Emergency rescue scheme generation method based on large model

Abstract

The invention provides an emergency rescue scheme generation method based on a large model, and relates to the field of emergency rescue and the field of artificial intelligence. The method comprises the following steps: processing historical case information and disaster related data by utilizing a large model according to preset constraint prompt information to obtain an initial rescue scheme, executing the initial rescue scheme by utilizing a simulation system to obtain an initial simulation execution result, evaluating the initial simulation execution result to obtain an initial evaluation result, and according to the constraint prompt information, executing the initial rescue scheme according to the initial evaluation result. And performing semantic understanding on the initial evaluation result by using the large model to obtain a target rescue scheme meeting scheme constraint conditions.

Description

Technical Field

[0001] The present invention relates to the fields of emergency rescue and artificial intelligence, and more specifically, to a method for generating an emergency rescue plan based on a large model. Background Art

[0002] In the event of natural disasters and man-made accidents, such as earthquakes, floods, and traffic accidents, timely and effective emergency rescue actions can safeguard the lives and property of people, as well as the economy and natural resources of the affected areas. Currently, historical data in the field of emergency rescue is scarce, and the available historical data is not fully utilized. The formulation of emergency rescue plans lacks effective data support and scientific evaluation methods, resulting in low rescue efficiency and even the possibility of missing the optimal rescue opportunity. Summary of the Invention

[0003] In view of this, the present invention provides a method, device, electronic device, medium and program product for generating an emergency rescue plan based on a large model.

[0004] One aspect of the present invention provides a method for generating an emergency rescue plan based on a large model, comprising:

[0005] Based on preset constraint prompt information, a large model is used to process historical case information and disaster-related data to obtain an initial rescue plan. The disaster-related data indicates the disaster impact and rescue guarantee status of the target area, and the constraint prompt information represents the plan constraints of the rescue plan for the target area.

[0006] Utilize the simulation system to execute the initial rescue plan and obtain the initial simulation execution result;

[0007] Evaluate the initial simulation execution results to obtain initial evaluation results, wherein the initial evaluation results are used to describe the difference between the initial rescue plan and the plan constraints;

[0008] According to the constraint prompt information, the large model is used to semantically understand the initial evaluation results and obtain the target rescue plan that meets the plan constraints.

[0009] A second aspect of the present invention provides a large-scale model-based emergency rescue plan generation device, comprising:

[0010] A processing module is used to process historical case information and disaster-related data using a large model based on preset constraint prompt information to obtain an initial rescue plan, wherein the disaster-related data indicates the disaster impact and rescue guarantee status of the target area, and the constraint prompt information represents the plan constraint conditions of the rescue plan for the target area;

[0011] An execution module, used to execute the initial rescue plan using the simulation system and obtain an initial simulation execution result;

[0012] An evaluation module is used to evaluate the initial simulation execution results to obtain an initial evaluation result, which is used to describe the difference between the initial rescue plan and the plan constraints;

[0013] The semantic understanding module is used to use the large model to perform semantic understanding of the initial evaluation results based on the constraint prompt information, and obtain the target rescue plan that meets the plan constraints.

[0014] A third aspect of the present invention provides an electronic device, comprising:

[0015] one or more processors;

[0016] a memory for storing one or more programs,

[0017] When the one or more programs are executed by the one or more processors, the one or more processors implement the method as described above.

[0018] The fourth aspect of the present invention further provides a computer-readable storage medium having a computer program or instructions stored thereon, which implements the steps of the above method when the computer program or instructions are executed by a processor.

[0019] The fifth aspect of the present invention further provides a computer program product, comprising a computer program or instructions, which implement the steps of the above method when executed by a processor.

[0020] According to an embodiment of the present invention, a simulation system is used to execute an initial rescue plan, deeply integrating the simulation system with a large model. The initial simulation execution results are evaluated, and the large model is used to semantically understand the initial evaluation results to obtain a target rescue plan that meets the plan's constraints. This forms a rapid and complete rescue plan generation and optimization route. In emergency situations, the rescue plan can also be quickly and automatically updated and optimized, more closely matching the actual conditions of the disaster-stricken area, improving the rescue plan's adaptability, accuracy, and safety in complex environments, thereby improving rescue efficiency. During the iterative optimization process, multiple rescue plans that meet the plan's objectives can be obtained, enriching the case data and partially resolving the problem of a lack of historical case data. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The above and other objects, features and advantages of the present invention will become more apparent through the following description of the embodiments of the present invention with reference to the accompanying drawings, in which:

[0022] Figure 1 A diagram schematically illustrates an application scenario of a method for generating an emergency rescue plan based on a large model according to an embodiment of the present invention;

[0023] Figure 2 The flowchart of the method for generating an emergency rescue plan based on a large model according to an embodiment of the present invention is schematically shown;

[0024] Figure 3 The schematic diagram shows the process of generating and iterating an emergency rescue plan based on a large model according to an embodiment of the present invention;

[0025] Figure 4 Schematically illustrates a closed-loop process according to an embodiment of the present invention;

[0026] Figure 5 A block diagram schematically illustrates a device for generating an emergency rescue plan based on a large model according to an embodiment of the present invention; and

[0027] Figure 6 A block diagram of an electronic device suitable for implementing a large model-based emergency rescue plan generation method according to an embodiment of the present invention is schematically shown. DETAILED DESCRIPTION

[0028] Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings. However, it should be understood that these descriptions are exemplary only and are not intended to limit the scope of the present invention. In the following detailed description, for ease of explanation, many specific details are set forth to provide a comprehensive understanding of embodiments of the present invention. However, it is apparent that one or more embodiments may also be implemented without these specific details. In addition, in the following description, descriptions of known structures and technologies are omitted to avoid unnecessary confusion of the concept of the present invention.

