Emergency rescue management optimization method and system based on AI algorithm

Abstract

The application relates to the technical field of emergency management, in particular to an emergency rescue management optimization method based on an AI algorithm.The method comprises the following steps: obtaining the quantity and position of various rescue resources and the state and performance of various rescue equipment in an emergency disaster area, performing rescue resource capacity evaluation, and simultaneously predicting the required emergency rescue resource demand of each emergency rescue task; performing emergency priority sequencing analysis on each emergency rescue task based on the required emergency rescue resource demand of each emergency rescue task, to generate the corresponding emergency rescue priority sequence in the current area; and performing emergency rescue deployment management optimization on the corresponding emergency rescue priority sequence in the current area by taking the maximization of rescue effect as a target and combining the corresponding emergency rescue resource capacity in the current area, to generate the corresponding emergency rescue resource deployment management decision suggestion.The application can improve the response speed and rescue effect of emergency rescue.

Description

Technical Field

[0001] This invention relates to the field of emergency management technology, and in particular to an AI-based method for optimizing emergency rescue management. Background Technology

[0002] In recent years, with the development of artificial intelligence (AI), big data, and the Internet of Things (IoT) technologies, emergency rescue management methods based on AI algorithms have gradually become a research hotspot in the emergency response field. AI technology, especially machine learning and deep learning algorithms, has significant advantages in large-scale data analysis, pattern recognition, and decision support. Intelligent decision-making through AI algorithms enables real-time monitoring of emergencies and intelligent scheduling and optimization of emergency resources, greatly improving the response speed and efficiency of emergency rescue. However, most existing methods rely on a single data source, such as meteorological data, geographic information, or event monitoring data, failing to effectively integrate data from multiple channels, resulting in insufficient decision support. Furthermore, traditional AI algorithms often focus on optimizing a single aspect, such as resource scheduling or personnel deployment, lacking comprehensive consideration of the complex relationships between multi-dimensional factors (such as resources, time, and location), thus reducing the response speed of emergency rescue. Summary of the Invention

[0003] Therefore, the present invention needs to provide an emergency rescue management optimization method and system based on AI algorithms to solve at least one of the above-mentioned technical problems.

[0004] To achieve the above objectives, an AI-based method for optimizing emergency rescue management includes the following steps:

[0005] Step S1: Obtain the quantity and location of various rescue resources and the status and performance of various rescue equipment in the emergency disaster area; based on the quantity and location of various rescue resources and the status and performance of various rescue equipment, conduct a rescue resource capability assessment to obtain the corresponding emergency rescue resource capability in the current area;

[0006] Step S2: Obtain the disaster type, disaster distribution scale, and surrounding population distribution scale corresponding to each emergency rescue task through the emergency disaster-stricken area; obtain the corresponding disaster spread speed and the number of affected people based on the disaster type, disaster distribution scale, and surrounding population distribution scale corresponding to each emergency rescue task, and predict the corresponding emergency rescue resource demand for each emergency rescue task;

[0007] Step S3: Based on the emergency rescue resource requirements for each emergency rescue task and in combination with the disaster type, disaster distribution scale, surrounding population distribution scale, disaster spread speed and the number of affected people, conduct an emergency priority ranking analysis for each emergency rescue task to generate the corresponding emergency rescue priority sequence in the current region.

[0008] Step S4: Optimize the emergency rescue allocation and management of the corresponding emergency rescue priority sequence in the current area by aiming to maximize the rescue effect and combining the corresponding emergency rescue resource capabilities in the current area, so as to generate emergency rescue resource allocation and management decision suggestions for emergency rescue tasks.

[0009] Furthermore, step S1 includes the following steps:

[0010] Step S11: Obtain the quantity and location of various rescue resources in the emergency disaster area;

[0011] Step S12: Obtain the status and performance of various rescue equipment in the emergency disaster area;

[0012] Step S13: Conduct equipment integrity assessment and analysis based on the status and performance of various types of rescue equipment to obtain the integrity probability of each type of rescue equipment;

[0013] Step S14: Based on the availability probability of various types of rescue equipment, perform resource reserve statistics on the quantity and location of various types of rescue resources to obtain the corresponding rescue resource reserve in the region;

[0014] Step S15: Assess the emergency rescue resource capacity of the disaster-stricken area based on the corresponding rescue resource reserves in the region, and obtain the corresponding emergency rescue resource capacity in the current region.

[0015] Furthermore, step S13 includes the following steps:

[0016] Step S131: Obtain the corresponding changes in the size of equipment components and the surface roughness of the equipment by the corresponding status of various rescue equipment;

[0017] Step S132: Based on the changes in the size of the equipment components and the surface roughness of the equipment corresponding to various types of rescue equipment, perform equipment wear assessment calculations to obtain the wear degree corresponding to various types of rescue equipment;

[0018] Step S133: Assign corresponding weights to each performance index within the performance of various types of rescue equipment through hierarchical analysis. Each performance index includes equipment response index, equipment operation index, and equipment capability index. Then, calculate the performance score by weighting the weights and corresponding performance indexes to obtain the performance score of each type of rescue equipment.

[0019] Step S134: Calculate the ratio between the wear and tear and performance scores of various types of rescue equipment to obtain the probability of good condition for each type of rescue equipment.

[0020] Furthermore, step S14 includes the following steps:

[0021] Step S141: Based on the integrity probability of various types of rescue equipment, perform integrity correction analysis on the quantity of various types of rescue resources. If the corresponding integrity probability is less than 50%, reduce the quantity of the corresponding type of rescue resource by one; otherwise, keep it unchanged. This process is used to correct and update the complete quantity of various types of rescue resources.

[0022] Step S142: Assess the accessibility of various types of rescue resources based on their locations to determine the accessibility of these resources for location storage.

[0023] Step S143: Verify and determine the complete quantity of various types of rescue resources based on the location and convenience of their storage, so as to obtain the corresponding rescue resource reserve quantity in the region.

[0024] Furthermore, step S142 includes the following steps:

[0025] By accurately determining the corresponding reserve locations of various rescue resources within the emergency disaster area, the regional reserve locations corresponding to various rescue resources can be obtained.

[0026] Based on the regional reserve locations corresponding to various types of rescue resources, the rescue transportation distance between the corresponding disaster relief mission locations within the emergency disaster area is calculated to obtain the rescue transportation distance between various types of rescue reserve resources and disaster relief mission locations.

[0027] The number of rescue transportation routes and traffic volume at transportation hubs corresponding to various types of rescue resources are determined based on the rescue transportation distance between various types of rescue reserve resources and the disaster relief mission locations.

[0028] Based on the number of rescue transportation routes corresponding to various types of rescue resources and the traffic volume of transportation hubs, the resource convenience of the locations corresponding to various types of rescue resources is assessed to obtain the location reserve convenience of various types of rescue resources.

[0029] Furthermore, step S2 includes the following steps:

[0030] Step S21: Obtain the disaster type, disaster distribution scale, and surrounding population distribution scale corresponding to each emergency rescue mission through the emergency disaster-stricken area;

[0031] Step S22: Set the corresponding disaster observation window size according to the corresponding disaster type, and perform a time series analysis of the disaster distribution scale and the surrounding population distribution scale corresponding to each emergency rescue task based on the disaster observation window size, so as to obtain the disaster distribution area change sequence and population distribution change sequence corresponding to each emergency rescue task under the corresponding window.

[0032] Step S23: Estimate the disaster spread rate based on the disaster distribution area change sequence corresponding to each emergency rescue task under the corresponding window, so as to obtain the disaster spread rate corresponding to each emergency rescue task;

[0033] Step S24: Estimate the affected population based on the population distribution change sequence corresponding to each emergency rescue task under the corresponding window, so as to obtain the number of affected people corresponding to each emergency rescue task;

[0034] Step S25: Based on the disaster spread rate and the number of affected people corresponding to each emergency rescue task, and combined with AI algorithms, predict the emergency rescue resource demand for the corresponding emergency rescue tasks in the disaster-stricken area, so as to predict the amount of emergency rescue resources required for each emergency rescue task.

[0035] Furthermore, the emergency rescue resource demand mentioned in step S25 includes the required quantities of rescue personnel, rescue equipment, and rescue supplies.

[0036] Furthermore, step S3 includes the following steps:

[0037] Step S31: By combining the emergency rescue resource requirements for each emergency rescue task with the corresponding disaster type, disaster distribution scale, surrounding population distribution scale, disaster spread speed and the number of affected people into the corresponding emergency rescue data vector, and forming the corresponding vector matrix, an emergency rescue task vector matrix is ​​generated.

