Emergency command decision generation method and system based on AI large model

Abstract

The invention discloses an emergency command decision generation method and system based on an AI large model. The method comprises the steps that accident visual data and pre-accident visual data of an emergency occurrence area are acquired; the accident visual data and the pre-accident visual data are processed through an AI large model, and an accident description file of the emergency occurrence area is generated; and according to the accident description file, matching an emergency command decision corresponding to the accident description file from preset alternative emergency command decisions.

Description

Technical Field

[0001] The present application belongs to the field of artificial intelligence technology, and specifically relates to an emergency command decision-making generation method and system based on an AI big model. Background Art

[0002] In relevant emergency command practices, emergency command decisions can be generated in the following ways: based on text templates and scenario materials, generated through keyword matching analysis. Or based on collected speech, the speech is converted into text, and then generated through the emergency resource scheduling analysis model.

[0003] However, the current status of the accident disaster scene or the rescue progress is an important basis for determining emergency command decisions. When faced with complex and changeable emergency situations, it is difficult to accurately describe the accident disaster scene or the rescue progress through text and voice, which will result in the generated emergency command decisions being unable to accurately handle the accident. Summary of the invention

[0004] The embodiments of the present application provide an emergency command decision-making generation method and system based on an AI big model, which can solve the problem that it is difficult to accurately describe complex accident disaster scenes or rescue progress using text and voice, which will result in the generated emergency command decisions being unable to accurately handle accidents.

[0005] In a first aspect, an embodiment of the present application provides a method for generating emergency command decisions based on an AI big model, the method comprising: obtaining accident visual data and pre-accident visual data of the area where the emergency occurred; processing the accident visual data and the pre-accident visual data through the AI ​​big model to generate an accident description file for the area where the emergency occurred; and according to the accident description file, matching the emergency command decision corresponding to the accident description file from preset alternative emergency command decisions.

[0006] In the second aspect, an embodiment of the present application provides an emergency command decision-making system based on an AI big model, the system comprising: an acquisition module for acquiring accident visual data of the area where the emergency occurs, and pre-accident visual data of the area where the emergency occurs; wherein the visual data comprises images or videos; a generation module for processing the accident visual data and the pre-accident visual data through the AI ​​big model to generate an accident description file for the area where the emergency occurs; a matching module for matching the emergency command decision corresponding to the accident description file from preset alternative emergency command decisions according to the accident description file.

[0007] In a third aspect, an embodiment of the present application provides an electronic device, which includes a processor, a memory, and a program or instruction stored in the memory and executable on the processor, wherein the program or instruction, when executed by the processor, implements the steps of the method described in the first aspect.

[0008] In a fourth aspect, an embodiment of the present application provides a readable storage medium, on which a program or instruction is stored, and when the program or instruction is executed by a processor, the steps of the method described in the first aspect are implemented.

[0009] In a fifth aspect, an embodiment of the present application provides a chip, comprising a processor and a communication interface, wherein the communication interface is coupled to the processor, and the processor is used to run a program or instruction to implement the steps of the method described in the first aspect.

[0010] In a sixth aspect, an embodiment of the present application provides a computer program product, comprising a computer program stored on a non-transitory computer-readable storage medium, wherein the computer program comprises a program or instructions, and when the program or instructions are executed, the steps of the method described in the first aspect are implemented.

[0011] In an embodiment of the present application, by acquiring accident visual data and pre-accident visual data of the area where the emergency occurred, the accident visual data and the pre-accident visual data are processed by an AI large model to generate an accident description file for the area where the emergency occurred. According to the accident description file, the emergency command decision corresponding to the accident description file is matched from preset alternative emergency command decisions, thereby generating an accident description file based on the accident visual data of the area where the emergency occurred, accurately assessing the specific on-site conditions of the area where the emergency occurred, and then accurately matching the emergency command decision based on the accident description file to improve the accuracy of the emergency command decision. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 It is a flowchart of an emergency command decision-making method based on an AI big model provided in an embodiment of the present application;

[0013] Figure 2 It is a structural diagram of an emergency command decision-making system based on an AI big model provided in an embodiment of the present application;

[0014] Figure 3 It is a structural diagram of another emergency command decision-making system based on an AI big model provided in an embodiment of the present application;

[0015] Figure 4 It is a flowchart of another emergency command decision-making method based on an AI big model provided in an embodiment of the present application;

[0016] Figure 5 It is a flowchart of a decision-making method for multiple decision nodes provided in an embodiment of the present application;

[0017] Figure 6 It is a structural schematic diagram of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0018] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0019] The terms "first", "second", etc. in the specification and claims of the present application are used to distinguish similar objects, and are not used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable under appropriate circumstances, so that the embodiments of the present application can be implemented in an order other than those illustrated or described here, and the objects distinguished by "first", "second", etc. are generally of one type, and the number of objects is not limited. For example, the first object can be one or more. In addition, "and / or" in the specification and claims represents at least one of the connected objects, and the character " / " generally indicates that the objects associated with each other are in an "or" relationship.