[0029] The terms used herein are only for describing specific embodiments and are not intended to limit the present invention. The terms "comprise", "include", etc. used herein indicate the presence of the features, steps, operations and / or components, but do not exclude the presence or addition of one or more other features, steps, operations or components.

[0030] All terms used herein (including technical and scientific terms) have the meanings commonly understood by those skilled in the art unless otherwise defined. It should be noted that the terms used herein should be interpreted as having a meaning consistent with the context of this specification and should not be interpreted in an idealized or overly rigid manner.

[0031] When expressions such as "at least one of A, B, and C, etc." are used, they should generally be interpreted in accordance with the meaning commonly understood by those skilled in the art (for example, "a system having at least one of A, B, and C" should include but is not limited to a system having A alone, B alone, C alone, A and B, A and C, B and C, and / or A, B, C, etc.).

[0032] During the implementation of this invention, the inventors discovered that developing emergency rescue plans relies on manual experience, consumes significant time and manpower, and lacks effective data support and scientific evaluation methods. This results in low rescue efficiency and can even lead to missed opportunities for rescue. Historical data on emergency rescue is scarce, and the available historical data is insufficiently utilized in the development of rescue plans, failing to provide scientific and robust support for their development.

[0033] In view of this, the present invention provides a method, apparatus, electronic device, medium, and program product for generating an emergency rescue plan based on a large model. The method includes: processing historical case information and disaster-related data using a large model based on preset constraint prompt information to obtain an initial rescue plan; executing the initial rescue plan using a simulation system to obtain an initial simulation execution result; evaluating the initial simulation execution result to obtain an initial evaluation result; and semantically understanding the initial evaluation result using the large model based on the constraint prompt information to obtain a target rescue plan that meets the plan's constraints.

[0034] Figure 1 The following schematically illustrates an exemplary system architecture 100 to which a large model-based emergency rescue plan generation method according to an embodiment of the present invention can be applied. Figure 1 The examples shown are merely examples of system architectures to which the embodiments of the present invention may be applied, to help those skilled in the art understand the technical content of the present invention, but do not mean that the embodiments of the present invention cannot be used in other devices, systems, environments or scenarios.

[0035] like Figure 1 As shown, the system architecture 100 according to this embodiment may include a first terminal device 101, a second terminal device 102, a third terminal device 103, a network 104, and a server 105. The network 104 is used as a medium for providing a communication link between the first terminal device 101, the second terminal device 102, the third terminal device 103, and the server 105. The network 104 may include various connection types, such as wired and / or wireless communication links, etc.

[0036] A user may use a first terminal device 101, a second terminal device 102, or a third terminal device 103 to interact with a server 105 via a network 104 to receive or send messages, etc. Various communication client applications may be installed on the first terminal device 101, the second terminal device 102, or the third terminal device 103, such as shopping applications, web browser applications, search applications, instant messaging tools, email clients, and / or social platform software (for example only).

[0037] The first terminal device 101 , the second terminal device 102 , and the third terminal device 103 may be various electronic devices having display screens and supporting web browsing, including but not limited to smart phones, tablet computers, laptop computers, desktop computers, and the like.

[0038] The server 105 may be a server that provides various services, such as a background management server (for example only) that supports websites browsed by users using the first terminal device 101, the second terminal device 102, and the third terminal device 103. The background management server may analyze and process received data such as user requests, and feed back processing results (e.g., web pages, information, or data obtained or generated based on user requests) to the terminal devices.

[0039] It should be noted that the large-model-based emergency rescue solution generation method provided in the embodiment of the present invention can generally be executed by the server 105. Accordingly, the large-model-based emergency rescue solution generation device provided in the embodiment of the present invention can generally be set in the server 105. The large-model-based emergency rescue solution generation method provided in the embodiment of the present invention can also be executed by a server or server cluster that is different from the server 105 and can communicate with the first terminal device 101, the second terminal device 102, the third terminal device 103 and / or the server 105. Accordingly, the large-model-based emergency rescue solution generation device provided in the embodiment of the present invention can also be set in a server or server cluster that is different from the server 105 and can communicate with the first terminal device 101, the second terminal device 102, the third terminal device 103 and / or the server 105. Alternatively, the large-model-based emergency rescue solution generation method provided in the embodiment of the present invention can also be executed by the first terminal device 101, the second terminal device 102, or the third terminal device 103, or by other terminal devices different from the first terminal device 101, the second terminal device 102, or the third terminal device 103. Correspondingly, the large model-based emergency rescue plan generation device provided in the embodiment of the present invention can also be set in the first terminal device 101, the second terminal device 102 or the third terminal device 103, or in other terminal devices different from the first terminal device 101, the second terminal device 102 or the third terminal device 103.

[0040] It should be understood that Figure 1 The number of terminal devices, networks and servers in the embodiment is merely illustrative. Any number of terminal devices, networks and servers may be provided as required.

[0041] Figure 2 The flowchart of the method for generating an emergency rescue plan based on a large model according to an embodiment of the present invention is schematically shown.

[0042] like Figure 2 As shown, the method includes operations S210 to S240.

[0043] In operation S210, based on preset constraint prompt information, the large model is used to process historical case information and disaster-related data to obtain an initial rescue plan.

[0044] According to an embodiment of the present invention, the disaster-related data indicates the disaster impact and rescue guarantee situation of the target area, and the constraint prompt information represents the solution constraint conditions of the rescue solution for the target area.

[0045] In an embodiment of the present invention, constraint prompt information includes rescue target information, which indicates the objectives to be achieved by rescue operations in a target area. Historical case information represents historical rescue information collected during historical rescue operations, such as historical rescue plans, historical rescue environments, and historical rescue conditions. The target area represents the area affected by the disaster.