[0038] Step S32: Calculate the eigenvalues ​​of each sub-vector corresponding to the emergency rescue task vector matrix to obtain the vector eigenvalues ​​corresponding to each emergency rescue task sub-vector;

[0039] Step S33: Based on the vector feature values ​​corresponding to each emergency rescue task sub-vector, sort the emergency rescue tasks in the disaster area from largest to smallest to generate the corresponding emergency rescue priority sequence in the current area.

[0040] Furthermore, step S4 includes the following steps:

[0041] Step S41: Obtain the corresponding emergency rescue resource transportation and availability constraints based on the emergency rescue resource capabilities within the current area;

[0042] Step S42: Optimize the emergency rescue allocation and management of each rescue task within the current emergency rescue priority sequence by taking the goal of maximizing rescue effectiveness and combining the constraints of emergency rescue resource transportation and availability within the current region. Based on the constraints, the emergency rescue resources corresponding to each rescue task are rationally arranged according to the transportation distance, transportation time, and resource availability factors, and the optimal corresponding emergency rescue resource transportation route is selected to generate emergency rescue resource allocation and management decision suggestions for the emergency rescue task.

[0043] Furthermore, the present invention also provides an AI-based emergency rescue management optimization system for executing the AI-based emergency rescue management optimization method described above. The AI-based emergency rescue management optimization system includes:

[0044] The rescue capability assessment module is used to obtain the quantity and location of various rescue resources and the status and performance of various rescue equipment in the emergency disaster area; based on the quantity and location of various rescue resources, the module assesses the rescue resource capability of various rescue equipment to obtain the corresponding emergency rescue resource capability in the current area.

[0045] The resource demand prediction module is used to obtain the disaster type, disaster distribution scale, and surrounding population distribution scale corresponding to each emergency rescue task through the emergency disaster-stricken area; based on the disaster type, disaster distribution scale, and surrounding population distribution scale corresponding to each emergency rescue task, it obtains the corresponding disaster spread speed and the number of affected people, and predicts the emergency rescue resource demand required for each emergency rescue task.

[0046] The emergency rescue ranking module is used to perform emergency priority ranking analysis on each emergency rescue task based on the corresponding emergency rescue resource requirements and in combination with disaster type, disaster distribution scale, surrounding population distribution scale, disaster spread speed and the number of affected people, so as to generate the corresponding emergency rescue priority sequence in the current area.

[0047] The emergency rescue management module is used to optimize the allocation and management of emergency rescue priorities within the current region by aiming to maximize rescue effectiveness and combining the corresponding emergency rescue resource capabilities within the current region, so as to generate emergency rescue resource allocation and management decision suggestions for emergency rescue tasks.

[0048] The beneficial effects of this invention are:

[0049] 1. The AI-based emergency rescue management optimization method proposed in this invention, compared with existing technologies, has the following advantages: by acquiring the quantity and location of various rescue resources within the disaster-stricken area, as well as the status and performance information of various rescue equipment, a comprehensive capability assessment of current emergency rescue resources can be conducted, thereby providing data support for subsequent resource allocation and rescue decisions. For example, if the number of medical equipment in a certain area is insufficient or its condition is poor, it can be identified early and supplemented or optimized in a timely manner, avoiding resource waste or missed rescue opportunities. In addition, understanding the performance of various rescue equipment can also ensure that resources can maximize their effectiveness during actual rescue operations, avoiding situations where resources are concentrated in too many or too few areas, ensuring that emergency response can cover the disaster-stricken area in a timely and comprehensive manner, thereby effectively integrating data from multiple channels. Secondly, by analyzing factors such as disaster type, disaster distribution scale, and surrounding population distribution scale, the resource requirements for each emergency rescue mission can be estimated. This analysis helps to accurately predict the needs of various rescue missions, ensuring that all missions receive sufficient resource support. By predicting variables such as the speed of disaster spread and the number of affected people, preparations for resource allocation can be made in advance. Predicting these factors not only allows for accurate assessment of the complexity of rescue missions and resource requirements but also enables adjustments to resource usage strategies based on the specific characteristics of the disaster area, avoiding excessive or insufficient resource allocation. Simultaneously, by analyzing the surrounding population distribution scale, the potential affected population can be predicted, allowing for targeted emergency evacuation and material distribution, ensuring the rationality and efficiency of resource use, and thus better supporting management decisions related to emergency rescue missions. Then, by prioritizing emergency rescue missions based on factors such as disaster type, disaster distribution scale, and surrounding population distribution scale, it can be ensured that limited rescue resources are prioritized for the most urgent and pressing rescue missions. The key to this step is that it maximizes resource utilization efficiency and avoids poor rescue results due to uneven resource allocation. For example, some disasters require immediate attention, such as chemical spills or fires. These situations usually pose a direct threat to human life and must be addressed with priority. Other disasters, while also serious, pose a greater threat to the environment and property and can be dealt with after other high-priority tasks are completed. By prioritizing these events, we can ensure that rescue operations can respond quickly and accurately to the most urgent needs. This approach allows for a comprehensive consideration of the complex relationships between multiple factors (such as resources, time, and location), greatly improving the timeliness and effectiveness of rescue work, thereby maximizing the response speed and effectiveness of emergency rescue.Finally, by optimizing resource allocation based on the goal of maximizing rescue effectiveness and combining regional resource capabilities and priorities, it can provide clear decision support for emergency management personnel, ensuring that all resources are allocated most effectively in complex emergency rescue processes. Through this optimization process, resource allocation can be adjusted in a timely manner according to the actual situation, so that the resource needs of each rescue task are effectively met, while avoiding resource waste. In large-scale disaster scenarios, optimized allocation can also ensure resource balance between different regions and different tasks, avoid excessive concentration or shortage of resources, and ensure that all tasks receive the necessary support on time.

[0050] 2. The AI-based emergency rescue management optimization system proposed in this invention consists of a rescue capability assessment module, a resource demand prediction module, an emergency rescue ranking module, and an emergency rescue management module. It can implement any AI-based emergency rescue management optimization method described in this invention. It is used to combine the operations between computer programs running on each module to realize the AI-based emergency rescue management optimization method. The internal structure of the system cooperates with each other, which can greatly reduce repetitive work and manpower input, and can quickly and effectively provide a more accurate and efficient AI-based emergency rescue management optimization process, thereby simplifying the operation process of the AI-based emergency rescue management optimization system. Attached Figure Description

[0051] Other features, objects, and advantages of the invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings:

[0052] Figure 1 This is a flowchart illustrating the steps of the AI-based emergency rescue management optimization method of the present invention.

[0053] Figure 2 for Figure 1 A detailed flowchart of step S1;

[0054] Figure 3 for Figure 2 A detailed flowchart of step S13. Detailed Implementation

[0055] The technical method of the present invention will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without inventive effort are within the scope of protection of the present invention.

[0056] Furthermore, the accompanying drawings are merely illustrative of the invention and are not necessarily drawn to scale. The same reference numerals in the drawings denote the same or similar parts, and therefore repeated descriptions of them will be omitted. Some block diagrams shown in the drawings are functional entities and do not necessarily correspond to physically or logically independent entities. These functional entities can be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor methods and / or microcontroller methods.

[0057] It should be understood that although the terms "first," "second," etc., may be used herein to describe various units, these units should not be limited by these terms. These terms are used merely to distinguish one unit from another. For example, without departing from the scope of the exemplary embodiments, a first unit may be referred to as a second unit, and similarly, a second unit may be referred to as a first unit. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0058] To achieve the above objectives, please refer to Figures 1 to 3 This invention provides an AI-based method for optimizing emergency rescue management, the method comprising the following steps:

[0059] Step S1: Obtain the quantity and location of various rescue resources and the status and performance of various rescue equipment in the emergency disaster area; based on the quantity and location of various rescue resources and the status and performance of various rescue equipment, conduct a rescue resource capability assessment to obtain the corresponding emergency rescue resource capability in the current area;

[0060] Step S2: Obtain the disaster type, disaster distribution scale, and surrounding population distribution scale corresponding to each emergency rescue task through the emergency disaster-stricken area; obtain the corresponding disaster spread speed and the number of affected people based on the disaster type, disaster distribution scale, and surrounding population distribution scale corresponding to each emergency rescue task, and predict the corresponding emergency rescue resource demand for each emergency rescue task;

[0061] Step S3: Based on the emergency rescue resource requirements for each emergency rescue task and in combination with the disaster type, disaster distribution scale, surrounding population distribution scale, disaster spread speed and the number of affected people, conduct an emergency priority ranking analysis for each emergency rescue task to generate the corresponding emergency rescue priority sequence in the current region.

