Intelligent digital platform based on cold region highway disaster emergency disposal

Abstract

The invention provides an intelligent digital platform based on cold region highway disaster emergency disposal, and relates to the technical field of cold region highway emergency management. Comprising a cold region highway disaster database and knowledge graph module, an intelligent emergency disposal scheme recommendation module, a dynamic resource allocation module, a data updating and maintenance module, a platform iterative optimization module and a man-machine interaction module. According to the method, by combining the geographical environment and the knowledge graph of the cold region, the recommendation scheme considers the particularity of disasters such as ice and snow, frost heaving and the like, such as snow-melting agent type selection at low temperature, avoiding blind application of a general scheme, improving the disposal accuracy by more than 40%, and by priority ranking and path optimization, the resource arrival time is shortened by 30-50%, resource idling and waste are avoided, and the method is suitable for popularization and application. The emergency cost is reduced by 20%, the efficiency of resource allocation of the platform is improved, and based on user feedback and new technology iteration, the platform can adapt to new disaster types in cold regions, such as road surface frost cracks caused by extreme cold waves, and the practicability is kept for a long time.

Description

Technical Field

[0001] This invention relates to the field of emergency management technology for highways in cold regions, and in particular to an intelligent digital platform for emergency response to highway disasters in cold regions. Background Technology

[0002] In cold regions, highways are often threatened by disasters such as low temperatures, rain, snow, and ice. These disasters seriously affect the safety of highway traffic and operational efficiency. Due to the special environment of low temperatures, snowfall, and frozen soil, highways in cold regions are prone to disasters such as snow and ice cover, road surface frost heave, and slope avalanches, which seriously affect traffic and driving safety.

[0003] Currently, the following problems exist in emergency response to highway disasters in cold regions:

[0004] The plan lacks specificity: it relies on human experience to formulate the plan without taking into account the geographical characteristics of cold regions (such as avoiding excessive use of de-icing agents in permafrost areas), which may easily lead to secondary disasters.

[0005] Inefficient resource allocation: The distribution of resources does not match the demand accurately, often resulting in situations where "snowplows are urgently needed on road sections but there are no equipment, while idle equipment is too far away";

[0006] Poor data timeliness: Disaster data is updated late, and historical cases are out of touch with the current environment, resulting in low reference value of the solutions;

[0007] The platform iterates slowly: it fails to optimize its functions by incorporating user feedback and new technologies, making it difficult to adapt to complex and ever-changing disaster scenarios in cold regions.

[0008] To address these issues, we propose an intelligent digital platform for emergency response to highway disasters in cold regions. Summary of the Invention

[0009] The problem this invention aims to solve is that in the emergency response to highway disasters in cold regions, there are issues such as inefficient resource allocation, poor data timeliness, slow platform iteration, and insufficient targeted solutions.

[0010] To address the aforementioned technical problems, this invention provides an intelligent digital platform for emergency response to highway disasters in cold regions. The intelligent digital platform specifically includes:

[0011] Cold Region Highway Disaster Database and Knowledge Graph Module: Used to store cold region highway disaster data and emergency response knowledge, including basic disaster data unit, historical response case unit, cold region geographical environment unit and knowledge graph construction unit;

[0012] Intelligent emergency response plan recommendation module: Based on database and knowledge graph, it generates emergency response measures and plans suitable for the current disaster through disaster feature extraction unit, plan matching unit and plan optimization unit;

[0013] Dynamic resource allocation module: Based on the severity of the disaster and the needs of response, it includes a resource information collection unit, a resource priority ranking unit, and an allocation route planning unit to achieve the rational allocation of human, material, and financial resources;

[0014] The data update and maintenance module includes a real-time data acquisition unit, a data verification unit, and a historical data archiving unit. It regularly updates the database to ensure the timeliness and accuracy of the data.

[0015] Platform Iteration and Optimization Module: This includes a user feedback collection unit, a functional performance analysis unit, and a technology upgrade unit, which dynamically optimizes platform functions based on user needs and technological advancements.

[0016] Human-computer interaction module: used to realize information interaction between users and the platform, including data input unit, result display unit and operation feedback unit.

[0017] Preferably, in the cold region highway disaster database and knowledge graph module:

[0018] The disaster basic data unit stores the disaster types of highways in cold regions, including ice and snow, frost heave, avalanches, etc., the time of occurrence, location (e.g., latitude and longitude + road section number), impact range, and severity level.

[0019] The historical case study section stores emergency response plans, response procedures, resource inputs, and response effectiveness evaluation data for highway disasters in cold regions over the past 10 years.