[0046] In an embodiment of the present invention, the constraint prompt information is used to constrain the rescue measures in the rescue plan output by the large model to meet the plan constraint conditions.

[0047] According to an embodiment of the present invention, the solution constraint conditions include at least one of the conditions for treating disaster victims, the conditions for repairing infrastructure, and the conditions for material demand.

[0048] For example, the conditions for treating disaster victims include searching for survivors, providing medical treatment to disaster victims, and safely transferring disaster victims; the conditions for repairing infrastructure include repairing transportation, ensuring communications, and repairing water and electricity supplies; and the conditions for material needs include the type, quantity, and location of rescue materials.

[0049] According to an embodiment of the present invention, using solution constraints can make the large model more consistent with the rescue needs and goals of the target area during the process of generating a rescue solution, thereby improving the accuracy and reliability of the rescue solution.

[0050] In operation S220 , the initial rescue plan is executed using the simulation system to obtain an initial simulation execution result.

[0051] In operation S230 , the initial simulation execution result is evaluated to obtain an initial evaluation result.

[0052] According to an embodiment of the present invention, the initial evaluation result is used to describe the difference between the initial rescue plan and the plan constraints.

[0053] In operation S240 , the initial evaluation result is semantically understood using the large model according to the constraint prompt information to obtain a target rescue solution that meets the solution constraint conditions.

[0054] In embodiments of the present invention, a targeted rescue plan can be used to guide rescue operations in a target area. For example, the targeted rescue plan may include parameters such as the number of rescue personnel, the type and quantity of heavy machinery, the quantity and type of rescue supplies such as tents and drinking water, and measures for mobilizing these supplies. Furthermore, the targeted rescue plan may also include other types of rescue measures, such as rescue routes and road and bridge repair measures.

[0055] According to embodiments of the present invention, a large model is used to semantically understand the initial assessment results based on constraint information to obtain information about initial solution deficiencies. Based on the constraint information and initial solution deficiencies, the large model is used to process historical case information and disaster-related data to obtain a target rescue solution that meets the solution constraints. Furthermore, the large model can be used to optimize the initial rescue solution based on the constraint information, initial solution deficiencies, and historical case information to obtain a target rescue solution that meets the solution constraints.

[0056] In an embodiment of the present invention, the initial plan defect information indicates deficiencies in the initial rescue plan. For example, deficiencies in the initial rescue plan in terms of rescue efficiency include an unreasonable rescue team route, resulting in an excessively long rescue time; deficiencies in the initial rescue plan in terms of resource utilization include an unscientific distribution plan for rescue supplies, resulting in waste of supplies; deficiencies in the initial rescue plan in terms of safety include the rescue team's failure to fully consider safety risks when performing the mission, resulting in injuries to rescue personnel, etc.

[0057] In an embodiment of the present invention, based on constraint prompt information, the large model uses semantic understanding of the initial assessment results. This allows the large model to reflect on the initial plan flaws represented by the initial assessment results. For example, the large model can examine whether the assumptions and decisions made during the generation of the initial rescue plan are reasonable, whether the topography and road network conditions of the target area are fully considered, whether the behavioral instructions in the initial rescue plan are accurate and effective, whether the rescue team's travel speed and search and rescue methods are reasonable, whether the initial rescue plan is flexible in dealing with complex environments and emergencies, and whether there are response measures in the event of secondary disasters. Based on the reflection results, a target rescue plan that meets the plan's constraints is obtained.

[0058] According to an embodiment of the present invention, a simulation system is used to execute an initial rescue plan, deeply integrating the simulation system with a large model. The initial simulation execution results are evaluated, and the large model is used to semantically understand the initial evaluation results to obtain a target rescue plan that meets the plan's constraints. This forms a rapid and complete rescue plan generation and optimization route. In emergency situations, the rescue plan can also be quickly and automatically updated and optimized, more closely matching the actual conditions of the disaster-stricken area, improving the rescue plan's adaptability, accuracy, and safety in complex environments, thereby improving rescue efficiency. During the iterative optimization process, multiple rescue plans that meet the plan's objectives can be obtained, enriching the case data and partially resolving the problem of a lack of historical case data.

[0059] According to an embodiment of the present invention, using a large model to semantically understand the initial evaluation result based on the constraint prompt information to obtain a target rescue plan that meets the plan constraints includes: when the n-1th evaluation result indicates that the n-1th rescue plan does not meet the plan objectives, using the large model to semantically understand the n-1th evaluation result and disaster-related data to obtain the nth rescue plan. According to an embodiment of the present invention, the first evaluation result is the initial evaluation result, and the first rescue plan is the initial rescue plan. According to an embodiment of the present invention, executing the nth rescue plan using a simulation system to obtain the nth simulation execution result; and evaluating the nth simulation execution result to obtain the nth evaluation result. According to an embodiment of the present invention, the nth evaluation result describes the difference between the nth rescue plan and the constraint prompt information. According to an embodiment of the present invention, when the nth evaluation result indicates that the nth rescue plan meets the plan constraints, obtaining the target rescue plan. According to an embodiment of the present invention, N ≥ n > 1, and N and n are positive integers.

[0060] In an embodiment of the present invention, if the initial assessment result indicates that the initial rescue plan does not meet the plan objectives, the large model is used to perform semantic understanding of the initial assessment result and disaster-related data to obtain a second rescue plan. If the second assessment result indicates that the second rescue plan does not meet the plan objectives, the large model is used to perform semantic understanding of the initial assessment result and disaster-related data to obtain a third rescue plan. This process is iterated until the Nth assessment result indicates that the Nth rescue plan meets the plan constraints, thereby obtaining the target rescue plan.

[0061] In an embodiment of the present invention, there may be multiple rescue solutions that meet the solution constraint condition, and at least one rescue solution is selected from the multiple rescue solutions that meet the solution constraint condition as the target rescue solution.