[0062] Step S4: Optimize the emergency rescue allocation and management of the corresponding emergency rescue priority sequence in the current area by aiming to maximize the rescue effect and combining the corresponding emergency rescue resource capabilities in the current area, so as to generate emergency rescue resource allocation and management decision suggestions for emergency rescue tasks.

[0063] In the embodiments of this invention, please refer to Figure 1 The diagram shown illustrates the steps of the AI-based emergency rescue management optimization method of the present invention. In this example, the AI-based emergency rescue management optimization method includes the following steps:

[0064] Step S1: Obtain the quantity and location of various rescue resources and the status and performance of various rescue equipment in the emergency disaster area; based on the quantity and location of various rescue resources and the status and performance of various rescue equipment, conduct a rescue resource capability assessment to obtain the corresponding emergency rescue resource capability in the current area;

[0065] In this embodiment of the invention, information on various rescue resources and equipment within the emergency disaster area is acquired using Global Positioning System (GPS) devices and sensor networks. Regarding rescue resources, GPS tags are equipped on each material box and equipment unit, such as fire hose boxes and fire extinguisher boxes, to transmit location information in real time. Simultaneously, an inventory management system accurately records the quantity of various rescue resources; for example, a fire supply warehouse may contain 500 reels of fire hose and 300 fire extinguishers. For rescue equipment, various sensors are installed on fire trucks, cranes, and other equipment, such as vibration sensors to monitor engine operation and pressure sensors to monitor hydraulic system performance. Based on this data, the Analytic Hierarchy Process (AHP) is used to assess rescue resource capabilities, constructing an assessment model. Resource quantity, location convenience, and equipment integrity rate are used as assessment indicators. Emergency rescue experts are invited to score the importance of each indicator and determine its weight; for example, resource quantity has a weight of 0.4, location convenience has a weight of 0.3, and equipment integrity rate has a weight of 0.3. The quantitative assessment result of the emergency rescue resource capabilities in the current area is calculated, such as a comprehensive score of 80 points (out of 100).

[0066] Step S2: Obtain the disaster type, disaster distribution scale, and surrounding population distribution scale corresponding to each emergency rescue task through the emergency disaster-stricken area; obtain the corresponding disaster spread speed and the number of affected people based on the disaster type, disaster distribution scale, and surrounding population distribution scale corresponding to each emergency rescue task, and predict the corresponding emergency rescue resource demand for each emergency rescue task;

[0067] In this embodiment of the invention, satellite remote sensing technology and a Geographic Information System (GIS) are used to acquire relevant information on various emergency rescue tasks within the disaster-stricken area. Satellite remote sensing images are used to determine the disaster type using image recognition algorithms. For example, fires exhibit specific high-temperature and smoke characteristics, while earthquakes can be identified through ground deformation and other signs. Within the GIS system, high-precision topographic data and remote sensing images are combined to measure the scale of the disaster distribution. For instance, if a forest fire has a burned area of ​​3 square kilometers, population census data and a real-time population monitoring system are used to determine the surrounding population distribution, such as a population of 2000 within a 5-kilometer radius of the fire. This data is then used to assess the speed of disaster spread. The method employs time series analysis to estimate the rate of disaster spread by comparing satellite images from different time periods, such as a fire spreading 0.5 square kilometers per day. The number of affected people is estimated by analyzing factors such as population flow trends and the extent of the affected area. Based on this information, a deep learning-based Long Short-Term Memory (LSTM) network model is used to predict the demand for emergency rescue resources. The model uses disaster type, scale, spread rate, and population as input features to predict, for example, that a fire rescue mission requires 300 firefighters, 20 fire trucks, and 50 tons of fire extinguishing agents. Finally, the predicted emergency rescue resource demand for each emergency rescue mission is obtained.

[0068] Step S3: Based on the emergency rescue resource requirements for each emergency rescue task and in combination with the disaster type, disaster distribution scale, surrounding population distribution scale, disaster spread speed and the number of affected people, conduct an emergency priority ranking analysis for each emergency rescue task to generate the corresponding emergency rescue priority sequence in the current region.

[0069] In this embodiment of the invention, emergency priority ranking analysis of various emergency rescue tasks is performed using Python data analysis libraries (such as NumPy and pandas) and the Analytic Hierarchy Process (AHP). This is achieved by integrating the emergency rescue resource requirements for each task with disaster type, disaster distribution scale, surrounding population distribution scale, disaster spread speed, and the number of affected people into a single dataset to construct a hierarchical model. The target layer prioritizes emergency rescue tasks, while the criteria layer considers resource demand, disaster severity, and the impact on the affected population. Experts in the field of emergency rescue are invited to analyze the criteria layer. The relative importance of the indicators is compared pairwise to construct a judgment matrix. For example, if experts believe that the degree of resource demand is slightly more important than the degree of disaster severity, the corresponding position in the judgment matrix is ​​assigned a value of 3. By calculating the eigenvector of the judgment matrix, the weight of each criterion layer indicator is obtained. Then, the data set of each emergency rescue task is quantitatively scored, and a comprehensive score is calculated based on the weights. The tasks are sorted from high to low according to the comprehensive score to generate an emergency rescue priority sequence in the current area. For example, task A (earthquake rescue) has a comprehensive score of 90 points, task B (fire rescue) has a comprehensive score of 85 points, and the priority sequence is [task A, task B, ...].

[0070] Step S4: Optimize the emergency rescue allocation and management of the corresponding emergency rescue priority sequence in the current area by aiming to maximize the rescue effect and combining the corresponding emergency rescue resource capabilities in the current area, so as to generate emergency rescue resource allocation and management decision suggestions for emergency rescue tasks.

[0071] In this embodiment of the invention, a linear programming algorithm is used to optimize the allocation and management of emergency rescue resources within the current area, taking into account the available emergency rescue resources. For example, for a fire rescue mission in the priority sequence, the required resources are 250 firefighters, 15 fire trucks, and 40 tons of extinguishing agents. There are three fire rescue bases in the current area: Base A has 150 firefighters, 10 fire trucks, and 20 tons of extinguishing agents, and is located 10 kilometers from the fire scene; Base B has 100 firefighters, 8 fire trucks, and 15 tons of extinguishing agents, and is also located 15 kilometers from the fire scene; Base C has 80 firefighters, 15 fire trucks, and 40 tons of extinguishing agents. Five fire trucks and 10 tons of fire extinguishing agent are deployed, located 8 kilometers from the fire scene. The objective function is the rescue effectiveness (e.g., fire extinguishing speed, success rate of rescuing affected people). Constraints include the quantity of resources at each base, transportation distance, and resource availability. A linear programming algorithm is used to solve the problem, resulting in the following allocations: 150 firefighters, 10 fire trucks, and 20 tons of fire extinguishing agent from base A; 100 firefighters, 5 fire trucks, and 15 tons of fire extinguishing agent from base B; and 0 firefighters, 0 fire trucks, and 5 tons of fire extinguishing agent from base C. The optimal transportation route is determined, such as from base A to the fire scene via main road X and secondary road Y. This information is used to generate emergency rescue resource allocation and management decision recommendations for the emergency rescue mission.

[0072] Furthermore, step S1 includes the following steps:

[0073] Step S11: Obtain the quantity and location of various rescue resources in the emergency disaster area;

[0074] Step S12: Obtain the status and performance of various rescue equipment in the emergency disaster area;

[0075] Step S13: Conduct equipment integrity assessment and analysis based on the status and performance of various types of rescue equipment to obtain the integrity probability of each type of rescue equipment;

[0076] Step S14: Based on the availability probability of various types of rescue equipment, perform resource reserve statistics on the quantity and location of various types of rescue resources to obtain the corresponding rescue resource reserve in the region;

[0077] Step S15: Assess the emergency rescue resource capacity of the disaster-stricken area based on the corresponding rescue resource reserves in the region, and obtain the corresponding emergency rescue resource capacity in the current region.

[0078] As an embodiment of the present invention, reference is made to... Figure 2 As shown, Figure 1 A detailed flowchart of step S1 is shown below. In this embodiment, step S1 includes the following steps:

[0079] Step S11: Obtain the quantity and location of various rescue resources in the emergency disaster area;

[0080] In this embodiment of the invention, by deploying positioning base stations in emergency disaster areas and using ultra-wideband (UWB) positioning technology, various rescue resources are accurately located. Each type of rescue resource is equipped with a UWB tag, such as a special UWB tag installed on fire hose reels. At the same time, barcode or QR code technology is used to count the number of rescue resources. In the material warehouse, when workers load fire hose reels onto vehicles, they scan the QR code on the hose using a barcode scanner, and the system automatically records the quantity. When the vehicle carrying the hose enters the disaster area, the positioning base station receives the UWB tag signal and accurately obtains the location coordinates of the hose reels in the disaster area. Assuming that 50 fire hose reels are deployed in this rescue operation, by scanning the barcode to count the quantity and combining it with UWB positioning, the specific distribution location of these 50 hose reels in the disaster area can be clearly known, such as 10 reels stored at a fire rescue point in a certain street.