[0020] In the following, in combination with the accompanying drawings, the emergency command decision-making method and system based on the AI ​​big model provided by the embodiment of the present application are described in detail through specific embodiments and their application scenarios.

[0021] Figure 1 The flowchart of a method for generating emergency command decisions based on an AI large model provided by an embodiment of the present application is shown, and the method can be executed by an electronic device. Figure 1 , the method may include the following steps.

[0022] Step 102, acquiring accident visual data of the area where the emergency occurs and pre-accident visual data.

[0023] Among them, the accident visual data of the emergency area can be obtained through remote sensing technology or monitoring equipment in the emergency area. For the monitoring equipment in the emergency area, the control authority of the monitoring equipment in the emergency area is first obtained by using a security authentication algorithm and an access control algorithm, and then the communication status of the monitoring equipment is judged according to the TCP / UDP port opening of the monitoring equipment by using a Transmission Control Protocol (TCP) / User Datagram Protocol (UDP) port detection algorithm, and finally the accident visual data collected regularly by the monitoring equipment in the emergency area is called, such as images or videos.

[0024] Among them, pre-accident visual data can be obtained through City Information Modeling (CIM) / Building Information Modeling (BIM). The CIM / BIM model includes various information about the area or building before the emergency occurs. For example, the material, size and function of the building, the geographical location of the area or building, and multimedia materials such as images and videos.

[0025] Step 104, the accident visual data and the pre-accident visual data are processed by an AI big model to generate an accident description file of the area where the emergency occurred.

[0026] Among them, the AI ​​big model includes a layer that can process visual data efficiently and accurately. By processing the accident visual data and pre-accident visual data through the AI ​​big model, it is possible to obtain an accident description file of the area where the emergency occurred, so as to accurately assess the specific on-site conditions of the emergency area, such as the damage to the area where the emergency occurred and the damage to the buildings in the area where the emergency occurred.

[0027] Step 106: According to the accident description file, an emergency command decision corresponding to the accident description file is matched from preset alternative emergency command decisions.

[0028] Among them, the preset alternative emergency command decisions can be composed according to the circumstances of emergencies that have occurred in the history and the emergency command decisions formulated for each emergency.

[0029] In an embodiment of the present application, by acquiring accident visual data and pre-accident visual data of the area where the emergency occurred, the accident visual data and the pre-accident visual data are processed by an AI large model to generate an accident description file for the area where the emergency occurred. According to the accident description file, the emergency command decision corresponding to the accident description file is matched from preset alternative emergency command decisions, thereby generating an accident description file based on the accident visual data of the area where the emergency occurred, accurately assessing the specific on-site conditions of the area where the emergency occurred, and then accurately matching the emergency command decision based on the accident description file to improve the accuracy of the emergency command decision.

[0030] In one implementation, the above step 104 processes the accident visual data and the pre-accident visual data through an AI big model to generate an accident description file of the area where the emergency occurred, which may include the following steps.

[0031] Step 1041, compare and analyze the accident visual data with the pre-accident visual data through the deep learning layer of the AI ​​big model to obtain the comprehensive loss rate of the area where the emergency occurred.

[0032] Among them, the deep learning layer of the AI ​​large model detects and classifies the accident objects in the emergency area based on the acquired accident visual data, and compares and analyzes the situation with the pre-accident visual data of the emergency area to obtain the difference before and after the accident in the emergency area, thereby determining the comprehensive loss rate of the emergency area.

[0033] Step 1042: Determine the accident level of the area where the emergency occurs based on the comprehensive loss rate.

[0034] Among them, the accident level of the emergency area is determined by the comprehensive loss rate, so as to summarize the accident situation in the emergency area through the accident level, so as to quickly know the accident situation in the emergency area. Optionally, the accident level can be divided into four levels: extremely serious, serious, major and general.

[0035] Step 1043, the accident visual data and the accident level are processed by the natural language layer of the AI ​​big model to generate text, and obtain an accident description file of the area where the emergency occurred.

[0036] The accident description file may include a description of the on-site situation in the area where the emergency occurred, the accident level, damage to the accident object, etc. An example of an accident description file may be that a building B at a location A has suffered a major level of damage.

