Post-disaster regional operation risk assessment method based on Beidou technology

Abstract

The invention relates to the technical field of risk assessment based on big data, and provides a post-disaster regional operation risk assessment method based on a Beidou technology, and the method comprises the steps: building an impact analysis model based on disaster types through multi-dimensional information, and combining with the condition of a target work site, so as to assess the risk of post-disaster field operation. In order to achieve the purpose, the post-disaster regional operation risk assessment method based on the Beidou technology is provided and comprises the steps that daily multi-source data collection and standardization processing are carried out on a target region, and indoor and outdoor scene dynamic judgment models are established respectively for risk index calculation; and after the disaster occurs, detecting the target area again, comparing the detection result with the comprehensive risk index, and accurately evaluating the risk again. According to the method, after-calamity risk assessment is converted from fuzzy judgment to quantitative accuracy, operation preparation is converted from experience driving to data driving, and direction guidance is provided for after-calamity safe and efficient operation in the power industry through risk assessment.

Description

Technical Field

[0001] The present invention relates to the technical field of risk assessment based on big data, and more specifically, to a post-disaster regional operation risk assessment method based on Beidou technology. Background Art

[0002] Even in extreme situations, power systems cannot guarantee safety. Existing technologies require post-disaster power equipment reconstruction, a core requirement for safeguarding people's livelihoods, industrial operations, and emergency rescue efforts. Rapid power restoration is essential to prevent the spread of secondary disasters and maintain social order. Standardized solutions, such as drone damage assessment, rapid grid connection of modular equipment, disaster-resistant tower reinforcement, and mechanized construction, have proven effective in achieving a closed loop of "rapid repair, safe reconstruction, and long-term disaster preparedness." Power disaster sites are densely populated, and repair teams are complex. Obtaining personnel location information is challenging, and collaboration can easily lead to safety risks. During power emergency response, the complex on-site environment makes it difficult to access information on temporary hazards and various resources. Emergency personnel have limited access to on-site information and support. Furthermore, poor communication between the emergency command center and the site hinders the efficiency of on-site repairs. Of particular note, natural disasters can introduce new safety hazards to power equipment and its surrounding facilities. If these hazards are not promptly identified and addressed, accidents are inevitable. Although there are corresponding solutions with Beidou technology in the existing technology, these solutions cannot cover indoor operations.

[0003] In this regard, patent CN201711379457.8, entitled "A Method for Micrometeorological Observation Site Selection for Wind Disaster Monitoring of Transmission Lines in Complex Terrain," discloses a method. Based on statistical analysis of historical disaster data, this method proposes a rule-based spatial overlay method. This method integrates wind field simulation at different altitudes within a mesoscale weather forecast model with spatial analysis methods from a geographic information system (GIS) to select micrometeorological observation sites for transmission lines within a region. Using site observations and model simulations, the method dynamically monitors the distribution of wind fields at different altitudes within the region in real time, monitors the distribution of wind disasters affecting transmission line towers, and provides technical support for ensuring safe power operation. This method forms a comprehensive site selection process that combines indoor comprehensive simulation analysis with field surveys. This method effectively addresses the issue of unpredictable power outages at tower stations following wind disasters affecting transmission lines within a region. This novel invention offers highly representative site selection results, providing technical support for the deployment of a high-wind meteorological disaster monitoring network for transmission lines and is easily scalable and implementable.

[0004] As can be seen, the patent has already taken into account the analysis and model calculation of scenarios after natural disasters, and also considers different indoor and outdoor conditions. However, this solution has significant limitations. First, it can only be applied to complex terrain, which is a significant limitation. Second, it can only determine wind disaster monitoring and cannot consider or judge multi-parameter disaster fusion scenarios. Most importantly, the solution can only perform single predictions. After actual on-site operations based on the prediction content, it is impossible to determine whether the previous operation method is the optimal solution, thus making it impossible to further optimize the previous model. Summary of the Invention

[0005] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides a post-disaster regional operation risk assessment method based on Beidou technology, which establishes an impact analysis model for different disaster types through multi-dimensional information, and combines the conditions of the target work site to assess the risks of post-disaster on-site operations.

[0006] To achieve the above objectives, the present invention provides a method for post-disaster regional operation risk assessment based on BeiDou technology, comprising the following steps: Step 1: Collect and standardize daily multi-source data in the target area; Step 2: Establish dynamic judgment models for indoor and outdoor scenes respectively; Step 3: Based on the early warning of different disaster types, calculate the risk index for indoor and outdoor areas respectively; Step 4: Perform compound disaster coupling calculation for related disasters to obtain the comprehensive risk index of compound disasters in corresponding indoor or outdoor scenarios; Step 5: Conduct preparatory work based on the comprehensive risk index assessment; Step 6: After the disaster occurs, the target area is inspected again and the inspection results are compared with the comprehensive risk index obtained in step 4 to achieve data closure and model correction, thereby re-assessing the risk accurately.