[0081] Step S12: Obtain the status and performance of various rescue equipment in the emergency disaster area;

[0082] In this embodiment of the invention, various types of sensors are installed at various locations where fire and rescue equipment is stored in the emergency disaster area (e.g., a fire disaster area). For fire trucks, sensors are installed on key components such as the engine, water tank, and fire pump. For example, a gas composition sensor is installed at the engine air intake to monitor the intake air quality; a liquid level sensor is installed at the water tank to provide real-time feedback on the water level; and a speed sensor is installed on the fire pump shaft to monitor the pump's operating speed. The data collected by these sensors is transmitted to a data aggregation center via low-power wide-area network (LPWAN) technology, such as narrowband Internet of Things (NB-IoT). The data aggregation center uses data parsing algorithms to transform the raw sensor data into intuitive equipment status and performance parameters. For example, by parsing the gas composition sensor data, it can be determined whether the engine combustion is complete, thereby comprehensively obtaining the status and performance of various rescue equipment such as fire trucks.

[0083] Step S13: Conduct equipment integrity assessment and analysis based on the status and performance of various types of rescue equipment to obtain the integrity probability of each type of rescue equipment;

[0084] In this embodiment of the invention, based on previously acquired data on the status and performance of rescue equipment, fault tree analysis (FTA) is used to assess the equipment's integrity. Taking a crane as an example, a fault tree is constructed for the crane, with the crane's malfunction as the top event and engine failure, hydraulic system failure, and electrical system failure as intermediate events. Each intermediate event is further refined into specific bottom events, such as spark plug failure and fuel pump failure. Based on data monitored by sensors, it is determined whether a bottom event has occurred. If a bottom event has occurred, the probability of intermediate and top events is gradually derived upwards according to the logical relationship of the fault tree. Assuming that the analysis shows that the probability of spark plug failure is 0.05 and the probability of fuel pump failure is 0.03, the probability of engine failure is 0.07 after fault tree calculation. Therefore, the overall integrity probability of the crane is 1 - 0.07 = 0.93. A similar fault tree analysis method is used for various types of rescue equipment to obtain their corresponding integrity probabilities.

[0085] Step S14: Based on the availability probability of various types of rescue equipment, perform resource reserve statistics on the quantity and location of various types of rescue resources to obtain the corresponding rescue resource reserve in the region;

[0086] In this embodiment of the invention, resource reserves are statistically analyzed based on the previously obtained probabilities of the integrity of various types of rescue equipment, combined with the acquired information on the quantity and location of rescue resources. For example, for medical rescue resources, if there are 10 medical aid points, each aid point has 10 sets of emergency medical equipment, and the probabilities of the integrity of each set of emergency medical equipment are assessed as 0.9, then the actual usable quantity of these emergency medical equipment can be calculated as 10 × 10 × 0.9 = 90 sets. At the same time, based on the location information of each medical aid point, these usable emergency medical equipment are statistically analyzed according to their geographical location. For example, if there are 3 medical aid points in area A of the disaster area, the number of usable emergency medical equipment in this area is 3 × 10 × 0.9 = 27 sets. Similarly, for various types of rescue resources, such as fire rescue resources and living supplies rescue resources, the quantity and location of the equipment are statistically analyzed based on the probabilities of the equipment's integrity, and finally the corresponding rescue resource reserves in the entire area are obtained.

[0087] Step S15: Assess the emergency rescue resource capacity of the disaster-stricken area based on the corresponding rescue resource reserves in the region, and obtain the corresponding emergency rescue resource capacity in the current region.

[0088] In this embodiment of the invention, the Analytic Hierarchy Process (AHP) is used to assess the emergency rescue resource capacity of disaster-stricken areas. First, an assessment index system is determined, including the quantity, distribution rationality, and integrity rate of rescue resources. A judgment matrix is ​​constructed, and experts in the field of emergency rescue are invited to compare the relative importance of these indicators pairwise. For example, if an expert considers the quantity of rescue resources to be slightly more important than the distribution rationality, a value of 3 is assigned to the corresponding position in the judgment matrix. By calculating the eigenvector of the judgment matrix, the weight of each indicator is obtained. Then, the previously obtained data on the reserve of rescue resources is quantified according to each indicator. For example, the quantity of rescue resources is quantified according to different resource types, and the actual reserve is converted into a corresponding quantitative score. The distribution rationality is scored based on the uniformity of resource coverage in the disaster-stricken area. Finally, by combining the weights of each indicator and the quantitative scores, the comprehensive score of the emergency rescue resource capacity in the current area is calculated, thereby assessing the emergency rescue resource capacity.

[0089] Furthermore, step S13 includes the following steps:

[0090] Step S131: Obtain the corresponding changes in the size of equipment components and the surface roughness of the equipment by the corresponding status of various rescue equipment;

[0091] Step S132: Based on the changes in the size of the equipment components and the surface roughness of the equipment corresponding to various types of rescue equipment, perform equipment wear assessment calculations to obtain the wear degree corresponding to various types of rescue equipment;

[0092] Step S133: Assign corresponding weights to each performance index within the performance of various types of rescue equipment through hierarchical analysis. Each performance index includes equipment response index, equipment operation index, and equipment capability index. Then, calculate the performance score by weighting the weights and corresponding performance indexes to obtain the performance score of each type of rescue equipment.

[0093] Step S134: Calculate the ratio between the wear and tear and performance scores of various types of rescue equipment to obtain the probability of good condition for each type of rescue equipment.

[0094] As an embodiment of the present invention, reference is made to... Figure 3 As shown, Figure 2 A detailed flowchart of step S13 is shown in this embodiment. Step S13 includes the following steps:

[0095] Step S131: Obtain the corresponding changes in the size of equipment components and the surface roughness of the equipment by the corresponding status of various rescue equipment;

[0096] In this embodiment of the invention, a high-precision laser measuring instrument is used to acquire the dimensional changes of equipment components for various rescue equipment. Taking the ladder arm of a fire truck as an example, multiple laser measuring instrument measurement points are set at key parts of the ladder arm, such as the connection points of each section and the telescopic joints. The laser measuring instrument emits a laser beam, and the distance between the measurement point and the surface of the equipment component is accurately calculated by measuring the time it takes for the laser to reflect back, thereby acquiring the initial dimensional data of the component. Measurements are performed periodically (e.g., weekly), and the new measured data is compared with the initial data to determine the dimensional changes of the component. Simultaneously, a profilometer is used to measure the surface roughness of the equipment. The probe of the profilometer is slowly moved along the surface of the fire truck body, ladder arm, etc. The sensor of the probe generates an electrical signal based on the microscopic undulations of the surface. The internal circuit of the profilometer converts the electrical signal into a digital signal and calculates the surface roughness parameters according to a standard algorithm, thereby comprehensively acquiring the dimensional changes of equipment components and the surface roughness of various rescue equipment.

[0097] Step S132: Based on the changes in the size of the equipment components and the surface roughness of the equipment corresponding to various types of rescue equipment, perform equipment wear assessment calculations to obtain the wear degree corresponding to various types of rescue equipment;

[0098] In this embodiment of the invention, based on the obtained data on changes in the dimensions of equipment components and the surface roughness of the equipment, an empirical formula combined with database comparison is used to calculate the wear assessment of the equipment, establishing a rescue equipment wear assessment database. The database stores the correspondence between the range of component size changes, surface roughness values, and wear degree of different types of rescue equipment under different service years and operating conditions. For example, for the impeller of a fire pump, if the size change of its components exceeds 10% of the normal range in the database within a certain period of time, and the surface roughness increases by 0.5 μm, by querying the database and combining it with the pre-set empirical formula: Wear degree = Size change influence coefficient × Size change ratio + Surface roughness influence coefficient × Surface roughness increase value, assuming the size change influence coefficient is 0.6 and the surface roughness influence coefficient is 0.4, the wear degree of the fire pump impeller can be calculated as 0.6 × 0.1 + 0.4 × 0.5 = 0.26. The wear degree corresponding to various types of rescue equipment can be deduced in this way.

[0099] Step S133: Assign corresponding weights to each performance index within the performance of various types of rescue equipment through hierarchical analysis. Each performance index includes equipment response index, equipment operation index, and equipment capability index. Then, calculate the performance score by weighting the weights and corresponding performance indexes to obtain the performance score of each type of rescue equipment.