[0062] Figure 3 The process of generating and iterating an emergency rescue plan based on a large model according to an embodiment of the present invention is schematically illustrated.

[0063] like Figure 3As shown, the large model is used to generate the n-1 rescue plan 301. The simulation system executes the n-1 rescue plan 301, obtaining the n-1 simulation execution result 302. The n-1 simulation execution result 302 is evaluated to obtain the n-1 evaluation result 303. It is then determined whether the n-1 evaluation result 303 satisfies the plan objectives. If the plan objectives are met, a target rescue plan 305 is obtained. If the plan objectives are not met, the large model is used to perform semantic understanding on the n-1 evaluation result 303, and n is incremented by one value to obtain the n-1 rescue plan. The above process is then repeated.

[0064] According to an embodiment of the present invention, through the closed-loop route of "solution generation-simulation execution-result evaluation-optimization solution", an effective and rapid iterative optimization process for rescue solutions that do not meet the solution objectives is formed. A target rescue solution that meets the solution objectives can be obtained in a short time, and multiple rescue solutions that meet the solution objectives can also be obtained, enriching the case database and partially solving the problem of lack of historical case data.

[0065] According to an embodiment of the present invention, an initial rescue plan is executed using a simulation system to obtain an initial simulation execution result, including: determining initial simulation rescue measure parameters based on the rescue description information of the initial rescue plan; and executing the initial rescue plan using a simulation system based on the initial simulation rescue measure parameters to obtain an initial simulation execution result.

[0066] In an embodiment of the present invention, the rescue description information represents the description text for the rescue measures in the initial rescue plan. For example, the description text for the rescue measures may be "rescue team A sets out from the assembly point M and travels along the route X to the disaster site N, with an estimated travel time of T."

[0067] In an embodiment of the present invention, simulation parameters are determined based on the initial simulated rescue measure parameters; based on the simulation parameters, the initial rescue plan is executed using a simulation system to obtain an initial simulation execution result. The simulation parameters represent parameters related to rescue and the rescue environment that the simulation system needs to set to simulate the initial rescue plan.

[0068] In an embodiment of the present invention, a simulation system is used to execute behavioral instructions in the initial rescue plan, simulating the actions of the rescue team, the transportation and distribution of rescue supplies, the reactions of the affected people, and other processes to obtain initial simulation execution results.

[0069] For example, the simulation system can be used to simulate the difficulty and time it takes for rescue teams to search for trapped people in the rubble, simulate the impact of road capacity and load limitations of transportation vehicles on the transportation of rescue supplies, and simulate the survival status of disaster victims and changes in rescue needs in secondary disasters.

[0070] According to an embodiment of the present invention, the simulation system can provide real-time feedback on the initial simulation execution results, including the movement status of the rescue team, the consumption of rescue materials, the rescue situation of the affected people, etc., and predict in advance the problems that may arise during the rescue process, so that the initial rescue plan can be optimized in time.

[0071] According to an embodiment of the present invention, determining initial simulation rescue measure parameters based on the rescue description information of the initial rescue plan includes: determining disaster key data and rescue demand attribute data that match the disaster-related data from the rescue description information.

[0072] According to an embodiment of the present invention, key disaster data includes at least one of disaster location data, traffic data, material data, and medical condition data. Rescue requirement attribute data characterizes rescue methods specific to the key disaster data. Quantitative processing is performed on the key disaster data and rescue requirement attribute data to generate initial simulated rescue measure parameters that match the input of the simulation system.

[0073] In embodiments of the present invention, key disaster data can represent information such as rescue team information, location information, supplies information, vehicle information, route information, and medical information. Rescue requirement attribute data includes rescue time data, travel speed data, travel path data, and fuzzy description data specific to the key disaster data. For example, fuzzy description data specific to the key disaster data could include "Rescue Team A is advancing rapidly" or "Using large-capacity vehicles to transport rescue supplies."

[0074] In an embodiment of the present invention, data processing and conversion tools can be used to determine the initial simulation rescue measures parameters. For example, the rescue description information of the initial rescue plan can be "rescue team A departs from the assembly point M and travels along the route X to the disaster-stricken point N, with an estimated travel time of T." The disaster-critical data for this rescue description information are "rescue team A," "assembly point M," "route X," and "disaster-stricken point N," and the rescue demand attribute data for this rescue description information is "travel time T." After quantifying the above-mentioned disaster-critical data and rescue demand attribute data, the initial simulation rescue measures parameters that match the input end of the simulation system are obtained as "rescue team A's departure time T1," "rescue team A's travel speed V," and "rescue team A's arrival time T2," etc.

[0075] According to an embodiment of the present invention, determining disaster-related key data and rescue demand attribute data that match disaster-related data can make the simulation environment of the simulation system closer to the actual environmental conditions of the target area, making the simulation execution results more accurate and reliable.

[0076] According to an embodiment of the present invention, evaluating the initial simulation execution results to obtain the initial evaluation results includes: performing a multi-dimensional quantitative evaluation on the initial simulation execution results based on multiple evaluation conditions to obtain multiple quantitative evaluation information; and using a large model to process the multiple quantitative evaluation information to obtain the initial evaluation results.

[0077] In embodiments of the present invention, quantitative assessment information can characterize the differences between various dimensions of the initial rescue plan and the plan's objectives. The initial simulation execution results are quantitatively evaluated in multiple dimensions based on multiple evaluation criteria. For example, rescue efficiency is assessed by counting the number of people successfully rescued and the time it takes to complete the rescue mission; resource utilization is assessed by calculating the consumption of rescue supplies and the efficiency of the rescue team's operations; and safety is assessed by analyzing the safety risks of the rescue team and the affected individuals.