[0037] In an embodiment of the present application, the accident visual data and the pre-accident visual data are compared and analyzed by the deep learning layer of the AI ​​big model, for example, the comparison of the damaged area and the total area, the comparison of the conditions of various types of accident objects in the area where the emergency occurred before and after the accident, etc., to obtain the comprehensive loss rate of the area where the emergency occurred. Then, based on the comprehensive loss rate, the accident level of the area where the emergency occurred is determined, and the accident visual data and the accident level are processed through the natural language layer of the AI ​​big model to generate text to obtain an accident description file of the area where the emergency occurred, so as to extract core, important, or useful information from the various information obtained and analyzed, so as to generate an accident description file that accurately describes the area where the emergency occurred, so as to accurately match the emergency command decision according to the accident description file.

[0038] In one implementation, the above-mentioned step 1041 compares and analyzes the accident visual data with the pre-accident visual data through the deep learning layer of the AI ​​big model to obtain the comprehensive loss rate of the area where the emergency occurred, which may include the following steps.

[0039] Step 21, constructing a first virtual area model of the area where the emergency occurred based on the accident visual data; and constructing a second virtual area model of the area where the emergency occurred based on the pre-accident visual data.

[0040] Among them, the first virtual area model is a virtual area model before the accident occurs in the emergency area, and the second virtual area model is a virtual area model after the accident occurs in the emergency area. The virtual area model can be a three-dimensional model to facilitate comparative analysis of the conditions of the emergency area before and after the accident and the conditions of various types of accident objects therein.

[0041] Step 22, comparing and analyzing the first virtual area model and the second virtual area model, respectively obtaining the scale loss rate, structure loss rate and function loss rate of each type of accident object in the emergency occurrence area.

[0042] For example, the integrity of a building in the second virtual area model before the accident is 100%, while the integrity in the first virtual area model after the accident is 20%. By comparing the two, the loss of the building can be obtained. Furthermore, the scale loss rate, structural loss rate, and functional loss rate of each type of accident object in the area where the emergency occurred can be determined based on more detailed parameters and attributes. For example, the scale loss rate can be determined by the area of ​​the area where the emergency occurred before and after the accident, or by the number of each type of accident object (such as buildings) in the area where the emergency occurred and the total number of objects before the accident. The scale loss rate D of a large community building scale = Number of damaged buildings / total number of buildings.

[0043] Step 23, determining the comprehensive loss rate of the area where the emergency occurs according to the scale loss rate, structure loss rate and function loss rate of each type of accident object.

[0044] Optionally, the loss situation in the area where the emergency occurred can be represented by vectors and matrices. For example, the area where the emergency occurred includes four buildings. Through comparative analysis of the AI ​​big model, we can get D = (D scale , D structure , D function ), where D is the comprehensive loss ratio, D scale is the scale loss rate, D structure is the structural loss rate, D function is the functional loss rate, D scale , D structure , D function All of them are column vectors with a length of 4. The three are concatenated into D, which is used to represent the loss situation in the area where the emergency occurred. D can intuitively represent the loss situation of various types of accident objects in three aspects, and can be used to more accurately represent the loss of accident objects with large single bodies, irregular shapes or special functions, such as nuclear power / electric facilities, large commercial entities, etc.

[0045] In an embodiment of the present application, based on the AI ​​big model, according to the accident visual data, a first virtual area model of the emergency area is constructed to reproduce the scene after the accident in the emergency area. And according to the pre-accident visual data, a second virtual area model of the emergency area is constructed to reproduce the scene before the accident in the emergency area. Thus, by comparing and analyzing the first virtual area model with the second virtual area model, the scale loss rate, structural loss rate and functional loss rate of each type of accident object in the emergency area can be obtained respectively, and then the comprehensive loss rate of the emergency area is determined according to the scale loss rate, structural loss rate and functional loss rate of each type of accident object. Based on visual data that can more accurately describe the status of the emergency area before and after the accident, such as images or videos of the scene, a more accurate comprehensive loss rate of the emergency area can be obtained to generate an accurate accident description file and determine an accurate emergency command decision.

[0046] In one implementation, the above step 23 determines the comprehensive loss rate of the emergency event occurrence area according to the scale loss rate, structural loss rate and functional loss rate of each type of accident object, including the following steps.

[0047] Step 231, constructing a comprehensive loss matrix of the emergency event area according to the scale loss rate, structure loss rate, function loss rate of each type of accident object and the scale weight corresponding to the scale loss rate, the structure weight corresponding to the structure loss rate, and the function weight corresponding to the function loss rate.