[0100] In this embodiment of the invention, the Analytic Hierarchy Process (AHP) is used to assign corresponding weights to various performance indicators within the performance of different types of rescue equipment. Taking a crane as an example, a hierarchical model is first constructed. The target layer is the crane performance evaluation, the criterion layer consists of equipment response indicators, equipment operation indicators, and equipment capability indicators, and the indicator layer is further refined into specific indicators such as start-up response time, operational stability, and lifting weight. Experts in the field of emergency rescue are invited to compare the relative importance of each indicator in the criterion layer pairwise to construct a judgment matrix. For example, if an expert considers the equipment capability indicator to be slightly more important than the equipment response indicator, a value of 3 is assigned to the corresponding position in the judgment matrix. By calculating the eigenvector of the judgment matrix, the various criteria are obtained. The weights of the layer indicators are assigned as follows: assuming the weight of the equipment response indicator is 0.2, the weight of the equipment operation indicator is 0.3, and the weight of the equipment capability indicator is 0.5. Then, each layer indicator is quantitatively scored. For example, a startup response time within 1 second scores 10 points, good operational stability scores 8 points, and lifting weight reaching the rated value scores 10 points. The performance score is then calculated by weighting the indicators according to their respective weights: Performance Score = Equipment Response Indicator Weight × Equipment Response Indicator Score + Equipment Operation Indicator Weight × Equipment Operation Indicator Score + Equipment Capability Indicator Weight × Equipment Capability Indicator Score. That is, the crane performance score = 0.2 × 10 + 0.3 × 8 + 0.5 × 10 = 9.4. Thus, the performance scores corresponding to various types of rescue equipment are obtained.

[0101] Step S134: Calculate the ratio between the wear and tear and performance scores of various types of rescue equipment to obtain the probability of good condition for each type of rescue equipment.

[0102] In this embodiment of the invention, the probability of equipment being in good working order is determined by calculating the ratio of wear level and performance score of various types of rescue equipment. Assuming a certain type of fire rescue equipment has a wear level of 0.3, a performance score of 8, and a reference score of 10 (i.e., full marks), the formula for calculating the probability of equipment being in good working order is: Probability of good working order = (Reference performance score - Wear level × Performance score) / Reference performance score. Substituting the data, we get the probability of equipment being in good working order = (10 - 0.3 × 8) / 10 = 0.76. This formula is used for all types of rescue equipment, and the obtained wear level and performance score are used to calculate the probability of good working order for each type of rescue equipment, providing key data support for subsequent statistics on rescue resource reserves and assessment of emergency rescue resource capabilities.

[0103] Furthermore, step S14 includes the following steps:

[0104] Step S141: Based on the integrity probability of various types of rescue equipment, perform integrity correction analysis on the quantity of various types of rescue resources. If the corresponding integrity probability is less than 50%, reduce the quantity of the corresponding type of rescue resource by one; otherwise, keep it unchanged. This process is used to correct and update the complete quantity of various types of rescue resources.

[0105] In this embodiment of the invention, the quantity of each type of rescue resource is corrected based on the availability probability of the corresponding rescue equipment. Taking fire extinguishers as an example, assuming there are 50 fire extinguishers, the availability probability of these devices, such as pressure testing equipment and nozzle patency testing equipment, is found to be 40%, less than 50%, based on the initial status assessment of the fire extinguisher's associated equipment. Therefore, according to the rules, the number of fire extinguishers is reduced by one, resulting in a corrected number of 49 fire extinguishers. If the availability probability of the testing equipment associated with a certain type of fire hose is 60%, greater than 50%, the number of that type of fire hose remains unchanged at the initial count of 30. For each type of rescue resource, such as first aid kits and fire axes, this logic is strictly followed, and the quantity is corrected based on the availability probability of the corresponding rescue equipment, thereby updating and correcting the available quantity of each type of rescue resource.

[0106] Step S142: Assess the accessibility of various types of rescue resources based on their locations to determine the accessibility of these resources for location storage.

[0107] In this embodiment of the invention, Geographic Information System (GIS) technology is used to assess the accessibility of various types of rescue resources based on their locations. Location information of these resources, such as latitude and longitude coordinates, is imported into the GIS system. This information is then analyzed in conjunction with road network data, population distribution data, and building distribution data of the disaster-stricken area. For example, medical emergency points located along main roads in disaster-stricken areas with dense surrounding populations have high accessibility because rescue personnel can quickly reach these points to obtain supplies and rapidly provide services to a large number of affected people. Conversely, rescue supply storage points located in remote mountainous areas with poor transportation and sparse surrounding populations have lower accessibility. The spatial analysis function of the GIS system is used to quantitatively assess the location of each rescue resource, setting accessibility from 0 to 10 points and assigning a score based on factors such as surrounding transportation and population to determine the accessibility of various types of rescue resources.

[0108] Step S143: Verify and determine the complete quantity of various types of rescue resources based on the location and convenience of their storage, so as to obtain the corresponding rescue resource reserve quantity in the region.

[0109] In this embodiment of the invention, the corrected quantity of complete supplies is verified and determined based on the location convenience of various types of rescue resources obtained previously. For example, for a certain type of fire sandbag, the corrected quantity is 80 bags, and the location convenience score is 8 points (high score means high convenience). Considering the convenience of the location, the easy access to the supplies, and the ability to efficiently serve the disaster area, the reserve of these 80 fire sandbags is considered reasonable and is included in the regional rescue resource reserve statistics. However, for another type of fireproof clothing, the location convenience score is only 3 points (low score means low convenience). Even if the corrected quantity is 50 sets, the reserve will be adjusted according to the actual situation because it is inconvenient to access and difficult to effectively serve the disaster area. For example, the quantity may be increased to ensure that rescue needs can be met even in unfavorable locations. Finally, by comprehensively considering the situation of various rescue resources, the corresponding rescue resource reserve in the region is obtained.

[0110] Furthermore, step S142 includes the following steps:

[0111] By accurately determining the corresponding reserve locations of various rescue resources within the emergency disaster area, the regional reserve locations corresponding to various rescue resources can be obtained.

[0112] In this embodiment of the invention, by combining the Global Positioning System (GPS) and the Geographic Information System (GIS), the storage locations of various types of rescue resources in the emergency disaster area are accurately determined. High-precision GPS positioning devices are equipped for each type of rescue resource storage facility, such as fire-fighting material warehouses and medical rescue points, to acquire their latitude and longitude coordinates in real time. This coordinate information is then imported into the GIS system. In the system, by matching it with detailed disaster area map data, the specific locations of various rescue resources can be clearly and intuitively marked, thus obtaining the corresponding regional storage locations for each type of rescue resource. For example, after obtaining the coordinates of a large emergency material storage center through GPS positioning, the GIS system clearly shows that it is located in the southern part of the city center of the disaster area, with surrounding roads, buildings, and other geographical information readily available, facilitating subsequent analysis of its relationship with disaster relief mission locations.

[0113] Preferably, the rescue transportation distance between corresponding disaster relief mission locations within the emergency disaster area is calculated based on the regional reserve locations corresponding to various types of rescue resources, so as to obtain the rescue transportation distance between various types of rescue reserve resources and disaster relief mission locations;

[0114] In this embodiment of the invention, by utilizing the path analysis function in the GIS system, based on the regional reserve locations corresponding to various rescue resources, the rescue transportation distance between corresponding disaster relief task locations within the emergency disaster area is calculated. Taking a fire rescue task as an example, the locations of the fire-fighting material reserve warehouse and the fire location are marked in the GIS system. Based on the built-in road network data, path search algorithms such as Dijkstra's algorithm or A* algorithm are used to calculate the shortest path from the fire-fighting material reserve warehouse to the fire location. The length of this path is the rescue transportation distance. Assuming that the shortest rescue transportation distance from fire-fighting material reserve warehouse A to a fire scene is calculated by the system to be 5 kilometers, the rescue transportation distance between various rescue reserve resources and disaster relief task locations can be obtained in this way, providing basic data for subsequent analysis.