[0007] Preferably, in step 1, the data collection scope is determined based on all power facilities in the target area, and information inside the power facilities and the surrounding environment within a specified range centered on the power facilities are collected. The collection frequency is to obtain parameters daily, and to obtain parameters frequently after receiving a disaster forecast.

[0008] Preferably, basic environmental data, meteorological data and equipment status data are obtained based on the positioning of power facilities; the outdoor data in the basic environmental data is obtained through Beidou static positioning coordinates, and the topography, geological parameters, and external building attributes are obtained, and the indoor data in the basic environmental data is obtained by indoor Bluetooth devices; meteorological data is real-time information such as rainfall, wind speed, temperature, magnitude, ice thickness, and historical disaster data obtained through Beidou short messages and meteorological satellites; the equipment status data is the operating parameters and security status of all power equipment in the area.

[0009] Preferably, non-quantitative data is converted into numerical indicators during the data acquisition process; all parameters are normalized to eliminate dimensional differences, and new data is updated in real time; all data are associated with device ID and geographic coordinates to ensure data traceability.

[0010] Preferably, in step 2, real-time data is used to distinguish the indoor and outdoor attributes of power facilities in the target area, a weight coefficient α is introduced, and outdoor Beidou data and indoor Bluetooth sensor data are dynamically integrated to establish indoor and outdoor switching thresholds. The formula is "scene determination result = α × outdoor eigenvalue + (1-α) × indoor eigenvalue", ensuring continuous and seamless scene determination.

[0011] Preferably, in steps 3 and 4, differentiated indoor and outdoor calculation logics are designed for the four core disasters of floods, typhoons, earthquakes, and ice disasters, respectively. Through historical data and real-time warnings, disaster combinations with synergistic effects are identified and coupled calculations are performed.

[0012] Preferably, in step 5, preparatory work is carried out based on the comprehensive risk index, and different risk levels are divided according to the index and marked on the map, and different entry strategies are deployed for different risk levels. In high-risk areas, emergency equipment is deployed within 3km first, and emergency repair materials are stored for ≥3 days. Two or more emergency repair teams are deployed and trained, and restricted areas are designated, detour routes are developed, and equipment is reinforced and sealed. In medium-risk areas, preventive tests are conducted on key equipment, and vulnerable parts are stored at 1.5 times the normal level. A phased maintenance plan and indoor operation safety inspection process are developed. Low-risk areas maintain real-time monitoring and update the index every 12 hours to ensure that vehicles and communication equipment are in good condition without the need for additional special deployment.

[0013] Preferably, in step 6, the pre-disaster comprehensive risk index is compared with the actual post-disaster damage situation, and the disaster model and coupling coefficient are revised; the entry strategy during maintenance is adjusted based on the revised disaster model and coupling coefficient. This step can directly improve the accuracy and effectiveness of pre-disaster preparations by comparing the pre-disaster prediction with the actual post-disaster situation and revising the disaster model and coupling coefficient. On the one hand, by quantifying the deviation, the risk level classification logic can be optimized, so that the resource allocation of "high / medium / low risk areas" in pre-disaster preparations is more in line with actual needs, avoiding waste or shortage of resources due to misjudgment; on the other hand, the revised model can more accurately capture the synergistic effect of disasters, making the preparatory plan for complex disasters more targeted and reducing safety hazards.

[0014] Preferably, the method further includes step 7: based on comparative analysis of actual data obtained after the disaster occurs and the predicted results, the model parameters are corrected to optimize the accuracy of risk determination.

[0015] Preferably, the revised model is verified and iterated, and by establishing a data benchmark library of disaster type fusion scenarios, a graded correction rule table divided by the degree of deviation is formulated, and a standard template suitable for different disasters and indoor and outdoor scenarios is designed.

[0016] The present invention brings the following beneficial effects through the technical solution: 1. Multi-source data fusion and precise standardization, centered around power facilities, integrates Beidou positioning, indoor Bluetooth data, meteorological satellite information, equipment status parameters, and historical disaster data to build a multi-dimensional data system. By digitizing non-quantitative data and normalizing parameters, linking device IDs with geographic coordinates, we ensure data traceability and eliminate dimensional discrepancies, laying a precise data foundation for risk assessment while balancing efficiency and real-time performance.