[0078] In an embodiment of the present invention, a large model is used to perform semantic understanding based on the quantitative evaluation information to obtain an initial evaluation result. For example, if the initial rescue plan differs from the plan's goal in terms of rescue efficiency, it may be that the rescue team's route was not reasonable, resulting in a long rescue time. If the initial rescue plan differs from the plan's goal in terms of resource utilization, it may be that the rescue material distribution plan was not scientific enough, resulting in material waste. If the initial rescue plan differs from the plan's goal in terms of safety, it may be that the rescue team did not fully consider safety risks during the rescue operation, resulting in injuries to rescue personnel.

[0079] In embodiments of the present invention, whether a rescue plan meets the plan objectives can also be determined based on the quantitative evaluation information and a preset threshold. For example, the preset threshold can be set to 80, or can be adjusted based on actual conditions. If the quantitative evaluation information indicates that the score of the Nth rescue plan is greater than 80, then the Nth rescue plan can be used as the target rescue plan.

[0080] According to the embodiments of the present invention, the rescue plan is evaluated from multiple dimensions such as rescue efficiency, resource utilization, and safety, which can comprehensively and objectively measure the disadvantages of the plan and perform targeted optimization of the rescue plan.

[0081] According to an embodiment of the present invention, before using the large model to process historical case information and disaster-related data, the method also includes: inputting disaster area geographic information and real-time monitoring information into the simulation system, and outputting disaster-related data, wherein the real-time monitoring information includes disaster information and rescue information.

[0082] In an embodiment of the present invention, the geographical information of the disaster area includes topographic information, road network information, building distribution information, etc. of the target area. The disaster information includes disaster severity information, disaster impact information, disaster location information, etc. The rescue information includes rescue resource information.

[0083] For example, if the disaster is an earthquake, disaster information may include the earthquake magnitude, epicenter location, aftershocks, distribution of affected people, casualties, etc. Rescue information may include the number of available rescue teams, their professional skills, equipment status, and the types, quantities, and storage locations of rescue supplies.

[0084] According to an embodiment of the present invention, using a simulation system to output disaster-related data can limit the unlimited divergence of thinking of the large model in the process of generating a rescue plan, improve the correspondence between input data and output data, and thus improve the authenticity and accuracy of the simulation system in executing the rescue plan.

[0085] According to an embodiment of the present invention, a plurality of target rescue solutions that meet solution constraints are taken as samples and stored in a sample library; the sample library is used to train a large model to obtain an optimized large model.

[0086] For example, using earthquake scenarios as an example, samples in the sample library are preprocessed, with historical case information and disaster-related data input. Based on the sample constraints, a large model is used to generate sample rescue plans. A loss function is then used to process the sample rescue plans and labeled rescue plans associated with the sample historical case information to obtain loss information. The large model is then fine-tuned using this loss information until the match between the sample rescue plans generated by the fine-tuned large model and the labeled rescue plans, such as the amount of rescue supplies and the sequence of rescue actions, meets preset requirements. This results in a trained large model. This trained large model can be used in the methods provided in the aforementioned embodiments.

[0087] In an embodiment of the present invention, three complete closed-loop routes are formed by using "simulation data input-solution generation-simulation execution" as the first closed loop, "solution generation-simulation execution-result evaluation-optimization solution" as the second closed loop, and "solution generation-optimization solution-optimization model-solution generation" as the third closed loop.

[0088] Figure 4 The closed-loop process according to an embodiment of the present invention is schematically illustrated.

[0089] like Figure 4 As shown, the first closed loop C410 includes: using the simulation system 404 to output disaster-related data and input it into the large model 402, using the large model 402 to generate the n-1th rescue plan 401 based on the disaster-related data, obtaining the n-1th simulation rescue measure parameters based on the n-1th rescue plan 401, and inputting the n-1th simulation rescue measure parameters into the simulation system 404 to execute the n-1th rescue plan 401.

[0090] The second closed loop C420 includes: generating an n-1th rescue plan 401 using the large model 402, and performing a simulation and evaluation based on the n-1th rescue plan 401 to obtain an n-1th evaluation result 406. Based on the n-1th evaluation result 406, it is determined whether the n-1th rescue plan 401 meets the plan objectives. If not, the n-1th rescue plan 401 is optimized, and n is incremented by one unit to obtain the nth rescue plan. The above process is then repeated.

[0091] The third closed loop C430 includes: obtaining a target rescue plan 408 based on the n-1 rescue plan 401, storing the target rescue plan 408 as a case in the sample library 405, and training the large model 402 based on the sample library 405. The trained large model 402 can be used in the method provided in the above embodiment.

[0092] According to an embodiment of the present invention, the first closed loop can limit the large model from unlimited divergent thinking in the process of generating rescue plans, improve the correspondence between input data and output data, and thus improve the authenticity and accuracy of the simulation system in executing rescue plans. The second closed loop forms an effective and rapid iterative optimization process for rescue plans that do not meet the plan objectives. It can obtain a target rescue plan that meets the plan objectives in a short period of time, and can also obtain multiple rescue plans that meet the plan objectives, enriching the case database and partially solving the problem of lack of historical case data. The third closed loop can improve the pertinence of the large model in generating emergency rescue scenarios and rescue plans, thereby improving the accuracy and reliability of the target rescue plan, shortening the time to determine the target rescue plan, and improving rescue efficiency.

[0093] Figure 5 A block diagram of a large model-based emergency rescue plan generation device according to an embodiment of the present invention is schematically shown.

[0094] like Figure 5 As shown, the large model-based emergency rescue plan generation device 500 includes a processing module 510, an execution module 520, an evaluation module 530, and a semantic understanding module 540.