[0115] Preferably, the number of rescue transportation routes and the traffic volume of transportation hubs corresponding to various types of rescue resources are determined based on the rescue transportation distance between various types of rescue reserve resources and the disaster relief mission location;

[0116] In this embodiment of the invention, by leveraging traffic big data analysis technology, the number of rescue transportation routes and traffic volume at transportation hubs corresponding to various types of rescue resources are determined based on the rescue transportation distance between various types of rescue reserve resources and disaster relief mission locations. Historical traffic flow data and road capacity data within the disaster area are collected to construct a traffic model. When the transportation distance of a certain type of rescue resource from its reserve location to the disaster relief mission location is known, such as the transportation distance from medical rescue point B to a temporary resettlement site for disaster victims being 3 kilometers, the traffic model will analyze possible transportation routes from medical rescue point B to the resettlement site based on historical traffic data and current road conditions. Assuming that the analysis reveals 3 feasible routes, and by analyzing the historical traffic flow data of transportation hubs (such as major intersections and bridges) along these routes, the possible traffic volume of each transportation hub during the emergency rescue period is calculated. For example, in similar emergency situations in the past, a certain transportation hub has an average of 200 vehicles passing through per hour. In this way, the number of rescue transportation routes and traffic volume at transportation hubs corresponding to various types of rescue resources are determined.

[0117] Preferably, the location of each type of rescue resource is assessed based on the number of rescue transportation routes corresponding to each type of rescue resource and the traffic volume of transportation hubs, so as to obtain the location reserve convenience of each type of rescue resource.

[0118] In this embodiment of the invention, the Analytic Hierarchy Process (AHP) is used to assess the location convenience of various rescue resources based on the number of rescue transportation routes and the traffic volume of transportation hubs corresponding to different types of rescue resources. This yields the location reserve convenience of each type of rescue resource. A hierarchical model is constructed, with the target layer representing the location convenience assessment of rescue resources and the criterion layers representing the number of rescue transportation routes and the traffic volume of transportation hubs. Experts in the field of emergency rescue are invited to compare the relative importance of each indicator in the criterion layer pairwise, constructing a judgment matrix. Assuming that experts consider the number of rescue transportation routes slightly more important than the traffic volume of transportation hubs, a value of 3 is assigned to the corresponding position in the judgment matrix. The calculation... The eigenvectors of the judgment matrix are used to derive the weights of each criterion layer indicator. For example, the weight of the number of rescue transport routes is 0.6, and the weight of the traffic volume of the transport hub is 0.4. The location of various rescue resources is quantitatively scored based on the number of rescue transport routes and the traffic volume of the transport hub. For example, if a rescue resource location has 5 rescue transport routes and a relatively small traffic volume at the transport hub, the score is 8 points. The score is calculated according to the weights: Location reserve convenience = weight of the number of rescue transport routes × score of the number of rescue transport routes + weight of the traffic volume of the transport hub × score of the traffic volume of the transport hub = 0.6 × 8 + 0.4 × 6 = 7.2 points. The location reserve convenience corresponding to various rescue resources is obtained in this way.

[0119] Furthermore, step S2 includes the following steps:

[0120] Step S21: Obtain the disaster type, disaster distribution scale, and surrounding population distribution scale corresponding to each emergency rescue mission through the emergency disaster-stricken area;

[0121] In this embodiment of the invention, relevant information about the emergency disaster area is obtained by using satellite remote sensing technology, geographic information system (GIS), and population census data. High-resolution satellite remote sensing images are processed using image recognition algorithms to identify different land cover types, thereby determining the disaster type corresponding to each emergency rescue task, such as fire or flood. In the GIS system, high-precision terrain data and satellite imagery are loaded, and combined with the image recognition results, the scale of the disaster distribution is accurately measured using area calculation tools. At the same time, population distribution data around the disaster area is extracted from the latest population census database, and the population distribution scale of the surrounding area is filtered according to the area boundary range. For example, in a fire rescue mission in a city, satellite remote sensing images identify the fire area, the GIS system measures the fire distribution scale to be 2 square kilometers, and the population distribution scale within a 2-kilometer radius is found to be 5,000 people through population census data.

[0122] Step S22: Set the corresponding disaster observation window size according to the corresponding disaster type, and perform a time series analysis of the disaster distribution scale and the surrounding population distribution scale corresponding to each emergency rescue task based on the disaster observation window size, so as to obtain the disaster distribution area change sequence and population distribution change sequence corresponding to each emergency rescue task under the corresponding window.

[0123] In this embodiment of the invention, the size of the disaster observation window is set according to the type of disaster. For fire disasters, which spread rapidly, the observation window is set to 1 hour; for flood disasters, which develop relatively slowly, the observation window is set to 6 hours. In the GIS system, time series analysis tools are used to analyze the disaster distribution scale and surrounding population distribution scale corresponding to each emergency rescue task with the set observation window. For example, for the above-mentioned fire rescue task, the fire distribution scale is measured once per hour, and the measurement is repeated 5 times to obtain the fire distribution area change sequence [2.1 square kilometers, 2.3 square kilometers, 2.5 square kilometers, 2.7 square kilometers, 2.9 square kilometers]. At the same time, the surrounding population distribution scale is statistically analyzed at 1-hour intervals to obtain the population distribution change sequence, such as [4800 people, 4500 people, 4200 people, 3900 people, 3600 people]. This is used to obtain the change sequence corresponding to each emergency rescue task under the corresponding window.

[0124] Step S23: Estimate the disaster spread rate based on the disaster distribution area change sequence corresponding to each emergency rescue task under the corresponding window, so as to obtain the disaster spread rate corresponding to each emergency rescue task;

[0125] In this embodiment of the invention, the disaster spread rate is estimated by using a linear regression algorithm based on the disaster distribution area change sequence corresponding to each emergency rescue task within the corresponding window. Taking fire rescue task as an example, the area value in the fire distribution area change sequence is used as the dependent variable, and time (in units of the observation window) is used as the independent variable. For example, the time sequence corresponding to the fire distribution area change sequence is [1 hour, 2 hours, 3 hours, 4 hours, 5 hours]. The linear relationship between area and time is calculated by the linear regression algorithm. Assuming the regression equation is y = 0.2x + 2 (where y is the fire distribution area and x is the time), the slope of the equation, 0.2, is the disaster spread rate corresponding to the fire within the observation window, in square kilometers per hour. The disaster spread rate corresponding to each emergency rescue task is obtained in this way.

[0126] Step S24: Estimate the affected population based on the population distribution change sequence corresponding to each emergency rescue task under the corresponding window, so as to obtain the number of affected people corresponding to each emergency rescue task;

[0127] In this embodiment of the invention, the affected population is estimated using a moving average method based on the population distribution change sequence corresponding to each emergency rescue task under the corresponding window. Taking the aforementioned fire rescue task as an example, for the population distribution change sequence [4800 people, 4500 people, 4200 people, 3900 people, 3600 people], the moving average window is set to 3. First, the average of the first three data points is calculated, i.e., (4800+4500+4200) / 3 = 4500 people; then, moving one data point forward, the average of (4500+4200+3900) / 3 = 4200 people is calculated, and so on, until the estimated number of affected people is finally obtained. Assuming that after the moving average calculation, the number of affected people corresponding to the fire rescue task is stable at around 4000 people, the number of affected people corresponding to each emergency rescue task is obtained in this way.

[0128] Step S25: Based on the disaster spread rate and the number of affected people corresponding to each emergency rescue task, and combined with AI algorithms, predict the emergency rescue resource demand for the corresponding emergency rescue tasks in the disaster-stricken area, so as to predict the amount of emergency rescue resources required for each emergency rescue task.

[0129] In this embodiment of the invention, based on the disaster spread rate and the number of affected people corresponding to each emergency rescue task, a deep learning-based Long Short-Term Memory (LSTM) network model combined with AI algorithms is used to predict the emergency rescue resource requirements for the corresponding emergency rescue tasks within the disaster-stricken area. The disaster spread rate, the number of affected people, and other relevant features (such as the one-hot encoding of the disaster type) are used as input data to train the LSTM model. For example, for a fire rescue task, the fire spread rate (e.g., 0.2 square kilometers / hour), the number of affected people (4,000 people), and the fire type encoding (e.g., 1 represents a forest fire, 2 represents an urban fire, etc.) are input into the trained LSTM model. After learning and calculation, the model outputs the emergency rescue resource requirements corresponding to each emergency rescue task. For example, it is predicted that the fire rescue task requires 300 firefighters, 20 fire trucks, and 50 tons of fire extinguishing agents, thus predicting the emergency rescue resource requirements corresponding to each emergency rescue task.

[0130] Furthermore, the emergency rescue resource demand mentioned in step S25 includes the required quantities of rescue personnel, rescue equipment, and rescue supplies.

[0131] Furthermore, step S3 includes the following steps:

[0132] Step S31: By combining the emergency rescue resource requirements for each emergency rescue task with the corresponding disaster type, disaster distribution scale, surrounding population distribution scale, disaster spread speed and the number of affected people into the corresponding emergency rescue data vector, and forming the corresponding vector matrix, an emergency rescue task vector matrix is ​​generated.