[0017] 2. Achieve dynamic adaptation and seamless switching between indoor and outdoor scenes. Design a scene determination model based on real-time data. Distinguish indoor and outdoor attributes through features such as the number of satellites and Bluetooth signal strength. Introduce a weight coefficient α to achieve dynamic fusion of "outdoor Beidou data + indoor Bluetooth / sensor data" to solve the problem of scene faults in complex environments and ensure that risk assessments accurately match actual scenes.

[0018] 3. Achieve coordinated assessment of disaster types and complex disasters. Design differentiated indoor and outdoor risk index calculation logic for the four core disasters of floods, typhoons, earthquakes, and ice disasters. At the same time, identify related disasters through historical data and real-time warnings, introduce coupled calculation models to quantify synergistic effects, and achieve comprehensive coverage from single disasters to complex disasters, avoiding the one-sidedness of traditional single disaster assessments.

[0019] 4. Implement tiered, precise preparation and dynamic operational guidance. Based on a comprehensive risk index, high, medium, and low risk levels are defined, with corresponding differentiated preparation strategies. Map identification and tiered strategies enable precise resource allocation and operational safety management.

[0020] 5. Achieve data closure and model iterative optimization, and build a detection-based logical closure: compare post-disaster detection results with pre-disaster comprehensive risk indexes, quantify deviations, and correct disaster model parameters and coupling coefficients; continuously optimize model accuracy by establishing a disaster-scenario data benchmark library, a hierarchical correction rule table, and a standard template to ensure that assessment capabilities dynamically improve as data accumulates.

[0021] In summary, the technical solution described in this invention achieves the transformation of post-disaster risk assessment from fuzzy judgment to quantitative precision, and operational preparation from experience-driven to data-driven through the deep integration of Beidou technology and multi-dimensional data. Through risk assessment, it provides directional guidance for safe and efficient post-disaster operations in the power industry. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] Figure 1 This is a flow chart of the post-disaster regional operation risk assessment method based on Beidou technology of the present invention; Figure 2 Schematic diagram of indoor and outdoor switching of the post-disaster regional operation risk assessment method based on Beidou technology of the present invention; Figure 3 Schematic diagram of indoor status detection during prediction of the post-disaster regional operation risk assessment method based on Beidou technology of the present invention; Figure 4 This is a schematic diagram of indoor risks obtained through the post-disaster regional operation risk assessment method based on Beidou technology of the present invention. DETAILED DESCRIPTION

[0023] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0024] While existing technologies within the power industry's disaster emergency response system have established a comprehensive safety control loop encompassing risk prevention, process management, and emergency preparedness, significant technical deficiencies remain. Current systems lack dynamic recognition of the on-site environment, particularly in areas with blurred boundaries between indoor and outdoor scenes, such as partially damaged buildings, lacking effective real-time identification mechanisms. More critically, existing risk assessment methods primarily rely on disaster prediction data, resulting in significant discrepancies between the actual disaster situation and the actual severity of the damage. Furthermore, waiting until the disaster is fully resolved before conducting assessments and deploying emergency resources can significantly delay repairs. This conflict between inaccurate predictions and delayed post-disaster response results in the system's inability to accurately pre-position resources before a disaster occurs, and difficulty ensuring the rapid deployment of emergency resources after a disaster. Currently, the industry lacks a technical solution that balances both accurate assessments and timely responses, a gap that directly impacts the efficiency and safety of power facility repairs.

[0025] like Figure 1 As shown, to solve the above problem, the present invention provides a post-disaster regional operation risk assessment method based on BeiDou technology, comprising the following steps: Step 1: Daily multi-source data collection and standardization are conducted within the target area. The data collection scope is determined based on all power facilities within the target area. Information is collected from the power facility interior and the surrounding environment within a specified distance from the power facility. Parameters are collected daily, and more frequently after receiving disaster forecasts. During the data acquisition process, non-quantitative data is converted to numerical indicators. All parameters are normalized to eliminate dimensional differences, and new data is updated in real time. All data is linked to device IDs and geographic coordinates to ensure traceability.

[0026] Step 2: Develop dynamic judgment models for indoor and outdoor scenarios. Using real-time data, we distinguish the indoor and outdoor attributes of power facilities within the target area. We introduce a weighting factor, α, and dynamically integrate outdoor Beidou data with indoor Bluetooth sensor data to establish switching thresholds between indoor and outdoor scenarios. The formula is "Scene Judgment Result = α × Outdoor Eigenvalue + (1-α) × Indoor Eigenvalue," ensuring seamless and continuous scene judgment.