[0095] Processing module 510 is used to use a large model to process historical case information and disaster-related data based on preset constraint prompt information to obtain an initial rescue plan, wherein the disaster-related data indicates the disaster impact and rescue guarantee situation in the target area, and the constraint prompt information represents the plan constraint conditions of the rescue plan for the target area.

[0096] The execution module 520 is used to execute the initial rescue plan using the simulation system to obtain an initial simulation execution result.

[0097] The evaluation module 530 is used to evaluate the initial simulation execution result to obtain an initial evaluation result, where the initial evaluation result is used to describe the difference between the initial rescue plan and the plan constraints.

[0098] The semantic understanding module 540 is used to perform semantic understanding on the initial evaluation result based on the constraint prompt information using the large model to obtain a target rescue solution that meets the solution constraint conditions.

[0099] According to an embodiment of the present invention, a simulation system is used to execute an initial rescue plan, deeply integrating the simulation system with a large model. The initial simulation execution results are evaluated, and the large model is used to semantically understand the initial evaluation results to obtain a target rescue plan that meets the plan's constraints. This forms a rapid and complete rescue plan generation and optimization route. In emergency situations, the rescue plan can also be quickly and automatically updated and optimized, more closely matching the actual conditions of the disaster-stricken area, improving the rescue plan's adaptability, accuracy, and safety in complex environments, thereby improving rescue efficiency. During the iterative optimization process, multiple rescue plans that meet the plan's objectives can be obtained, enriching the case data and partially resolving the problem of a lack of historical case data.

[0100] According to an embodiment of the present invention, the semantic understanding module 540 includes a semantic understanding submodule, an execution submodule, an evaluation submodule, and a target solution confirmation submodule.

[0101] The semantic understanding submodule is used to use the large model to perform semantic understanding on the n-1th evaluation result and the disaster-related data to obtain the nth rescue plan when the n-1th evaluation result indicates that the n-1th rescue plan does not meet the plan goal, wherein the first evaluation result is the initial evaluation result and the first rescue plan is the initial rescue plan.

[0102] The execution submodule is used to execute the nth rescue plan using the simulation system to obtain the nth simulation execution result.

[0103] An evaluation submodule is configured to evaluate the nth simulation execution result to obtain an nth evaluation result, wherein the nth evaluation result describes the difference between the nth rescue solution and the constraint prompt information.

[0104] The target solution confirmation submodule is used to obtain a target rescue solution when the Nth evaluation result indicates that the Nth rescue solution meets the solution constraint conditions, wherein N≥n>1, and N and n are positive integers.

[0105] According to an embodiment of the present invention, the execution module 520 includes a parameter determination submodule and an initial solution execution submodule.

[0106] The parameter determination submodule is used to determine the initial simulation rescue measure parameters according to the rescue description information of the initial rescue plan.

[0107] The initial plan execution submodule is used to execute the initial rescue plan using the simulation system according to the initial simulation rescue measure parameters to obtain the initial simulation execution result.

[0108] According to an embodiment of the present invention, the parameter determination submodule includes a data determination unit and a quantization processing unit.

[0109] A data determination unit is used to determine disaster key data and rescue demand attribute data that match the disaster-related data from the rescue description information, wherein the disaster key data includes at least one of disaster-stricken location data, traffic data, material data and medical condition data, and the rescue demand attribute data represents the rescue method for the disaster key data.

[0110] The quantitative processing unit is used to perform quantitative processing on the disaster key data and rescue demand attribute data to obtain the initial simulation rescue measure parameters that match the input end of the simulation system.

[0111] According to an embodiment of the present invention, the evaluation module 530 includes a quantitative evaluation submodule and an information processing submodule.

[0112] The quantitative evaluation submodule is used to perform a multi-dimensional quantitative evaluation on the initial simulation execution result based on multiple evaluation conditions to obtain multiple quantitative evaluation information.

[0113] The information processing submodule is used to use the large model to process the plurality of quantitative evaluation information to obtain the initial evaluation result.

[0114] According to an embodiment of the present invention, the large model-based emergency rescue plan generation device 500 further includes an information input module.

[0115] The information input module is used to input the geographical information of the disaster area and the real-time monitoring information into the simulation system and output the disaster-related data, wherein the real-time monitoring information includes disaster information and rescue information.

[0116] According to an embodiment of the present invention, the solution constraints include conditions for treating disaster victims, conditions for repairing infrastructure, and conditions for material needs.

[0117] According to an embodiment of the present invention, the large model-based emergency rescue plan generation device 500 further includes a storage module and a training module.

[0118] The storage module is used to store multiple target rescue solutions that meet the solution constraints as samples in a sample library.

[0119] The training module is used to train the large model using the sample library to obtain the optimized large model.

[0120] Any number of the modules, submodules, units, and subunits according to embodiments of the present invention, or at least part of the functionality of any number of these units, can be implemented in a single module. Any one or more of the modules, submodules, units, and subunits according to embodiments of the present invention can be split into multiple modules for implementation. Any one or more of the modules, submodules, units, and subunits according to embodiments of the present invention can be at least partially implemented as a hardware circuit, such as a field programmable gate array (FPGA), a programmable logic array (PLA), a system on a chip, a system on a substrate, a system on a package, an application-specific integrated circuit (ASIC), or can be implemented in hardware or firmware using any other reasonable method of circuit integration or packaging, or can be implemented in any one of the three implementation methods of software, hardware, and firmware, or any appropriate combination of any of these. Alternatively, one or more of the modules, submodules, units, and subunits according to embodiments of the present invention can be at least partially implemented as a computer program module that, when executed, can perform the corresponding functionality.