[0133] In this embodiment of the invention, an emergency rescue task vector matrix is ​​constructed using Python's NumPy library. First, for each emergency rescue task, the required emergency rescue resources, such as the number of rescue personnel, equipment, and supplies, are integrated with the corresponding disaster type (which can be numerically encoded, e.g., 1 represents earthquake, 2 represents fire, etc.), disaster distribution scale (in square kilometers), surrounding population distribution scale (specific number of people), disaster spread rate (area change per unit time, e.g., square kilometers / hour), and the number of affected people. For example, a fire rescue mission requires 200 firefighters, 15 fire trucks, and 30 tons of extinguishing agents. The disaster type code is 2, the disaster distribution area is 1.5 square kilometers, the surrounding population distribution area is 3,000 people, the disaster spread rate is 0.1 square kilometers / hour, and the number of affected people is 1,000. These data are combined into a vector [200,15,30,2,1.5,3000,0.1,1000]. The same operation is performed on all emergency rescue missions, and these vectors are used as rows to construct a two-dimensional vector matrix, thereby generating the emergency rescue mission vector matrix.

[0134] Step S32: Calculate the eigenvalues ​​of each sub-vector corresponding to the emergency rescue task vector matrix to obtain the vector eigenvalues ​​corresponding to each emergency rescue task sub-vector;

[0135] In this embodiment of the invention, the eigenvalues ​​of each sub-vector within the previously generated emergency rescue task vector matrix are calculated using Python's NumPy library. NumPy provides efficient linear algebra calculation capabilities. For each sub-vector (i.e., each row of data) in the matrix, an eigenvalue decomposition algorithm is used. For example, for a row vector [200, 15, 30, 2, 1.5, 3000, 0.1, 1000] in the emergency rescue task vector matrix, relevant NumPy functions, such as `numpy.linalg.eigvals`, are called to process the vector. This function calculates eigenvalues ​​reflecting the vector's characteristics based on the relationships between its elements. The calculated eigenvalues ​​comprehensively reflect the characteristics of the emergency rescue task in terms of resource requirements, disaster conditions, etc. For example, after calculation, the eigenvalue of this sub-vector is 5.67 (the specific value is derived from the calculation). The eigenvalues ​​corresponding to each emergency rescue task sub-vector are obtained in this way.

[0136] Step S33: Based on the vector feature values ​​corresponding to each emergency rescue task sub-vector, sort the emergency rescue tasks in the disaster area from largest to smallest to generate the corresponding emergency rescue priority sequence in the current area.

[0137] In this embodiment of the invention, the emergency rescue tasks within the disaster area are prioritized by using Python's sorting function to process the vector feature values ​​corresponding to each emergency rescue task sub-vector, in descending order. For example, assuming the vector feature values ​​corresponding to five emergency rescue task sub-vectors are [8.2, 5.67, 6.5, 4.1, 7.8], the Python `sorted` function combined with the `reverse=True` parameter is used to sort these feature values, resulting in [8.2, 7.8, 6.5, 5.67, 4.1]. Simultaneously, the emergency rescue task number corresponding to each feature value is recorded. Based on the sorting result, the corresponding emergency rescue tasks are arranged in descending order of feature value, generating the corresponding emergency rescue priority sequence within the current area. For instance, if feature value 8.2 corresponds to task A and 7.8 corresponds to task C, then the emergency rescue priority sequence is [task A, task C, ...]. This clearly identifies which emergency rescue tasks need to be prioritized, providing a basis for the rational allocation of rescue resources.

[0138] Furthermore, step S4 includes the following steps:

[0139] Step S41: Obtain the corresponding emergency rescue resource transportation and availability constraints based on the emergency rescue resource capabilities within the current area;

[0140] In this embodiment of the invention, the transportation and availability constraints of emergency rescue resources within the current area are obtained by leveraging a Geographic Information System (GIS) and a resource management database. The GIS system stores detailed road network data, including road length, traffic capacity, and congestion information. By connecting with the resource management database, the inventory quantity, storage location, and availability status of various emergency rescue resources (such as equipment condition and material sufficiency) can be obtained. For example, if there are three fire-fighting material storage points in an area, the resource management database shows that storage point A has 50 tons of fire extinguishing agent, storage point B has 30 tons, and storage point C has 20 tons. The number of transport vehicles and their availability at each storage point are also recorded. Simultaneously, the GIS system shows that the road length from storage point A to a fire rescue mission location is 10 kilometers, the road capacity is 50 vehicles per hour, and the current congestion index is 0.3. Based on this information, the transportation and availability constraints of emergency rescue resources are determined, including restrictions on the quantity of resources available and road traffic restrictions during transportation.

[0141] Step S42: Optimize the emergency rescue allocation and management of each rescue task within the current emergency rescue priority sequence by taking the goal of maximizing rescue effectiveness and combining the constraints of emergency rescue resource transportation and availability within the current region. Based on the constraints, the emergency rescue resources corresponding to each rescue task are rationally arranged according to the transportation distance, transportation time, and resource availability factors, and the optimal corresponding emergency rescue resource transportation route is selected to generate emergency rescue resource allocation and management decision suggestions for the emergency rescue task.

[0142] In this embodiment of the invention, a linear programming algorithm is used to maximize the rescue effect. Combined with previously obtained constraints on emergency rescue resource transportation and availability, the emergency rescue allocation and management of each rescue task within the current region's emergency rescue priority sequence is optimized. Taking an earthquake rescue task as an example, this task is at the forefront of the priority sequence. The required rescue resources are: 200 rescue personnel, 5 life detectors, and 300 medical first-aid kits. Based on resource transportation and availability constraints, multiple resource supply points are available. Assume that transporting rescue personnel from rescue base A is 15 kilometers away, with an estimated transportation time of 2 hours, and that base has 300 deployable rescue personnel; transporting life detectors from warehouse B is 8 kilometers away, with a transportation time of 1 hour... For example, warehouse B has 8 life detectors; transporting medical first aid kits from medical supply reserve center C is 12 kilometers away and takes 1.5 hours. Center C has 400 medical first aid kits. Using a linear programming algorithm, the rescue effect (such as the number of people successfully rescued, rescue efficiency, etc.) is used as the objective function, and transportation distance, transportation time, resource availability, etc. are used as constraints to solve the problem. Finally, it is determined that 200 rescuers should be dispatched from rescue base A, 5 life detectors should be dispatched from warehouse B, and 300 medical first aid kits should be dispatched from medical supply reserve center C. The optimal transportation route is also determined. For example, the optimal route from rescue base A to the earthquake rescue point is through main road A and secondary road B. Based on this, emergency rescue resource allocation and management decision-making suggestions are generated for the emergency rescue mission.

[0143] Furthermore, the present invention also provides an AI-based emergency rescue management optimization system for executing the AI-based emergency rescue management optimization method described above. The AI-based emergency rescue management optimization system includes:

[0144] The rescue capability assessment module is used to obtain the quantity and location of various rescue resources and the status and performance of various rescue equipment in the emergency disaster area; based on the quantity and location of various rescue resources, the module assesses the rescue resource capability of various rescue equipment to obtain the corresponding emergency rescue resource capability in the current area.

[0145] The resource demand prediction module is used to obtain the disaster type, disaster distribution scale, and surrounding population distribution scale corresponding to each emergency rescue task through the emergency disaster-stricken area; based on the disaster type, disaster distribution scale, and surrounding population distribution scale corresponding to each emergency rescue task, it obtains the corresponding disaster spread speed and the number of affected people, and predicts the emergency rescue resource demand required for each emergency rescue task.

[0146] The emergency rescue ranking module is used to perform emergency priority ranking analysis on each emergency rescue task based on the corresponding emergency rescue resource requirements and in combination with disaster type, disaster distribution scale, surrounding population distribution scale, disaster spread speed and the number of affected people, so as to generate the corresponding emergency rescue priority sequence in the current area.

[0147] The emergency rescue management module is used to optimize the allocation and management of emergency rescue priorities within the current region by aiming to maximize rescue effectiveness and combining the corresponding emergency rescue resource capabilities within the current region, so as to generate emergency rescue resource allocation and management decision suggestions for emergency rescue tasks.

[0148] Therefore, the embodiments should be considered as exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of the equivalents of the application are intended to be included within the invention.

[0149] The above description is merely a specific embodiment of the present invention, enabling those skilled in the art to understand or implement the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the invention. Therefore, the present invention is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features of the invention herein.