[0027] Step 3: Based on the early warning of different disaster types, calculate the risk index for indoor and outdoor areas respectively; Here, we design a disaster classification mathematical model that integrates multi-source data, referring to the regression analysis and neural network frameworks in big data. The details are as follows: 1. Taking rainstorm combined with flood as an example, a single disaster type impact analysis model is established.

[0028] Input variables: Meteorological parameters: 24-hour rainfall R (mm), rainfall duration T (h), peak water level H (m) Site parameters: Altitude E (m), slope S (°), soil moisture content W (%), drainage capacity D (mm / h) Model formula: Danger index

[0029] Weight It is obtained by training the reference linear regression equation based on the historical data of rainstorm disaster cases in the past 10 years.

[0030] Threshold: A hazard index greater than or equal to 0.8 indicates high risk, between 0.5 and 0.8 indicates medium risk, and otherwise indicates low risk. It should be noted that the hazard index proposed here is based on existing safety specifications. For example, the "GB / T28588-2012 Guidelines for Emergency Management of Power System Disasters" clearly divides power disasters into four levels: extremely serious (level I), serious (level II), relatively serious (level III) and general (level IV). The "DL / T1301-2013 Specification for the Preparation of Emergency Plans for Power System Emergencies" The corresponding relationship between the quantitative indicators of equipment damage and geographical environment parameters has been further supplemented. In addition, there is also "Q / GDW11838-2018 Power Grid Meteorological Disaster Warning Classification". It is based on these standards that the judgment conditions and implementation plans for indoor and outdoor scenarios are generated as shown in the following table. 1. Flood or heavy rain scenarios

[0031] 2. Typhoon scene

[0032] 3. Earthquake scenario

[0033] 2. Take the coupled disaster model of typhoon combined with heavy rain as an example; Input variables: Typhoon parameters: maximum wind speed V (m / s), gust coefficient Cg; Heavy rain parameters: R, T, H are the same as those of the single-hazard model.

[0034] Coupling formula:

[0035] The wind resistance threshold of the equipment.

[0036] The coupling coefficient is used to reflect the synergistic effect of wind and rain, and is fitted through historical composite disaster data, such as

[0037] =0.002.

[0038] As weight, when typhoon dominates ,otherwise .

[0039] 3. Model Iteration Optimization Mechanism Correction Factor: ; are model parameters (such as α, β, λ), and η is the learning rate (0.01-0.1); Actual represents the actual post-disaster risk, such as the equipment damage rate; Predict represents the model output; and X represents the input feature vector. After each post-disaster operation, the parameters are updated with new data to improve prediction accuracy.

[0040] For single disasters, a gradient boosting tree is used to handle nonlinear relationships, such as the threshold effect between rainfall and waterlogging. For complex disasters, a neural network is used to learn coupling terms. Both solutions utilize edge computing to deploy lightweight models, ensuring millisecond-level response. This model is trained using quantitative parameters and historical data to achieve computable and iterative hazard zone classification, adhering to the document's closed-loop logic of "prediction-actual comparison-model refinement."

[0041] Step 4: Perform compound disaster coupling calculations for related disasters to obtain a comprehensive risk index for compound disasters in corresponding indoor or outdoor scenarios. In the following embodiments, four core disasters, namely floods, typhoons, earthquakes, and ice disasters, are mainly targeted. Differentiated calculation logic is designed for indoor and outdoor use. Through historical data and real-time warnings, disaster combinations with synergistic effects are identified and coupled calculations are performed.

[0042] Step 5: Conduct preparatory work based on the comprehensive risk index assessment; Step 6: After the disaster occurs, the target area is inspected again and the inspection results are compared with the comprehensive risk index obtained in step 4 to achieve data closure and model correction, thereby re-assessing the risk accurately.

[0043] The following is a specific embodiment of the present invention: taking a specific substation in a certain area as an example: Step 1: Daily multi-source data collection and standardization of substations Data collection scope: Covers all power facilities within the substation, including the substation building, transmission lines, distribution rooms, and surrounding environment. For power facilities, equipment status data is obtained, including power equipment operating parameters and security status.

[0044] Basic positioning and environmental data includes Beidou static positioning coordinates, dynamic trajectories with a sampling frequency of 1Hz; indoor Bluetooth AOA base station signal strength and multi-axis sensor data; and topographical features, geological parameters, building wind resistance ratings, waterproof thresholds, and seismic intensity. After determining an area, meteorological and disaster warning data for that area is obtained: real-time warning information such as rainfall, wind speed, temperature, earthquake magnitude, and ice thickness is obtained through Beidou short messages and meteorological satellites. Historical disaster data is also retrieved, typically showing the impact range and equipment damage rate of similar disasters over the past 10 years.

