown, based on the same inventive concept as the aforementioned embodiment 1, this embodiment of the present application provides a post-earthquake disaster impact assessment platform using multi-source data fusion, the system comprising:

[0094] The spatial grid obtains unit 1, which is used to perform spatial interpolation processing on the power outage distribution map, the equipment abnormality distribution map and the population distribution heat map to obtain the spatial grid.

[0095] The region type obtaining unit 2 is configured to mark the region type of the spatial grid to obtain the region type.

[0096] The population coefficient obtaining unit 3 is used to calculate the population density according to the regional type in combination with the population distribution heat map to obtain the regional type population coefficient.

[0097] The distribution map generating unit 4 is configured to combine the region type and the region type population coefficient based on a multi-source data fusion algorithm to generate a comprehensive disaster level distribution map.

[0098] The disaster level determination unit 5 is configured to classify the comprehensive disaster level distribution map according to a preset disaster level classification rule to determine a final disaster level.

[0099] Furthermore, the spatial grid obtaining unit 1 is used to perform the following steps:

[0100] Obtain the tower offline rate in the power outage area, extract the corresponding geographic coordinate information, and generate a power outage distribution map; obtain the base station decommissioning rate in the communication equipment function abnormality area, extract the corresponding geographic coordinate information, and generate an equipment abnormality distribution map; obtain the population thermal data of the densely populated area, extract the corresponding geographic coordinate information, and generate a population distribution heat map; perform spatial interpolation processing on the power outage distribution map, the equipment abnormality distribution map, and the population distribution heat map based on a spatial interpolation algorithm to obtain the spatial grid.

[0101] Furthermore, the spatial grid obtaining unit 1 is used to perform the following steps:

[0102] The spatial interpolation algorithm includes an inverse distance weighted interpolation algorithm, and the resolution of the spatial grid is set according to the spatial area and / or population density.

[0103] Furthermore, the region type obtaining unit 2 is configured to perform the following steps:

[0104] If the tower offline rate is higher than the first preset threshold and the base station deservice rate is lower than the second preset threshold, it is marked as a power outage dominated area; if the base station deservice rate is higher than the second preset threshold and the tower offline rate is lower than the first preset threshold, it is marked as an equipment abnormality dominated area; if the tower offline rate is higher than the first preset threshold and the base station deservice rate is higher than the second preset threshold, it is marked as a dual impact area; combine the power outage dominated area, the equipment abnormality dominated area and the dual impact area to obtain the area type.

[0105] Furthermore, the spatial grid obtaining unit 1 is used to perform the following steps:

[0106] If the power outage-dominated area and the equipment abnormality-dominated area are staggered, a spatial smoothing algorithm is used to smooth the power outage-dominated area and the equipment abnormality-dominated area to obtain an optimized power outage-dominated area and an optimized equipment abnormality-dominated area; the optimized power outage-dominated area and the optimized equipment abnormality-dominated area are added to the power outage-dominated area and the equipment abnormality-dominated area respectively.

[0107] Furthermore, the population coefficient obtaining unit 3 is used to perform the following steps:

[0108] According to the preset population density classification rules, the population thermal data of the population distribution heat map is classified to obtain the population density weight; based on the functional geographic object information of the spatial grid, a hot spot area compensation factor is generated; according to the population density weight and combined with the hot spot area compensation factor, the population density correction calculation of the area type is performed to obtain the population coefficient of the area type.

[0109] Furthermore, the distribution map generating unit 4 is configured to perform the following steps:

[0110] The multi-source data fusion algorithm includes tower offline rate weights, base station decommissioning rate weights and population density coefficient weights, and is obtained by minimum error fitting based on historical samples of labeled disaster levels.

[0111] A post-earthquake disaster impact assessment platform with multi-source data fusion provided by an embodiment of the present invention can execute a post-earthquake disaster impact assessment method with multi-source data fusion provided by any embodiment of the present invention, and has functional modules and beneficial effects corresponding to the execution method.

[0112] Although the present application makes various references to certain modules in the system according to the embodiments of the present application, any number of different modules may be used and run on the user terminal and / or server, and the various units and modules included are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional units are only for the convenience of distinguishing each other and are not used to limit the scope of protection of the present invention.