Claims

1. An AI-based method for optimizing emergency rescue management, characterized in that, Includes the following steps: Step S1: Obtain the quantity and location of various rescue resources and the status and performance of various rescue equipment in the emergency disaster area; assess the rescue resource capacity based on the quantity and location of various rescue resources and the status and performance of various rescue equipment to obtain the corresponding emergency rescue resource capacity in the current area; Step S1 includes the following steps: Step S11: Obtain the quantity and location of various rescue resources in the emergency disaster area; Step S12: Obtain the status and performance of various rescue equipment in the emergency disaster area; Step S13: Conduct equipment integrity assessment and analysis based on the status and performance of various types of rescue equipment to obtain the integrity probability of each type of rescue equipment; Step S14: Based on the operational probability of various types of rescue equipment, statistical analysis is performed on the quantity and location of various rescue resources to obtain the corresponding rescue resource reserves in the region; Step S14 includes the following steps: Step S141: Based on the integrity probability of various types of rescue equipment, perform integrity correction analysis on the quantity of various types of rescue resources. If the corresponding integrity probability is less than 50%, reduce the quantity of the corresponding type of rescue resource by one; otherwise, keep it unchanged. This process is used to correct and update the complete quantity of various types of rescue resources. Step S142: Assess the accessibility of various types of rescue resources based on their locations to determine the accessibility of these resources for future use. Step S142 includes the following steps: By accurately determining the corresponding reserve locations of various rescue resources within the emergency disaster area, the regional reserve locations corresponding to various rescue resources can be obtained. Based on the regional reserve locations corresponding to various types of rescue resources, the rescue transportation distance between the corresponding disaster relief mission locations within the emergency disaster area is calculated to obtain the rescue transportation distance between various types of rescue reserve resources and disaster relief mission locations. The number of rescue transportation routes and traffic volume at transportation hubs corresponding to various types of rescue resources are determined based on the rescue transportation distance between various types of rescue reserve resources and the disaster relief mission locations. Based on the number of rescue transportation routes corresponding to various types of rescue resources and the traffic volume of transportation hubs, the resource convenience of the locations corresponding to various types of rescue resources is assessed to obtain the location reserve convenience of various types of rescue resources. Step S143: Verify and determine the complete quantity of various types of rescue resources based on the convenience of their location and storage, so as to obtain the corresponding rescue resource reserve quantity in the region. Step S15: Assess the emergency rescue resource capacity of the disaster-stricken area based on the corresponding rescue resource reserves in the area to obtain the corresponding emergency rescue resource capacity in the current area; Step S2: Obtain the disaster type, disaster distribution scale, and surrounding population distribution scale corresponding to each emergency rescue task through the emergency disaster-stricken area; based on the disaster type, disaster distribution scale, and surrounding population distribution scale corresponding to each emergency rescue task, obtain the corresponding disaster spread rate and the number of affected people, and predict the emergency rescue resource demand required for each emergency rescue task; wherein, Step S2 includes the following steps: Step S21: Obtain the disaster type, disaster distribution scale, and surrounding population distribution scale corresponding to each emergency rescue mission through the emergency disaster-stricken area; Step S22: Set the corresponding disaster observation window size according to the corresponding disaster type, and perform a time series analysis of the disaster distribution scale and the surrounding population distribution scale corresponding to each emergency rescue task based on the disaster observation window size, so as to obtain the disaster distribution area change sequence and population distribution change sequence corresponding to each emergency rescue task under the corresponding window. Step S23: Estimate the disaster spread rate based on the disaster distribution area change sequence corresponding to each emergency rescue task under the corresponding window, so as to obtain the disaster spread rate corresponding to each emergency rescue task; Step S24: Estimate the affected population based on the population distribution change sequence corresponding to each emergency rescue task under the corresponding window, so as to obtain the number of affected people corresponding to each emergency rescue task; Step S25: Based on the disaster spread rate and the number of affected people corresponding to each emergency rescue task, and combined with AI algorithms, predict the emergency rescue resource demand for the corresponding emergency rescue tasks in the emergency disaster area, so as to predict the amount of emergency rescue resources required for each emergency rescue task. Step S3: Based on the emergency rescue resource requirements for each emergency rescue task and in combination with the disaster type, disaster distribution scale, surrounding population distribution scale, disaster spread speed and the number of affected people, conduct an emergency priority ranking analysis for each emergency rescue task to generate the corresponding emergency rescue priority sequence in the current region. Step S4: Optimize the emergency rescue allocation and management of the corresponding emergency rescue priority sequence in the current area by aiming to maximize the rescue effect and combining the corresponding emergency rescue resource capabilities in the current area, so as to generate emergency rescue resource allocation and management decision suggestions for emergency rescue tasks.

2. The emergency rescue management optimization method based on AI algorithm according to claim 1, characterized in that, Step S13 includes the following steps: Step S131: Obtain the corresponding changes in the size of equipment components and the surface roughness of the equipment by the corresponding status of various rescue equipment; Step S132: Based on the changes in the size of the equipment components and the surface roughness of the equipment corresponding to various types of rescue equipment, perform equipment wear assessment calculations to obtain the wear degree corresponding to various types of rescue equipment; Step S133: Assign corresponding weights to each performance index within the performance of various types of rescue equipment through hierarchical analysis. Each performance index includes equipment response index, equipment operation index, and equipment capability index. Then, calculate the performance score by weighting the weights and corresponding performance indexes to obtain the performance score of each type of rescue equipment. Step S134: Calculate the ratio between the wear and tear and performance scores of various types of rescue equipment to obtain the probability of good condition for each type of rescue equipment.

3. The emergency rescue management optimization method based on AI algorithm according to claim 1, characterized in that, The emergency rescue resource demand mentioned in step S25 includes the required quantities of rescue personnel, rescue equipment, and rescue supplies.

4. The emergency rescue management optimization method based on AI algorithm according to claim 1, characterized in that, Step S3 includes the following steps: Step S31: By combining the emergency rescue resource requirements for each emergency rescue task with the corresponding disaster type, disaster distribution scale, surrounding population distribution scale, disaster spread speed and the number of affected people into the corresponding emergency rescue data vector, and forming the corresponding vector matrix, an emergency rescue task vector matrix is ​​generated. Step S32: Calculate the eigenvalues ​​of each sub-vector corresponding to the emergency rescue task vector matrix to obtain the vector eigenvalues ​​corresponding to each emergency rescue task sub-vector; Step S33: Based on the vector feature values ​​corresponding to each emergency rescue task sub-vector, sort the emergency rescue tasks in the disaster area from largest to smallest to generate the corresponding emergency rescue priority sequence in the current area.

5. The emergency rescue management optimization method based on AI algorithm according to claim 1, characterized in that, Step S4 includes the following steps: Step S41: Obtain the corresponding emergency rescue resource transportation and availability constraints based on the emergency rescue resource capabilities within the current area; Step S42: Optimize the emergency rescue allocation and management of each rescue task within the current emergency rescue priority sequence by taking the goal of maximizing rescue effectiveness and combining the constraints of emergency rescue resource transportation and availability within the current region. Based on the constraints, the emergency rescue resources corresponding to each rescue task are rationally arranged according to the transportation distance, transportation time, and resource availability factors, and the optimal corresponding emergency rescue resource transportation route is selected to generate emergency rescue resource allocation and management decision suggestions for the emergency rescue task.

6. An emergency rescue management optimization system based on AI algorithms, characterized in that, The AI-based emergency rescue management optimization system is used to execute the AI-based emergency rescue management optimization method as described in claim 1, and includes: The rescue capability assessment module is used to obtain the quantity and location of various rescue resources and the status and performance of various rescue equipment in the emergency disaster area; based on the quantity and location of various rescue resources, the module assesses the rescue resource capability of various rescue equipment to obtain the corresponding emergency rescue resource capability in the current area. The resource demand prediction module is used to obtain the disaster type, disaster distribution scale, and surrounding population distribution scale corresponding to each emergency rescue task through the emergency disaster-stricken area; based on the disaster type, disaster distribution scale, and surrounding population distribution scale corresponding to each emergency rescue task, it obtains the corresponding disaster spread speed and the number of affected people, and predicts the emergency rescue resource demand required for each emergency rescue task. The emergency rescue ranking module is used to perform emergency priority ranking analysis on each emergency rescue task based on the corresponding emergency rescue resource requirements and in combination with disaster type, disaster distribution scale, surrounding population distribution scale, disaster spread speed and the number of affected people, so as to generate the corresponding emergency rescue priority sequence in the current area. The emergency rescue management module is used to optimize the allocation and management of emergency rescue priorities within the current region by aiming to maximize rescue effectiveness and combining the corresponding emergency rescue resource capabilities within the current region, so as to generate emergency rescue resource allocation and management decision suggestions for emergency rescue tasks.

Images