[0020] The cold region geographical environment unit stores environmental parameters such as road section altitude, temperature, snow thickness, and soil freezing depth, as well as basic information such as highway grade, pavement structure, and distribution of surrounding rescue points;

[0021] The knowledge graph construction unit constructs a knowledge graph linking "disaster-solution-resource-environment" by recognizing entities, including disaster types, response measures, and resource types, and extracting relationships, including "snow and ice disaster - need to spread de-icing agent" and "frost heave - need to repair road surface".

[0022] Preferably, in the intelligent emergency response solution recommendation module:

[0023] The disaster feature extraction unit extracts features such as disaster type, severity level, and affected road sections by using sensor data, such as road surface temperature and snow thickness, and user-uploaded information, such as photos of disaster sites.

[0024] The solution matching unit is based on knowledge graphs and uses similarity algorithms, such as cosine similarity, to match the optimal initial solution in historical cases.

[0025] The scheme optimization unit combines real-time environmental data, such as the snowfall forecast for the next 24 hours, to adjust the initial scheme, such as increasing the number of snow removal equipment, and outputs three schemes with priority ranking.

[0026] Preferably, in the dynamic resource allocation module:

[0027] The resource information collection unit acquires real-time data on available resources, including snowplows, de-icing agent reserves, number of rescue personnel, emergency funds, and current resource location.

[0028] The resource priority ranking unit prioritizes resource needs based on the severity level of the disaster, with level 5 being the highest. For example, large snow removal equipment is prioritized for level 5 disasters.

[0029] The allocation route planning unit is based on a GIS map and uses Dijkstra's algorithm to plan the shortest route for resource transportation, avoid interrupted road sections, and generate an allocation schedule.

[0030] Preferably, in the data update and maintenance module:

[0031] The real-time data acquisition unit updates data once an hour through IoT sensors, road condition sensors every 5 kilometers, weather stations, and user reports.

[0032] The data verification unit employs cross-validation, such as comparing sensor data with on-site photos, eliminating outliers, and ensuring data accuracy ≥ 95%.

[0033] The historical data archiving unit stores historical disaster data on an annual basis, forming a structured dataset for model training.

[0034] Preferably, in the platform iterative optimization module:

[0035] The user feedback collection unit gathers feedback from decision-makers on the proposal through satisfaction ratings and written evaluations.

[0036] The functional performance analysis unit statistically analyzes indicators such as scheme adoption rate and resource allocation time to identify platform shortcomings, such as low accuracy of scheme recommendations for a certain type of disaster.

[0037] The technology upgrade unit introduces the latest algorithms, such as reinforcement learning to optimize resource allocation, and technologies, such as drones for disaster inspection, and updates platform functions once a quarter.

[0038] Preferably, in the human-computer interaction module:

[0039] Data input unit: Supports multiple input methods, including automatic sensor upload, manual input via mobile APP, and batch import from Excel;

[0040] Results Display Unit: The solution is displayed in a visual interface, such as in the form of a flowchart, resource allocation route, such as GIS map annotation, and treatment effect prediction, such as in the form of a timeline.

[0041] Operation feedback unit: Provides real-time feedback on the status of user operations, such as "submitting disaster information", such as "received, generating a plan". In case of abnormalities, such as network interruption, it displays retry instructions.

[0042] The technical effects and advantages of this invention are as follows:

[0043] 1. This invention combines the geographical environment of cold regions with knowledge graphs to recommend solutions that take into account the special characteristics of disasters such as snow and ice and frost heave. For example, it selects de-icing agents at low temperatures, avoids blindly applying general solutions, and improves the accuracy of disposal by more than 40%. Through priority ranking and route optimization, the resource arrival time is shortened by 30% to 50%, avoiding resource idleness and waste, reducing emergency costs by 20%, thereby improving the efficiency of resource allocation on the platform.

[0044] 2. This invention ensures that the solution is generated based on the latest information by updating data every hour with an accuracy rate of over 95%, thus solving the problem of historical data lag. Based on user feedback and new technology iteration, the platform can adapt to new types of disasters in cold regions, such as road surface freezing and cracking caused by extreme cold waves, and maintain its practicality in the long term. Attached Figure Description

[0045] Figure 1 This is a schematic diagram of the overall structure of the present invention. Detailed Implementation

[0046] This invention provides an intelligent digital platform for emergency response to highway disasters in cold regions, such as... Figure 1 As shown, it includes:

[0047] Cold Region Highway Disaster Database and Knowledge Graph Module;

[0048] As the platform's "data hub," it stores and links data and knowledge related to highway disasters in cold regions.