[0121] For example, any number of the processing module 510, the execution module 520, the evaluation module 530, and the semantic understanding module 540 can be combined into a single module / unit / sub-unit, or any one of these modules / units / sub-units can be split into multiple modules / units / sub-units. Alternatively, at least part of the functionality of one or more of these modules / units / sub-units can be combined with at least part of the functionality of other modules / units / sub-units and implemented in a single module / unit / sub-unit. According to an embodiment of the present invention, at least one of the processing module 510, the execution module 520, the evaluation module 530, and the semantic understanding module 540 can be at least partially implemented as a hardware circuit, such as a field programmable gate array (FPGA), a programmable logic array (PLA), a system on a chip, a system on a substrate, a system on a package, an application-specific integrated circuit (ASIC), or can be implemented in hardware or firmware by any other reasonable means of circuit integration or packaging, or can be implemented in any one of the three implementation methods of software, hardware, and firmware, or any appropriate combination of any of these. Alternatively, at least one of the processing module 510 , the execution module 520 , the evaluation module 530 , and the semantic understanding module 540 may be at least partially implemented as a computer program module, which may perform corresponding functions when executed.

[0122] Figure 6A block diagram of an electronic device suitable for implementing a large model-based emergency rescue plan generation method according to an embodiment of the present invention is schematically shown. Figure 6 The electronic device shown is only an example and should not limit the functions and scope of use of the embodiments of the present invention.

[0123] like Figure 6 As shown, an electronic device 600 according to an embodiment of the present invention includes a processor 601, which can perform various appropriate actions and processes based on programs stored in a ROM (Read Only Memory) 602 or programs loaded from a storage unit 608 into a RAM (Random Access Memory) 603. Processor 601 may include, for example, a general-purpose microprocessor (e.g., a CPU), an instruction set processor and / or related chipsets, and / or a specialized microprocessor (e.g., an application-specific integrated circuit (ASIC)). Processor 601 may also include onboard memory for caching. Processor 601 may include a single processing unit or multiple processing units for performing the different actions of the method flow according to an embodiment of the present invention.

[0124] RAM 603 stores various programs and data required for the operation of electronic device 600. Processor 601, ROM 602, and RAM 603 are interconnected via bus 604. Processor 601 executes the programs in ROM 602 and / or RAM 603 to perform various operations according to the method flow of the embodiment of the present invention. It should be noted that the programs may also be stored in one or more memories other than ROM 602 and RAM 603. Processor 601 may also execute the programs stored in one or more memories to perform various operations according to the method flow of the embodiment of the present invention.

[0125] According to an embodiment of the present invention, electronic device 600 may further include an input / output (I / O) interface 605, which is also connected to bus 604. Electronic device 600 may also include one or more of the following components connected to I / O interface 605: an input section 606 including a keyboard, mouse, etc.; an output section 607 including devices such as a cathode ray tube (CRT), liquid crystal display (LCD), and speakers; a storage section 608 including a hard disk; and a communication section 609 including a network interface card such as a LAN card or modem. Communication section 609 performs communication processing via a network such as the Internet. A drive 610 is also connected to I / O interface 606 as needed. Removable media 611, such as a magnetic disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed in drive 610 as needed, so that computer programs read from the removable media can be installed into storage section 608 as needed.

[0126] According to an embodiment of the present invention, the method flow according to an embodiment of the present invention can be implemented as a computer software program. For example, an embodiment of the present invention includes a computer program product, which includes a computer program carried on a computer-readable storage medium, and the computer program includes a program code for executing the method shown in the flowchart. In such an embodiment, the computer program can be downloaded and installed from the network through the communication part 609, and / or installed from the removable medium 611. When the computer program is executed by the processor 601, the above-mentioned functions defined in the system of the embodiment of the present invention are executed. According to an embodiment of the present invention, the system, device, apparatus, module, unit, etc. described above can be implemented by a computer program module.

[0127] The present invention also provides a computer-readable storage medium, which may be included in the device / apparatus / system described in the above embodiments, or may exist independently and not incorporated into the device / apparatus / system. The computer-readable storage medium carries one or more programs, which, when executed, implement the method according to the embodiments of the present invention.

[0128] According to embodiments of the present invention, a computer-readable storage medium may be a non-volatile computer-readable storage medium. Examples include, but are not limited to, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the present invention, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device.

[0129] For example, according to an embodiment of the present invention, the computer-readable storage medium may include the ROM 602 and / or the RAM 603 described above and / or one or more memories other than the ROM 602 and the RAM 603 .

[0130] An embodiment of the present invention also includes a computer program product, which includes a computer program, which contains program code for executing the method provided by the embodiment of the present invention. When the computer program product is run on an electronic device, the program code is used to enable the electronic device to implement the large model-based emergency rescue plan generation method provided by the embodiment of the present invention.

[0131] When the computer program is executed by the processor 601, the above functions defined in the system / device of the embodiment of the present invention are performed. According to the embodiment of the present invention, the above-described systems, devices, modules, units, etc. can be implemented by computer program modules.

[0132] In one embodiment, the computer program may be stored on a tangible storage medium such as an optical storage device or a magnetic storage device. In another embodiment, the computer program may be transmitted and distributed in the form of a signal on a network medium, downloaded and installed via the communication portion 609, and / or installed from a removable medium 611. The program code contained in the computer program may be transmitted using any appropriate network medium, including but not limited to wireless, wired, or any suitable combination thereof.

[0133] According to an embodiment of the present invention, the program code for executing the computer program provided by the embodiment of the present invention can be written in any combination of one or more programming languages. Specifically, these computer programs can be implemented using high-level procedural and / or object-oriented programming languages, and / or assembly / machine languages. Programming languages include, but are not limited to, languages such as Java, C++, Python, "C" or similar programming languages. The program code can be executed entirely on the user computing device, partially on the user device, partially on a remote computing device, or entirely on a remote computing device or server. In the case of a remote computing device, the remote computing device can be connected to the user computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computing device (for example, using an Internet service provider to connect via the Internet).