[0045] Convert non-quantitative data into numerical indicators, such as "IP65 waterproof rating" corresponds to a water depth threshold of 0.5m, "soft soil foundation" is assigned a value of 1.2, and "hard soil foundation" is assigned a value of 0.8; All parameters are normalized (mapped to the interval [0,1]) to eliminate dimensional differences (such as converting wind speed into the ratio of "actual wind speed / equipment wind resistance threshold"). A dynamic database is established to update new data in real time, and the equipment ID and geographic coordinates are associated to ensure data traceability.

[0046] Step 2: Establish dynamic judgment models for indoor and outdoor scenes respectively like Figure 2 As shown in the figure, real-time data is used to distinguish the "indoor" and "outdoor" attributes of power facilities in the substation, providing a scenario basis for subsequent risk calculations.

[0047] The judgment basis and logic are as follows: Outdoor scenarios are determined based on positioning signals: ≥4 Beidou / GNSS satellites, stable RTK differential data reception, and positioning accuracy ≤10cm. This is combined with environmental characteristics, including the absence of obstructions from enclosed structures such as transmission towers and outdoor switch stations, and terrain such as open areas, mountains, and river valleys. Identifiable equipment, such as towers and cable trenches, is labeled "outdoor equipment" and marked for any indoor protective features, such as waterproofing and anti-smash features.

[0048] like Figure 3 As shown, indoor scene determination includes positioning signals: the number of satellites is less than 3, the Bluetooth AOA signal strength is ≥-70dBm, and the "Beidou + Bluetooth + multi-axis sensor" hybrid positioning is enabled. The environment is characterized by enclosed structures such as power distribution rooms, substation control rooms, and underground cable shafts, with obstructions such as walls and roofs.

[0049] Identifiable equipment, such as switchgear and battery packs, is labeled "indoor equipment" and has an IP65 waterproof rating and protection against electric shock. For semi-enclosed areas, such as collapsed factory buildings and covered outdoor equipment areas, a weighting factor α is introduced. The weighting factor is the proportion of outdoor signals, with a value of 0 < α < 1. Outdoor Beidou data is dynamically integrated with indoor Bluetooth / sensor data using the formula "scene determination result = α × outdoor eigenvalue + (1-α) × indoor eigenvalue," ensuring seamless scene determination.

[0050] Step 3: Calculate risk indices for indoor and outdoor areas based on early warnings based on disaster types Adaptation by disaster type and scenario: Differentiated calculation logic is designed for indoor and outdoor scenarios for the four core disasters of floods, typhoons, earthquakes, and ice disasters. The risk index is represented by a numerical value in the [0,1] range. The higher the value, the higher the risk.

[0051] Specific calculation logic: Floods / Heavy Rain: Outdoor: Based on the calculation of "rainfall / drainage capacity", "peak water level - site elevation" and "soil moisture content", focus on the risk of tower foundation scour and line inundation: "Risk index = 0.4 × (rainfall / drainage capacity) + 0.3 × (peak water level - elevation) / historical highest water level + 0.3 × soil moisture content"; Indoor: Based on calculations based on "water depth / waterproof threshold," "difference between drainage rate and water rise rate," and "equipment insulation damage probability," focus on the risks of leakage and equipment short circuits in distribution rooms. Outdoor: Calculated based on "wind speed × gust coefficient / tower wind resistance level" and "impact duration / tolerance duration", focus on the risks of tower overturning and line disconnection, "risk index = 0.6 × (wind speed × gust coefficient / wind resistance level) + 0.4 × (impact duration / 24h)".

[0052] Indoor: Calculated based on "actual wind speed / building wind resistance rating" and "door and window damage rate", focus on the risk of equipment damage caused by wall deformation and sudden changes in air pressure: "Risk index = 0.7×(wind speed / building wind resistance rating)+0.3×door and window damage rate".

[0053] earthquake: Outdoor: Calculated based on "magnitude / 8," "epicenter distance / 50km," and "equipment tilt angle / safety threshold," combined with the soil coefficient (1.2 for soft soil and 0.8 for hard soil). Focus on the risks of tower collapse and landslide burial: "Risk index = 0.5 × (magnitude / 8) × exp (-epicenter distance / 50) + 0.5 × (tilt angle / 3°) × soil coefficient." Indoor: Calculated based on "wall crack width / 5mm", "load-bearing structure stress / design limit" and "equipment falling probability", focus on the risk of building collapse and equipment damage: "Risk index = 0.6×(crack width / 5mm)+0.3×(stress / limit value)+0.1×fall probability".