[0048] In the above example, the emergency area includes four buildings, D scale , D structure , D function are all column vectors of length 4, D = k scale ·D scale +k structure ·D structure +k function ·D function , forming a 4×1 comprehensive loss matrix D. Each element in D represents the total loss of each type of accident object in three aspects. The loss situation is described by the sum of scale loss rate, structural loss rate, functional loss rate and their weights.

[0049] Step 232, superimpose the elements of the comprehensive loss matrix to obtain the comprehensive loss rate of the area where the emergency occurs.

[0050] In an embodiment of the present application, based on the scale loss rate, structural loss rate, functional loss rate of each type of accident object and the scale weight corresponding to the scale loss rate, the structural weight corresponding to the structural loss rate, and the functional weight corresponding to the functional loss rate, a comprehensive loss matrix that can describe the total loss situation of each type of accident object in three aspects is constructed, and each element corresponds to the total loss situation of each type of accident object. The total loss situations of various types of accident objects are then superimposed, that is, the elements of the comprehensive loss matrix are superimposed to obtain the comprehensive loss rate of the area where the emergency occurs, and then the accident level of the area where the emergency occurs is determined based on the comprehensive loss rate, so as to generate accurate accident description files and determine accurate emergency command decisions.

[0051] In one implementation, the preset alternative emergency command decision includes emergency event information and handling actions corresponding to the emergency event information; the above-mentioned step 106 matches the emergency command decision corresponding to the accident description file from the preset alternative emergency command decisions according to the accident description file, and may include the following steps.

[0052] Step 1061, extracting key words according to the weight of each word in the accident description file.

[0053] Among them, by dividing the accident description file, a word set, a punctuation mark set and a grammatical structure set are obtained. The weight of each word in the accident description file is expressed by calculating the distance of each word in the word set. The distance of each word can be calculated by the following formula:

[0054] Where L is the total number of words, D i represents the distance of the i-th word, and the frequency of words and the relationship between adjacent words are considered in the calculation process; p i represents the frequency of the i-th word, W represents the total number of words in the accident description file, and tanh represents the hyperbolic tangent function, which compresses the data into the range of [-1,1] to better capture the relationship between words; p i ⊙p i+1 The Hadamard product representing the frequency of adjacent words is used to measure the relationship between words. The distances of each word are then combined to form the accident description matrix A, and key words are selected from the weights of each word in A. The selection method can be to select a preset number of words with the highest weight ranking. In one example, A can be:

[0055]

[0056] Step 1062: Match the keyword with the emergency event information to determine target emergency event information that matches the keyword.

[0057] Among them, since the key words describe the area where the emergency occurs more accurately, the key words are matched with the emergency information in the preset alternative emergency command decision, and the target emergency information matching the key words is determined. The preset alternative emergency command decision includes the emergency information and the handling actions corresponding to the emergency information, so that the corresponding handling actions can be determined according to the target emergency information.

[0058] Step 1063: Determine the target handling action corresponding to the accident description file according to the target emergency information.

[0059] In an embodiment of the present application, keywords are extracted from the accident description file, and the keywords are matched with the emergency information in the preset alternative emergency command decision to determine the target emergency information that matches the keyword. Since the preset alternative emergency command decision includes the emergency information and the handling action corresponding to the emergency information, the corresponding handling action can be determined according to the target emergency information, thereby realizing the accurate determination of the emergency command decision according to the accident description file.

[0060] In one implementation, after matching the emergency command decision corresponding to the accident description file from preset alternative emergency command decisions according to the accident description file in the above step 106, the method further includes the following steps.

[0061] Step 108, periodically acquiring visual data after the emergency command decision is executed in the area where the emergency occurs.

[0062] The visual data after the emergency command decision is executed in the emergency occurrence area may be visual data obtained after actual execution, or may be visual data obtained after execution based on a simulation method.

[0063] Step 110, analyzing the visual data to determine a feasibility assessment result of the emergency command decision;

[0064] Step 112: updating the preset alternative emergency command decision according to the feasibility assessment result.

[0065] In an embodiment of the present application, visual data after the emergency command decision is executed in the area where the emergency occurs is periodically obtained, the feasibility of the emergency command decision is analyzed based on the visual data, and then the preset alternative emergency command decision is updated according to the obtained feasibility assessment results, for example, the process of handling actions is adjusted, the execution method is improved, etc.

[0066] Figure 2The schematic diagram of the structure of an emergency command decision-making system based on an AI big model provided in an embodiment of the present application is shown. Figure 2 As shown, the system 200 includes: an acquisition module 21, a generation module 22 and a matching module 23.