[0049] Disaster Basic Data Unit: Collects and stores core disaster information, including:

[0050] Disaster type (e.g., "snowstorms causing road icing" or "frost heave causing road cracking");

[0051] Severity level (Level 1: Minor impact, Level 5: Traffic disruption + people stranded);

[0052] Spatiotemporal information (occurrence time accurate to the minute, location associated with highway station number and latitude and longitude);

[0053] Historical Case Studies Unit: Structured storage of cold-region highway disaster cases from 2014 to 2024. Each case includes:

[0054] Disposal plans (such as "spreading de-icing agent + mechanical snow removal");

[0055] Resource input (3 snowplows, 5 tons of de-icing agent);

[0056] Effectiveness assessment (traffic resumed within 2 hours, no secondary icing);

[0057] Cold Region Geographical Environment Unit: Integrating Static and Dynamic Environmental Data:

[0058] Static data: Highway grade (expressway / national highway), road surface material (asphalt / cement), location of nearby rescue stations;

[0059] Dynamic data: real-time air temperature (-15℃), snow depth (10cm), and soil freezing depth (80cm).

[0060] Knowledge graph construction unit: Entity recognition is performed using the BERT model, and relationships are built using graph neural networks (GNNs), for example:

[0061] "Blizzard Disaster" → "Response Measures": Spread de-icing agent (temperature ≥ -5℃), mechanical snow removal (temperature < -5℃, de-icing agent ineffective);

[0062] “Frost heave disaster” → “Prohibited measures”: Avoid heavy machinery driving over the road (which will exacerbate road surface cracking).

[0063] Intelligent emergency response solution recommendation module;

[0064] Based on data and knowledge, generate precise handling plans:

[0065] Disaster Feature Extraction Unit:

[0066] Automatic extraction: Real-time acquisition of the physical characteristics of disasters through sensors (temperature, humidity, snow depth sensors) deployed along the highway;

[0067] Manual supplement: Supports users to upload on-site photos (such as road collapse conditions), and extract detailed features (collapse area 2m×3m) through image recognition model (YOLOv8);

[0068] Output: Disaster type (snow and ice cover), severity level (level 3), affected road section (G101 line K50-K55 section).

[0069] Solution matching unit:

[0070] Retrieve similar historical cases based on knowledge graphs (such as "same road section, same level of blizzard disaster");

[0071] The cosine similarity algorithm was used to calculate the feature matching degree, and three historical schemes with a matching degree of ≥85% were selected as initial candidates.

[0072] Solution optimization unit:

[0073] Adjust the plan based on real-time environmental data (such as continuous snowfall in the next 6 hours), for example, optimize "2 snowplows" to "4 snowplows + 1 spreader".

[0074] Note the special precautions for cold regions (e.g., "snow melting agent dosage ≤ 20g / m²"). 2 (To prevent seepage into the permafrost and subsequent frost heave the following year);

[0075] Output: Three solutions ranked by “processing efficiency – cost – safety” and the reasons for recommendation.

[0076] Dynamic resource allocation module;

[0077] Achieving precise matching between resources and needs:

[0078] Resource information collection unit:

[0079] Connect to the highway management department's resource management system and synchronize data in real time: snowplows (10 available, 3 of which are at K30), de-icing agent reserves (50 tons, located at K60 warehouse), and rescue personnel (20 on standby).

[0080] Mark the resource status (e.g., whether the snowplow is equipped with snow chains or whether the de-icing agent is for cold regions).

[0081] Resource priority sorting unit:

[0082] Resource needs are prioritized based on the severity of the disaster: Level 5 disasters prioritize the allocation of "large snowplows + emergency medical teams", while Level 3 disasters prioritize the allocation of "small snowplows + general rescue personnel".

[0083] Resources of the same type are sorted in a secondary order based on their distance from the disaster site (equipment within 30 kilometers is given priority).

[0084] Deployment of route planning units:

[0085] Based on GIS maps, avoid closed sections of the road (such as the K45-K48 section which is interrupted by an avalanche);

[0086] Dijkstra's algorithm is used to calculate the shortest transportation route and generate a dispatch schedule (e.g., "snowplows depart from K30 and arrive at the disaster site in 40 minutes").

[0087] Data update and maintenance module;

[0088] Ensuring the timeliness and accuracy of data:

[0089] Real-time data acquisition unit:

[0090] Automatic data collection: The sensor uploads data once per hour, and the weather station updates the forecast every 30 minutes;

[0091] Manual reporting: Supports highway inspectors to upload on-site data (such as "Ice formation at K52") via mobile APP;

[0092] The data format is uniformly JSON to ensure cross-system compatibility.

[0093] Data verification unit:

[0094] Outlier detection: Data that deviates from the normal range is identified using the Z-score algorithm (e.g., "snow thickness 100cm" far exceeds the historical maximum value and is marked as an anomaly).