[0054] Ice disaster: Outdoor: Calculated based on "ice thickness / conductor tolerance thickness", "air temperature / -10°C", and "low temperature duration / 24 hours", with a focus on line disconnection and insulator ice flashover risks: "Risk index = 0.6 × (ice thickness / tolerance value) + 0.2 × (-air temperature / 10) + 0.2 × (low temperature duration / 24 hours)"); Indoor: Based on the calculation of "equipment operating temperature / lower limit threshold" and "heating system failure probability", focus on the risks of battery freezing damage and pipe rupture: "Risk index = 0.7×(lower limit temperature-actual temperature) / 20+0.3×heating failure probability".

[0055] Step 4: Perform compound disaster coupling calculations for related disasters to obtain a comprehensive risk index Identification of related disasters: Through historical data and real-time warnings, we can identify disaster combinations with synergistic effects. The judgment criteria are "the interval between two or more disasters is ≤6 hours and the overlap of the affected areas is ≥50%."

[0056] During the coupled calculation, the indoor or outdoor risk index of a single disaster in the corresponding scenario is first calculated. This step is based on the results of step 3. The coupling coefficient λ is introduced, ranging from 0.1 to 0.3, and is set based on the synergistic intensity of the disasters. For example, for a typhoon plus a rainstorm, λ = 0.2, and for an earthquake plus a landslide, λ = 0.3, this quantifies the mutual amplification effect between disasters. The comprehensive risk index is then calculated as Σ (single-hazard risk index × weight) + λ × Π (single-hazard risk index), where weights are assigned based on the dominance of the disaster.

[0057] Example: For a combined disaster of outdoor typhoon and rainstorm, the typhoon risk index is 0.7, the rainstorm risk index is 0.6, and the coupling coefficient is 0.2. The comprehensive risk index = 0.4×0.7+0.4×0.6+0.2×0.7×0.6=0.28+0.24+0.084=0.604, which represents the overall degree of danger after the superposition of the two disasters.

[0058] Step 5: Conduct preparatory work based on comprehensive risk index assessment Risk level classification: Based on the comprehensive risk index, three levels are defined: high risk (≥0.8), medium risk (0.5~0.8), and low risk (<0.5). Figure 4As shown, different risk levels are marked with different colors. The white area is marked as outdoor, and it switches after passing through the green door. Whether it is a high-risk area or a low-risk area, targeted preparatory measures are designed for rescue and maintenance.

[0059] Targeted preparatory measures in high-risk areas include resource allocation, prioritizing the deployment of emergency equipment within a 3km radius; stockpiling at least three days' worth of repair supplies; arranging two or more repair teams on standby; and ensuring that team members complete pre-emptive training for indoor and outdoor operations, wear smart helmets with tracking, and use Beidou short message terminals. No-entry zones and detour routes are established; outdoor equipment is reinforced in advance, and indoor equipment is waterproofed and sealed.

[0060] Equipment inspections are conducted in medium-risk areas: Key equipment, such as transformers and circuit breakers, undergo preventive insulation resistance and mechanical property tests. Consumable parts are also stockpiled at 1.5 times the normal usage.

[0061] Operation planning: Develop a phased maintenance plan to repair the main line first, then restore the branch line, and clarify the structural safety inspection process for indoor operations.

[0062] In low-risk areas, targeted preparatory measures include daily monitoring, maintaining real-time data transmission from Beidou positioning and sensors, and updating the risk index every 12 hours. They also ensure that emergency repair vehicles have sufficient fuel and communication equipment is fully charged. No additional special deployment is required.

[0063] Step 6: Re-inspect the substation after the disaster Testing timing: Start the first round of testing within 1 hour after the disaster ends, and complete full coverage within 24 hours.

[0064] Testing content and means: Environmental and topographic changes: Using drone aerial photography and Beidou positioning, we identify new risk points, such as buried areas after landslides and exposed tower foundations washed out by floodwaters. We use LiDAR to scan indoor building structures to detect post-disaster wall crack expansion and floor slab settlement. We use intelligent inspection robots and handheld terminals equipped with Beidou positioning to monitor equipment parameters. For indoor equipment in flooded areas, we focus on testing whether the watertightness threshold has been breached. During preparation, we conduct communication and positioning verification, testing the communication quality of Beidou short message terminals and 4G / 5G emergency base stations to ensure continuous positioning both indoors and outdoors.

[0065] Compare the pre-disaster comprehensive risk index with the actual post-disaster damage, and revise the disaster model and coupling coefficient; update the boundaries of the dangerous area based on the new detection data, and adjust the preparatory operation plan in step 5 to achieve the reorganization of the plan.