[0067] Among them, the acquisition module 21 is used to obtain the accident visual data of the area where the emergency occurred, and the pre-accident visual data of the area where the emergency occurred; wherein the visual data includes images or videos; the generation module 22 is used to process the accident visual data and the pre-accident visual data through the AI ​​big model to generate an accident description file of the area where the emergency occurred; the matching module 23 is used to match the emergency command decision corresponding to the accident description file from the preset alternative emergency command decisions according to the accident description file.

[0068] In one implementation, the generation module 22 can be used to compare and analyze the accident visual data with the pre-accident visual data through the deep learning layer of the AI ​​big model to obtain the comprehensive loss rate of the area where the emergency occurred; determine the accident level in the area where the emergency occurred based on the comprehensive loss rate; and perform text generation processing on the accident visual data and the accident level through the natural language layer of the AI ​​big model to obtain an accident description file for the area where the emergency occurred.

[0069] In one implementation, the generation module 22 can be used to construct a first virtual area model of the area where the emergency occurs based on the accident visual data; and to construct a second virtual area model of the area where the emergency occurs based on the pre-accident visual data; compare and analyze the first virtual area model with the second virtual area model to obtain the scale loss rate, structural loss rate and functional loss rate of each type of accident objects in the area where the emergency occurs; and determine the comprehensive loss rate of the area where the emergency occurs based on the scale loss rate, structural loss rate and functional loss rate of each type of accident objects.

[0070] In one implementation, the generation module 22 can be used to construct a comprehensive loss matrix for the area where the emergency occurs based on the scale loss rate, structural loss rate, functional loss rate of each type of accident object, as well as the scale weight corresponding to the scale loss rate, the structural weight corresponding to the structural loss rate, and the functional weight corresponding to the functional loss rate; and superimpose the elements of the comprehensive loss matrix to obtain the comprehensive loss rate for the area where the emergency occurs.

[0071] In one implementation, the preset alternative emergency command decision includes emergency event information and handling actions corresponding to the emergency event information; the matching module 23 can be used to extract key words according to the weights of each word in the accident description file; match the key words with the emergency event information to determine the target emergency event information that matches the key words; and determine the target handling actions corresponding to the accident description file based on the target emergency event information.

[0072] In one implementation, the above-mentioned system may also include an update module, which can be used to periodically obtain visual data after the emergency command decision is executed in the area where the emergency occurs; analyze the visual data to determine the feasibility assessment result of the emergency command decision; and update the preset alternative emergency command decision based on the feasibility assessment result.

[0073] Figure 3 FIG. 2 shows a schematic diagram of the structure of another emergency command decision-making system based on an AI big model provided in an embodiment of the present application. Figure 3 As shown, the system 300 includes: a basic information preparation module 31, an image acquisition module 32, an image processing module 33, an accident assessment module 34, a text generation module 35, a solution recommendation module 36 and a simulation improvement module 37.

[0074] The basic information preparation module 31 is used to obtain basic information such as time, structural criticality, functional criticality, etc. of the disaster-affected objects in the area of ​​concern and pre-disaster images or videos based on models such as CIM / BIM. Accident and disaster information and emergency command decisions in historical accident and disaster-affected areas are obtained to build alternative emergency command decisions.

[0075] The image acquisition module 32 includes an authority feedback unit, a communication detection unit and an image acquisition unit. The image acquisition module 32 obtains the control authority of the monitoring device in the area where the accident or disaster occurs through the authority feedback unit using the security authentication algorithm and the access control algorithm. Then, the communication detection unit uses the TCP / UDP port detection algorithm to determine the communication status of the monitoring device according to the TCP / UDP port opening status of the monitoring device. Finally, the image acquisition unit calls the monitoring device in the area where the accident or disaster occurs to collect accident or disaster images or videos at regular intervals.

[0076] The image analysis module 33 is used to detect and classify buildings in accident disaster images or videos through the AI ​​big model, and compare and analyze them with images or videos of the same area before the disaster, and output the scale loss rate, structural loss rate and functional loss rate of each disaster-stricken object (such as a building).

[0077] The accident assessment module 34 is used to determine the accident level of the area where the accident or disaster occurs based on the scale loss rate, structure loss rate and function loss rate of each disaster-affected object (such as a building).

[0078] The text generation module 35 includes a scanning and sorting unit and a text description unit. The scanning and sorting unit uses the optical fluid algorithm to extract the current status of each building in the accident disaster image or video, and sorts the damage degree information of each building according to the accident disaster level determination result. The text description unit calls the Natural Language Toolkit (NLTK) library to generate an accident disaster description file according to the completed sorting of the damage degree information of each building.