[0134] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of the systems, methods, and computer program products according to various embodiments of the present invention. In this regard, each box in the flowchart or block diagram may represent a module, program segment, or portion of code, which contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the boxes may occur in an order different from that marked in the accompanying drawings. For example, two boxes shown in succession may actually be executed substantially in parallel, or they may sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flowchart, as well as the combination of boxes in the block diagram or flowchart, may be implemented using a dedicated hardware-based system that performs the specified function or operation, or may be implemented using a combination of dedicated hardware and computer instructions. It will be understood by those skilled in the art that the features described in the various embodiments of the present invention may be combined and / or coupled in various ways, even if such combinations or couplings are not explicitly described in the present invention. In particular, the features described in the various embodiments of the present invention may be combined and / or coupled in various ways without departing from the spirit and teachings of the present invention. All such combinations and / or combinations fall within the scope of the present invention.

[0135] The above describes embodiments of the present invention. However, these embodiments are for illustrative purposes only and are not intended to limit the scope of the present invention. Although each embodiment has been described separately above, this does not mean that the measures in each embodiment cannot be advantageously used in combination. The present invention does not depart from the scope of the present invention, and those skilled in the art may make various substitutions and modifications, which are all intended to fall within the scope of the present invention.

Claims

1. A method for generating an emergency rescue plan based on a large model, comprising: Based on preset constraint prompt information, a large model is used to process historical case information and disaster-related data to obtain an initial rescue plan, wherein the disaster-related data indicates the disaster impact and rescue guarantee status of the target area, and the constraint prompt information represents the plan constraints of the rescue plan for the target area; Executing the initial rescue plan using a simulation system to obtain an initial simulation execution result; Evaluating the initial simulation execution result to obtain an initial evaluation result, wherein the initial evaluation result is used to describe the difference between the initial rescue plan and the plan constraint conditions; According to the constraint prompt information, the large model is used to perform semantic understanding on the initial evaluation result to obtain a target rescue plan that meets the plan constraint conditions.

2. The method according to claim 1, wherein The method of using the large model to perform semantic understanding on the initial evaluation result based on the constraint prompt information to obtain a target rescue solution that meets the solution constraint conditions includes: When the n-1th evaluation result indicates that the n-1th rescue plan does not meet the plan goal, the large model is used to perform semantic understanding on the n-1th evaluation result and the disaster-related data to obtain the nth rescue plan, wherein the first evaluation result is the initial evaluation result and the first rescue plan is the initial rescue plan; Executing the nth rescue plan using the simulation system to obtain an nth simulation execution result; evaluating the nth simulation execution result to obtain an nth evaluation result, wherein the nth evaluation result describes a difference between the nth rescue solution and the constraint prompt information; When the Nth evaluation result indicates that the Nth rescue solution satisfies the solution constraint condition, a target rescue solution is obtained, wherein N≥n>1, and N and n are positive integers.

3. The method according to claim 1, wherein The utilizing the simulation system to execute the initial rescue plan and obtain an initial simulation execution result includes: Determining initial simulation rescue measure parameters according to the rescue description information of the initial rescue plan; The initial rescue plan is executed using the simulation system according to the initial simulation rescue measure parameters to obtain the initial simulation execution result.

4. The method according to claim 3, wherein: The determining of initial simulation rescue measure parameters according to the rescue description information of the initial rescue plan includes: Determining, from the rescue description information, disaster-related key data and rescue requirement attribute data that match the disaster-related data, wherein the disaster-related key data includes at least one of disaster-affected location data, traffic data, material data, and medical condition data, and the rescue requirement attribute data represents a rescue method for the disaster-related key data; The disaster key data and rescue demand attribute data are quantified to obtain the initial simulation rescue measure parameters that match the input end of the simulation system.

5. The method according to claim 1, wherein The evaluating the initial simulation execution result to obtain the initial evaluation result includes: Performing a multi-dimensional quantitative evaluation on the initial simulation execution result based on multiple evaluation conditions to obtain multiple quantitative evaluation information; The large model is used to process a plurality of the quantitative evaluation information to obtain the initial evaluation result.

6. The method according to claim 1, wherein Before using the large model to process historical case information and disaster-related data, the method also includes: inputting disaster area geographic information and real-time monitoring information into the simulation system, and outputting the disaster-related data, wherein the real-time monitoring information includes disaster information and rescue information.

7. The method according to claim 1, wherein The constraints of the plan include conditions for treating disaster victims, conditions for repairing infrastructure, and conditions for material needs.

8. The method according to claim 1, wherein The method further comprises: Storing multiple target rescue solutions that meet the solution constraints as samples in a sample library; The large model is trained using the sample library to obtain the optimized large model.

9. A device for generating an emergency rescue plan based on a large model, comprising: a processing module configured to process historical case information and disaster-related data using a large model based on preset constraint prompt information to obtain an initial rescue plan, wherein the disaster-related data indicates the disaster impact and rescue guarantee status of the target area, and the constraint prompt information represents the plan constraint conditions of the rescue plan for the target area; An execution module, configured to execute the initial rescue plan using a simulation system to obtain an initial simulation execution result; An evaluation module, configured to evaluate the initial simulation execution result to obtain an initial evaluation result, wherein the initial evaluation result is used to describe the difference between the initial rescue plan and the plan constraint conditions; The semantic understanding module is used to perform semantic understanding on the initial evaluation result based on the constraint prompt information using the large model to obtain a target rescue plan that meets the plan constraint conditions.

10. An electronic device comprising: one or more processors; a memory for storing one or more programs, When the one or more programs are executed by the one or more processors, the one or more processors are enabled to implement the method according to any one of claims 1 to 8.

Images