[0095] Cross-validation: Compare sensor data with drone aerial images (e.g., if the error between the snow thickness sensor data and the aerial estimate is ≤5cm, the validation is passed);

[0096] Error data is automatically pushed to the administrator for review, ensuring that the accuracy of the data entering the database is ≥95%.

[0097] Historical data archiving unit:

[0098] Data is archived by year (e.g., "2023 Snow and Ice Disaster Dataset" and "2024 Frost Heave Case Studies").

[0099] Data security is ensured by using distributed storage (HDFS), and fast retrieval is supported by dimensions such as time and disaster type.

[0100] Platform iteration and optimization module;

[0101] Drive the platform's continuous evolution:

[0102] User feedback collection unit:

[0103] After the plan is implemented, an evaluation questionnaire will be sent to decision-makers, including "Plan Applicability (1-5 stars)", "Timeliness of Resource Allocation (1-5 stars)" and written suggestions (such as "We hope to add a special plan for avalanche disasters").

[0104] Monthly feedback reports are generated, and frequently occurring issues are identified (such as "the recommended dosage of snow melting agent is too high").

[0105] Functional performance analysis unit:

[0106] Key performance indicators (KPIs) monitoring: Solution adoption rate (target ≥ 80%), resource allocation time (target ≤ 30 minutes), user satisfaction (target ≥ 4.2 stars);

[0107] Weakness identification: If the adoption rate of a solution for a certain type of disaster (such as road damage caused by freeze-thaw cycles) is less than 60%, it is marked as an item that needs optimization.

[0108] Technology Upgrade Unit:

[0109] Algorithm upgrade: Introduce reinforcement learning (RL) to optimize resource allocation strategies, and train the model using historical data to improve path planning efficiency;

[0110] Technology integration: Integrating drone inspection functions, it can complete a 5-kilometer road section disaster survey within 10 minutes, replacing manual inspection (which takes 1 hour);

[0111] We release updated versions every quarter, simultaneously upgrading the knowledge graph and recommendation model.

[0112] Human-computer interaction module;

[0113] Enables convenient interaction between users and the platform:

[0114] Data input unit: Supports multiple input methods (automatic sensor upload, manual input via mobile APP, batch import from Excel);

[0115] Results Display Unit: Presents the solution (in flowchart form), resource allocation route (with GIS map annotations), and predicted treatment effects (in timeline form) through a visual interface;

[0116] - Operation Feedback Unit: Provides real-time feedback on the status of user operations (such as "Submit Disaster Information") ("Received, Generating Plan"), and displays retry instructions in case of abnormal situations (such as network interruption).

[0117] Example: Taking a "blizzard disaster on a highway section in a cold region, severity level 3" as an example, the platform operation process is explained as follows:

[0118] Data collection: Sensors upload "Road section K100-K105, snow thickness 8cm, temperature -8℃", users upload on-site photos (road surface icy);

[0119] Recommended solution:

[0120] Feature extraction: Identified as "snow and ice cover disaster, level 3, affecting 5 kilometers of highway section";

[0121] Matching historical cases: Retrieve similar blizzard cases on the same road section in 2022 (solution: 2 snowplows + spreading de-icing agent);

[0122] Optimized Plan: Based on the forecast of "continuous snowfall in the next 4 hours", the plan is optimized to "3 snowplows (including 1 snow sweeper) + 2 tons of cold-region-specific de-icing agent (dosage 15g / m³)". 2 The label reads, "Avoid excessive application on bridge sections (to prevent steel corrosion)."

[0123] Resource allocation:

[0124] Resources collected: Of the 10 snowplows, 3 are located on the K80 section (20 kilometers from the disaster site), and 30 tons of de-icing agent are stored (K90 warehouse);

[0125] Priority order: Prioritize the allocation of 3 snowplows on the K80 section and 2 tons of de-icing agent from the K90 warehouse;

[0126] Route planning: Generate route (K80→K100, avoiding the icy section from K95 to K98), estimated arrival time 30 minutes;

[0127] Feedback: After the solution was completed, the user reported that "the solution was practical and the resources were timely". The platform archived the case for subsequent model training.

[0128] This invention integrates data and knowledge on highway disasters in cold regions through modular design, enabling intelligent solution recommendation, dynamic resource allocation, and continuous platform optimization. This significantly improves the efficiency and accuracy of emergency response to highway disasters in cold regions, providing important technical support for ensuring transportation in these areas.