[0079] The solution recommendation module 36 is used to match the target emergency command decision from the preset alternative emergency command decisions according to the accident and disaster description file.

[0080] The simulation improvement module 37 is used to conduct real-time analysis of the accident and disaster occurrence area based on the accident and disaster images or videos obtained at regular intervals, and to improve the alternative emergency command decisions.

[0081] In this embodiment, the simulation improvement module includes a progress analysis unit, an adjustment and improvement unit, and a decision simulation unit. Among them, the progress analysis unit uses an evaluation and judgment algorithm to perform a feasibility analysis of alternative emergency command decisions based on accident and disaster images or videos collected in different time periods. The adjustment and improvement unit improves and adjusts the implementation process of each stage of the alternative emergency command decision based on the feasibility analysis results. The decision simulation unit uses a virtual twin model to simulate the implementation of the alternative emergency command decision that has completed the adjustment and improvement.

[0082] Figure 4 A flowchart of another emergency command decision-making method based on an AI big model provided in an embodiment of the present application is shown, see Figure 4 , the method may include the following steps.

[0083] Step 401, based on the CIM / BIM model, obtain pre-disaster images or videos of the target area; and obtain historical emergency information and disposal actions in the target area to establish alternative emergency command decisions.

[0084] The target area includes multiple areas of current concern. The form of alternative emergency command decision is condition-action, that is, the emergency information corresponds to the disposal action.

[0085] Step 402, obtaining accident or disaster images or videos of the area where the accident or disaster occurred.

[0086] Among them, the control authority of the monitoring equipment in the area where the accident or disaster occurs is obtained through the image acquisition module, and the communication port of the monitoring equipment is judged at the same time, and the monitoring equipment in the area where the accident or disaster occurs is called to regularly collect accident or disaster images or videos; the collected accident or disaster images or videos are pre-processed, for example, the collected accident or disaster images or videos are converted, cleaned and classified through the image processing module, and a virtual twin model of the area where the accident or disaster occurs is constructed to reflect the real-time situation of the area where the accident or disaster occurs.

[0087] Step 403, based on the AI ​​big model, compare and analyze the pre-disaster images or videos and the accident disaster images or videos, evaluate the accident disaster according to the comparison and analysis results, and determine the accident disaster level.

[0088] Step 404, generating an accident disaster description file according to the accident disaster level and the accident disaster image or video.

[0089] Step 405: Match the target emergency command decision from preset candidate emergency command decisions according to the accident and disaster description file.

[0090] Step 406, performing real-time analysis on the area where the accident or disaster occurred based on the accident or disaster images or videos acquired at regular intervals, and updating the alternative emergency command decision based on the analysis results.

[0091] In the embodiment of the present application, the accident disaster level is determined by obtaining accident disaster images or videos of the accident disaster area and comparing and analyzing them with the images or videos before the disaster. Then, an accident disaster description file is generated according to the accident disaster level and the accident disaster images or videos, and the target emergency command decision is matched from the preset alternative emergency command decisions according to the accident disaster description file, so that a more accurate emergency command decision can be obtained.

[0092] In addition, the emergency command decision-making system based on the AI ​​big model provided in this application can be applied to the following scenarios.

[0093] Pre-monitoring and early warning: The system uses an integrated sensor network and remote sensing technology to conduct real-time monitoring of potential risk points in the city, such as early warning of natural disasters such as earthquakes, floods, and typhoons.

[0094] Risk assessment: Using remote sensing technology, the system can assess areas in the city that may be affected by disasters, predict the losses that may be caused by disasters, and provide data support for emergency preparedness.

[0095] Emergency response to incidents: When a disaster occurs, the system can quickly generate emergency command decisions and mobilize rescue resources such as firefighting, medical care, and transportation to ensure a quick and effective response.

[0096] Command and rescue during the event: Through remote sensing technology, the emergency command center can understand the situation at the disaster site in real time, and guide the rescue team to carry out precise rescue through the generated emergency command decisions to avoid secondary disasters.

[0097] Resource allocation: The system optimizes the allocation of rescue resources based on the analysis results of images obtained by remote sensing technology to ensure that resources can be used effectively.

[0098] Post-disaster assessment: Using remote sensing technology, the system can conduct a detailed assessment of the damage caused by the disaster and provide decision support for post-disaster reconstruction.

[0099] Decision support: During the entire disaster management process, the images acquired by the system and the analysis results of data processing and analysis can help decision makers make more scientific and reasonable decision plans.

[0100] Figure 5 A schematic diagram of a decision-making method for multiple decision nodes provided in an embodiment of the present application is shown. Figure 5 ,The above emergency command decision generation method based on AI big model can be applied to any ,decision node, thus forming a complete decision scheme.