[0066] The above steps are sufficient to complete the assessment of the operational risks in the post-disaster area. However, in order to provide services more efficiently and accurately, the present invention also includes step 7: Based on the comparative analysis of actual data and predicted results, the model parameters are modified to optimize the risk determination accuracy. Core goal: Through quantitative comparison of actual post-disaster disaster data with pre-disaster model prediction results, identify model deviations, adjust parameters in a targeted manner, form a closed loop of "prediction-practice-correction", and continuously improve the accuracy of subsequent risk assessments.

[0067] 1. Actual Data Collection and Standardization Equipment damage data is collected to calculate the actual damage to power equipment in the target area, including the number of collapsed towers, the proportion of short-circuited switchgear, and the length of cable insulation failure. This data is categorized as "total damage / partial damage / intact" and converted into quantitative indicators, where total damage = 1, part damage = 0.5, and intact = 0. Post-disaster monitoring also provides operational feedback on actual risk parameters: The deviation between predicted and actual risk is recorded during the repair process. For example, if the model predicts low risk but actual equipment damage occurs, this is marked as a "missed prediction"; if a high risk prediction has no significant impact, this is marked as a "false positive." The actual data is then linked to the pre-disaster model's predicted objects (e.g., specific towers or distribution rooms) using their IDs and geographic coordinates to ensure a one-to-one correspondence between prediction and actual risk.

[0068] 2. Comparative Analysis of Forecast Results and Actual Data Risk index error: Calculate the absolute error |Rpred-Ract| and relative error |Rpred-Ract| / Ract between the "predicted risk index Rpred" and "actual risk index Ract" for single or combined disasters. Ract is inferred based on the actual damage rate and on-site parameters. For example, Ract = 0.9 for a fully damaged device and 0.6 for a half-damaged device. Regional consistency: Compare the spatial overlap between the high / medium / low risk areas predicted by the model and the actual disaster-affected areas. An overlap of less than 60% is considered a significant deviation. False positive rate statistics: count the “number of missed positives” and “number of false positives” and calculate the false positive rate.

[0069] Typical comparison scenario: For example, if the model predicts a risk index of an outdoor tower as Rpredicted = 0.6 during a combined typhoon and rainstorm disaster, indicating a medium risk, but the tower actually collapses due to water accumulation and strong winds, with Ractual = 0.9, indicating a high risk. The absolute error is 0.3, and it is necessary to analyze whether the coupling coefficient or the single disaster weight is too low.

[0070] If the indoor distribution room model predicts R = 0.7, indicating medium risk, but the actual water depth does not exceed the waterproof threshold R = 0.3, you need to check whether the "water depth weight" in the indoor flood model is too high.

[0071] 3. Model parameter correction rules and methods Single-hazard model weight: If the actual risk of a particular hazard continues to be higher than the predicted value, the weight is adjusted according to the formula "adjusted weight = original weight × (1 + average relative error)". For example, if the original weight is 0.5 and the average relative error is 0.3, the adjusted weight = 0.5 × 1.3 = 0.65.

[0072] Composite disaster coupling coefficient λ: If the actual risk of a composite disaster is significantly higher than the combined value of a single disaster, adjust it according to "λnew = λold × (1 + composite error mean)". For example, if the original λ = 0.2 and the composite error mean is 0.4, then λnew = 0.2 × 1.4 = 0.28.

[0073] Indoor and outdoor scene determination thresholds: If "scene misjudgment" frequently occurs in a certain area, adjust the satellite number threshold or Bluetooth signal strength threshold. After adjusting the parameters, ensure that the risk index remains in the [0, 1] range and that the sum of the individual disaster weights is 1 to avoid model logic failure due to over-correction.

[0074] 4. Verification and Iteration of the Revised Model Verification method: The revised model is tested using a recent disaster case or random samples from historical data that were not included in the revision, and new error metrics are calculated. If the mean absolute error drops from 0.2 to below 0.15 and the regional consistency improves to above 80%, the revision is considered effective. Simulations are conducted, inputting historical disaster data, and comparing the consistency of the model's risk assessment results before and after revision with the actual disaster situation to ensure the correct direction of the revision. Initial parameter revisions are completed within 72 hours after each disaster, and batch optimization is performed monthly using multiple batches of data. A model version library is established to record parameter changes, reasons, and verification results for each revision, facilitating the tracing of applicable scenarios for different versions.

[0075] 5. Output and Application Generate a correction report to clarify the cause of the deviation, the adjusted parameter values ​​and the scope of application; synchronize the corrected model to the edge computing device and cloud system for the next disaster warning and risk assessment to ensure that the optimization effect is implemented in real time.