[0129] It is understood that the present invention has been described through some embodiments, and those skilled in the art will recognize that various changes or equivalent substitutions can be made to these features and embodiments without departing from the spirit and scope of the invention. Furthermore, under the teachings of the present invention, these features and embodiments can be modified to adapt to specific situations and materials without departing from the spirit and scope of the invention. Therefore, the present invention is not limited to the specific embodiments disclosed herein, and all embodiments falling within the scope of the claims of this application are within the protection scope of the present invention.

Claims

1. An intelligent digital platform for emergency response to highway disasters in cold regions, characterized in that, The intelligent digital platform specifically includes: Cold Region Highway Disaster Database and Knowledge Graph Module: Used to store cold region highway disaster data and emergency response knowledge, including basic disaster data unit, historical response case unit, cold region geographical environment unit and knowledge graph construction unit; Intelligent emergency response plan recommendation module: Based on database and knowledge graph, it generates emergency response measures and plans suitable for the current disaster through disaster feature extraction unit, plan matching unit and plan optimization unit; Dynamic resource allocation module: Based on the severity of the disaster and the needs of response, it includes a resource information collection unit, a resource priority ranking unit, and an allocation route planning unit to achieve the rational allocation of human, material, and financial resources; The data update and maintenance module includes a real-time data acquisition unit, a data verification unit, and a historical data archiving unit. It regularly updates the database to ensure the timeliness and accuracy of the data. Platform Iteration and Optimization Module: This includes a user feedback collection unit, a functional performance analysis unit, and a technology upgrade unit, which dynamically optimizes platform functions based on user needs and technological advancements. Human-computer interaction module: used to realize information interaction between users and the platform, including data input unit, result display unit and operation feedback unit.

2. The intelligent digital platform for emergency response to highway disasters in cold regions according to claim 1, characterized in that, In the cold region highway disaster database and knowledge graph module: The basic disaster data unit stores the disaster type, occurrence time, location, impact range, and severity level for highways in cold regions; The historical case study section stores emergency response plans, response procedures, resource inputs, and response effectiveness evaluation data for highway disasters in cold regions over the past 10 years. The cold region geographical environment unit stores environmental parameters such as road section altitude, temperature, snow thickness, and soil freezing depth, as well as basic information such as highway grade, pavement structure, and distribution of surrounding rescue points; The knowledge graph construction unit constructs a knowledge graph linking "disaster-solution-resource-environment" through entity recognition and relation extraction.

3. The intelligent digital platform for emergency response to highway disasters in cold regions according to claim 1, characterized in that, In the intelligent emergency response solution recommendation module: The disaster feature extraction unit extracts features such as disaster type, severity level, and affected road sections through sensor data and user-uploaded information; The solution matching unit is based on a knowledge graph and uses a similarity algorithm to match the optimal initial solution from historical cases; The scheme optimization unit combines real-time environmental data to adjust the initial scheme and output three priority-ranked schemes.

4. The intelligent digital platform for emergency response to highway disasters in cold regions according to claim 1, characterized in that, In the dynamic resource allocation module: The resource information collection unit acquires real-time data on available resources, including snowplows, de-icing agent reserves, number of rescue personnel, emergency funds, and current resource location. The resource priority ranking unit ranks resource needs according to the severity of the disaster; The allocation route planning unit is based on a GIS map and uses Dijkstra's algorithm to plan the shortest route for resource transportation, avoid interrupted road sections, and generate an allocation schedule.

5. The intelligent digital platform for emergency response to highway disasters in cold regions according to claim 1, characterized in that, In the data update and maintenance module: The real-time data acquisition unit updates data once per hour through IoT sensors, weather stations, and user reports. The data verification unit uses cross-validation to remove outliers, ensuring data accuracy ≥ 95%. The historical data archiving unit stores historical disaster data on an annual basis, forming a structured dataset for model training.

6. The intelligent digital platform for emergency response to highway disasters in cold regions according to claim 1, characterized in that, In the platform iteration and optimization module: The user feedback collection unit gathers feedback from decision-makers on the proposal through satisfaction ratings and written evaluations. The functional performance analysis unit statistically analyzes indicators such as solution adoption rate and resource allocation time to identify platform shortcomings; The technology upgrade unit introduces the latest algorithms and technologies, and updates the platform's functions once a quarter.

7. The intelligent digital platform for emergency response to highway disasters in cold regions according to claim 1, characterized in that, In the human-computer interaction module: Data input unit: Supports multiple input methods, including automatic sensor upload, manual input via mobile APP, and batch import from Excel; Results Display Unit: Presents the solution, resource allocation route, and predicted treatment effect through a visual interface; Operation feedback unit: Provides real-time feedback on the status of user operations, and displays retry instructions in case of abnormal situations.

Images