[0101] Based on the same technical concept, the embodiment of the present application also provides an electronic device, which is used to execute the above-mentioned emergency command decision-making generation method based on the AI ​​big model. Figure 6 A schematic diagram of the structure of an electronic device for implementing each embodiment of the present application. Electronic devices may have relatively large differences due to different configurations or performances, and may include a processor (processor) 601, a communication interface (Communications Interface) 602, a memory (memory) 603 and a communication bus 604, wherein the processor 601, the communication interface 602, and the memory 603 communicate with each other through the communication bus 604. The processor 601 can call a computer program stored in the memory 603 and can run on the processor 601 to execute the various steps of the above-mentioned emergency command decision-making generation method embodiment based on the AI ​​large model, and can achieve the same technical effect. To avoid repetition, it will not be repeated here.

[0102] The above electronic device structure does not constitute a limitation on the electronic device. The electronic device may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently. For example, the input unit may include a graphics processing unit (GPU) and a microphone, and the display unit may be configured with a display panel in the form of a liquid crystal display, an organic light-emitting diode, etc. The user input unit includes a touch panel and at least one of other input devices. The touch panel is also called a touch screen. Other input devices may include, but are not limited to, a physical keyboard, function keys (such as volume control buttons, switch buttons, etc.), a trackball, a mouse, and a joystick, which will not be repeated here.

[0103] The memory can be used to store software programs and various data. The memory may mainly include a first storage area for storing programs or instructions and a second storage area for storing data, wherein the first storage area may store an operating system, an application program or instructions required for at least one function (such as a sound playback function, an image playback function, etc.), etc. In addition, the memory may include a volatile memory or a non-volatile memory, or the memory may include both volatile and non-volatile memories. Among them, the non-volatile memory may be a read-only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), or a flash memory. Volatile memory can be random access memory (RAM), static random access memory (SRAM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), double data rate synchronous dynamic random access memory (DDRSDRAM), enhanced synchronous dynamic random access memory (ESDRAM), synchronous link dynamic random access memory (SLDRAM) and direct memory bus random access memory (DRRAM).

[0104] The processor may include one or more processing units; optionally, the processor integrates an application processor and a modem processor, wherein the application processor mainly processes operations related to the operating system, user interface, and application programs, and the modem processor mainly processes wireless communication signals, such as a baseband processor. It is understandable that the modem processor may not be integrated into the processor.

[0105] An embodiment of the present application also provides a readable storage medium, on which a program or instruction is stored. When the program or instruction is executed by a processor, each process of the above-mentioned emergency command decision-making generation method embodiment based on the AI ​​big model is implemented, and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.

[0106] The processor is a processor in the electronic device described in the above embodiment. The readable storage medium includes a computer readable storage medium, such as a computer read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0107] An embodiment of the present application further provides a chip, which includes a processor and a communication interface, wherein the communication interface is coupled to the processor, and the processor is used to run programs or instructions to implement the various processes of the above-mentioned emergency command decision-making generation method embodiment based on the AI ​​big model, and can achieve the same technical effect. To avoid repetition, it will not be repeated here.

[0108] It should be understood that the chip mentioned in the embodiments of the present application can also be called a system-level chip, a system chip, a chip system or a system-on-chip chip, etc.

[0109] An embodiment of the present application also provides a computer program product, which includes a computer program stored on a non-transitory computer-readable storage medium, and the computer program includes a program or instructions. When the program or instructions are executed, the various processes of the above-mentioned emergency command decision-making generation method embodiment based on the AI ​​big model are implemented, and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.

[0110] It should be noted that, in this article, the terms "comprise", "include" or any other variant thereof are intended to cover non-exclusive inclusion, so that the process, method, article or device including a series of elements includes not only those elements, but also includes other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise one..." do not exclude the presence of other identical elements in the process, method, article or device including the element. In addition, it should be pointed out that the scope of the methods and devices in the embodiments of the present application is not limited to performing functions in the order shown or discussed, and may also include performing functions in a substantially simultaneous manner or in reverse order according to the functions involved, for example, the described method may be performed in an order different from that described, and various steps may also be added, omitted, or combined. In addition, the features described with reference to certain examples may be combined in other examples.

[0111] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above-mentioned embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method. Based on such an understanding, the technical solution of the present application, or the part that contributes to the prior art, can be embodied in the form of a software product, which is stored in a storage medium (such as ROM / RAM, a magnetic disk, or an optical disk), and includes a number of instructions for a terminal (which can be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.) to execute the methods described in each embodiment of the present application.