[0076] Through this step, the model's risk assessment error for single disasters can be reduced by more than 30%, and the regional consistency of complex disasters can be increased to more than 85%, providing a more accurate quantitative basis for subsequent preparatory work.

[0077] Those skilled in the art will appreciate that the modules and algorithm steps of the various examples disclosed herein can be implemented using electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application of the technical solution and the constraints of the invention. Professionals and technicians may use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of the present invention.

[0078] In addition, each functional module in each embodiment of the present invention may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module.

[0079] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any modifications or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in the present invention should be included in the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be based on the scope of protection of the claims.

[0080] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A post-disaster regional operation risk assessment method based on BeiDou technology is characterized by: The steps include: Step 1: Collect and standardize daily multi-source data in the target area; Step 2: Establish dynamic judgment models for indoor and outdoor scenes respectively; Step 3: Based on the early warning of different disaster types, calculate the risk index for indoor and outdoor areas respectively; Step 4: Perform compound disaster coupling calculation for related disasters to obtain the comprehensive risk index of compound disasters in corresponding indoor or outdoor scenarios; Step 5: Conduct preparatory work based on the comprehensive risk index assessment; Step 6: After the disaster occurs, the target area is inspected again and the inspection results are compared with the comprehensive risk index obtained in step 4 to achieve data closure and model correction, thereby re-assessing the risk accurately.

2. The post-disaster regional operation risk assessment method based on BeiDou technology according to claim 1 is characterized in that: In step 1, the data collection scope is determined based on all power facilities in the target area, and information about the indoor space of the power facilities and the surrounding environment within a specified range centered on the power facilities is collected. The collection frequency is daily acquisition of parameters, and frequent acquisition of parameters is performed after receiving a disaster forecast.

3. The post-disaster regional operation risk assessment method based on BeiDou technology according to claim 1 or 2, characterized in that: Based on the positioning of power facilities, basic environmental data, meteorological data and equipment status data are obtained; the outdoor data in the basic environmental data is obtained through Beidou static positioning coordinates, and the topography, geological parameters, and external building attributes are obtained. The indoor data in the basic environmental data is obtained by indoor Bluetooth devices; meteorological data is real-time information such as rainfall, wind speed, temperature, magnitude, ice thickness, and historical disaster data obtained through Beidou short messages and meteorological satellites; equipment status data is the operating parameters and security status of all power equipment in the area.

4. The post-disaster risk area operation positioning and guidance device incorporating Beidou technology according to claim 3 is characterized by: During the data acquisition process, non-quantitative data is converted into numerical indicators; all parameters are normalized to eliminate dimensional differences, and new data is updated in real time; all data is associated with device ID and geographic coordinates to ensure data traceability.

5. The post-disaster regional operation risk assessment method based on BeiDou technology according to claim 1 is characterized in that: In step 2, real-time data is used to distinguish the indoor and outdoor attributes of power facilities in the target area, and a weight coefficient α is introduced to dynamically integrate outdoor Beidou data and indoor Bluetooth sensor data to formulate indoor and outdoor switching thresholds.

6. The post-disaster regional operation risk assessment method based on BeiDou technology according to claim 1 is characterized in that: In steps 3 and 4, differentiated indoor and outdoor calculation logics are designed for the four core disasters of floods, typhoons, earthquakes, and ice disasters. Through historical data and real-time warnings, disaster combinations with synergistic effects are identified and coupled calculations are performed.

7. The post-disaster regional operation risk assessment method based on BeiDou technology according to claim 1 is characterized in that: In step 5, preparatory work is performed based on the comprehensive risk index, and different risk levels are divided according to the index and marked on the map, and different entry strategies are configured for different risk levels.

8. The post-disaster regional operation risk assessment method based on BeiDou technology according to claim 1 is characterized in that: In step 6, the pre-disaster comprehensive risk index is compared with the actual damage after the disaster, and the disaster model and coupling coefficient are revised; the entry strategy during maintenance is adjusted according to the revised disaster model and coupling coefficient.

9. The post-disaster regional operation risk assessment method based on BeiDou technology according to claim 1 or 8, characterized in that: It also includes step 7: based on the comparative analysis of actual data obtained after the disaster occurs and the predicted results, the model parameters are corrected to optimize the accuracy of risk assessment.

10. The post-disaster regional operation risk assessment method based on BeiDou technology according to claim 1 or 8, characterized in that: The revised model is verified and iterated, and by establishing a data benchmark library for disaster type fusion scenarios, a graded correction rule table divided by the degree of deviation is formulated, and a standard template suitable for different disasters and indoor and outdoor scenarios is designed.

Images