[0113] The above specific embodiments do not constitute a limitation to the scope of protection of this application. It should be understood by those skilled in the art that various modifications, combinations and substitutions can be made according to design requirements and other factors. Any modifications, equivalent replacements and improvements made within the spirit and principles of this application should be included in the scope of protection of this application. In some cases, the actions or steps recorded in this application can be performed in an order different from that in the embodiments and can still achieve the desired results. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

Claims

1. A post-earthquake disaster impact assessment method based on multi-source data fusion, characterized in that: include: Perform spatial interpolation processing on the power outage distribution map, equipment anomaly distribution map, and population distribution heat map to obtain a spatial grid; Marking the spatial grid with a region type to obtain a region type; Combined with the population distribution heat map, the population density is calculated according to the regional type to obtain the regional type population coefficient; Combining the regional type with the regional type population coefficient based on a multi-source data fusion algorithm to generate a comprehensive disaster level distribution map; According to the preset disaster grade classification rules, the comprehensive disaster grade distribution map is classified to determine the final disaster grade.

2. The post-earthquake disaster impact assessment method based on multi-source data fusion according to claim 1, characterized in that: Perform spatial interpolation processing on the power outage distribution map, equipment anomaly distribution map, and population distribution heat map to obtain a spatial grid, including: Obtain the tower offline rate in the power outage area, extract the corresponding geographic coordinate information, and generate a power outage distribution map; Obtain the base station outage rate in areas with abnormal communication equipment functions, extract the corresponding geographic coordinate information, and generate an equipment abnormality distribution map; Obtain population thermal data of densely populated areas, extract corresponding geographic coordinate information, and generate a population distribution heat map; The power outage distribution map, the equipment anomaly distribution map, and the population distribution heat map are spatially interpolated based on a spatial interpolation algorithm to obtain the spatial grid.

3. The post-earthquake disaster impact assessment method based on multi-source data fusion according to claim 2, characterized in that: The spatial interpolation algorithm includes an inverse distance weighted interpolation algorithm, and the resolution of the spatial grid is set according to the spatial area and / or population density.

4. The post-earthquake disaster impact assessment method based on multi-source data fusion according to claim 2, characterized in that: Marking the spatial grid with a region type to obtain the region type includes: If the tower offline rate is higher than a first preset threshold and the base station out-of-service rate is lower than a second preset threshold, it is marked as a power outage dominant area; If the base station decommissioning rate is higher than the second preset threshold and the tower offline rate is lower than the first preset threshold, it is marked as a device abnormality dominant area; If the tower offline rate is higher than the first preset threshold and the base station decommissioning rate is higher than the second preset threshold, it is marked as a double-impact area; The area type is obtained by combining the power outage dominant area, the equipment abnormality dominant area, and the dual impact area.

5. The post-earthquake disaster impact assessment method based on multi-source data fusion according to claim 2, characterized in that: Obtaining the spatial grid includes: If the power outage dominant region and the equipment abnormality dominant region are staggered, a spatial smoothing algorithm is used to smooth the power outage dominant region and the equipment abnormality dominant region to obtain an optimized power outage dominant region and an optimized equipment abnormality dominant region; The optimized power outage dominant region and the optimized equipment abnormality dominant region are added to the power outage dominant region and the equipment abnormality dominant region respectively.

6. The post-earthquake disaster impact assessment method based on multi-source data fusion according to claim 1, characterized in that: Combined with the population distribution heat map, the population density is calculated according to the regional type to obtain the regional type population coefficient, including: Classifying the population thermal data of the population distribution heat map according to a preset population density classification rule to obtain a population density weight; generating a hotspot area compensation factor based on the functional geographic object information of the spatial grid; According to the population density weight and in combination with the hotspot area compensation factor, a population density correction calculation is performed on the area type to obtain the population coefficient of the area type.

7. The multi-source data fusion post-earthquake disaster impact assessment method according to claim 1, characterized in that: The multi-source data fusion algorithm includes tower offline rate weights, base station decommissioning rate weights and population density coefficient weights, and is obtained by minimum error fitting based on historical samples of labeled disaster levels.

8. A multi-source data fusion post-earthquake disaster impact assessment platform, characterized by: A post-earthquake disaster impact assessment method for implementing the multi-source data fusion method according to any one of claims 1 to 7, the system comprising: The spatial grid is used to obtain cells, which are used to perform spatial interpolation processing on the power outage distribution map, the equipment abnormality distribution map and the population distribution heat map to obtain the spatial grid; A region type obtaining unit, configured to mark the spatial grid with a region type to obtain a region type; A population coefficient obtaining unit is used to calculate the population density according to the regional type in combination with the population distribution heat map to obtain the regional type population coefficient; a distribution map generating unit, configured to combine the regional type and the regional type population coefficient based on a multi-source data fusion algorithm to generate a comprehensive disaster level distribution map; The disaster level determination unit is used to classify the comprehensive disaster level distribution map according to a preset disaster level classification rule to determine a final disaster level.

Images