[0112] The embodiments of the present application are described above in conjunction with the accompanying drawings, but the present application is not limited to the above-mentioned specific implementation methods. The above-mentioned specific implementation methods are merely illustrative and not restrictive. Under the guidance of the present application, ordinary technicians in this field can also make many forms without departing from the purpose of the present application and the scope of protection of the claims, all of which are within the protection of the present application.

Claims

1. A method for generating emergency command decisions based on an AI big model, characterized in that: include: Obtain accident visual data and pre-accident visual data of the area where the emergency occurred; Processing the accident visual data and the pre-accident visual data through an AI big model to generate an accident description file of the area where the emergency occurred; According to the accident description file, an emergency command decision corresponding to the accident description file is matched from preset alternative emergency command decisions.

2. The method according to claim 1, characterized in that The step of processing the accident visual data and the pre-accident visual data through an AI big model to generate an accident description file of the area where the emergency occurred includes: Compare and analyze the accident visual data with the pre-accident visual data through the deep learning layer of the AI ​​big model to obtain the comprehensive loss rate of the area where the emergency occurred; Determine the accident level of the area where the emergency occurs based on the comprehensive loss rate; The accident visual data and the accident level are processed by the natural language layer of the AI ​​big model to generate text, so as to obtain an accident description file of the area where the emergency occurred.

3. The method according to claim 2, characterized in that The method of comparing and analyzing the accident visual data with the pre-accident visual data through the deep learning layer of the AI ​​big model to obtain the comprehensive loss rate of the area where the emergency occurred includes: Constructing a first virtual area model of the area where the emergency occurred based on the accident visual data; and constructing a second virtual area model of the area where the emergency occurred based on the pre-accident visual data; Comparing and analyzing the first virtual area model and the second virtual area model, respectively obtaining the scale loss rate, structural loss rate and functional loss rate of each type of accident object in the emergency occurrence area; The comprehensive loss rate of the area where the emergency occurs is determined based on the scale loss rate, structural loss rate and functional loss rate of each type of accident object.

4. The method according to claim 3, characterized in that Determining the comprehensive loss rate of the area where the emergency occurs based on the scale loss rate, structure loss rate and function loss rate of each type of accident object includes: Construct a comprehensive loss matrix of the area where the emergency occurs according to the scale loss rate, structure loss rate, function loss rate of each type of accident object, the scale weight corresponding to the scale loss rate, the structure weight corresponding to the structure loss rate, and the function weight corresponding to the function loss rate; The elements of the comprehensive loss matrix are superimposed to obtain the comprehensive loss rate of the area where the emergency occurs.

5. The method according to claim 1, characterized in that The preset alternative emergency command decision includes emergency event information and a handling action corresponding to the emergency event information; the matching of the emergency command decision corresponding to the accident description file from the preset alternative emergency command decisions according to the accident description file includes: Extracting key words according to the weights of each word in the accident description file; Matching the key words with the emergency event information to determine target emergency event information matching the key words; According to the target emergency event information, a target handling action corresponding to the accident description file is determined.

6. The method according to claim 1, characterized in that After matching the emergency command decision corresponding to the accident description file from preset alternative emergency command decisions according to the accident description file, the method further includes: Regularly acquiring visual data after the emergency command decision is executed in the area where the emergency occurs; Analyzing the visual data to determine a feasibility assessment result of the emergency command decision; According to the feasibility assessment result, the preset alternative emergency command decision is updated.

7. An emergency command decision-making system based on AI big model, characterized in that: include: An acquisition module, used to acquire accident visual data of the area where the emergency occurred, and pre-accident visual data of the area where the emergency occurred; wherein the visual data includes images or videos; A generation module, used to process the accident visual data and the pre-accident visual data through an AI big model to generate an accident description file of the area where the emergency occurred; A matching module is used to match the emergency command decision corresponding to the accident description file from preset alternative emergency command decisions based on the accident description file.

8. An electronic device, characterized in that: It includes a processor, a memory, and a program or instruction stored in the memory and executable on the processor. When the program or instruction is executed by the processor, the steps of the emergency command decision-making generation method based on the AI ​​large model are implemented as described in any one of claims 1 to 6.

9. A readable storage medium, characterized in that: The readable storage medium stores programs or instructions, and when the programs or instructions are executed by the processor, the steps of the emergency command decision-making generation method based on the AI ​​large model as described in any one of claims 1 to 6 are implemented.

10. A computer program product, characterized in that The computer program product includes a computer program stored on a non-transitory computer-readable storage medium, and the computer program includes a program or instructions, which, when executed, implement the steps of the emergency command decision-making method based on the AI ​​big model as described in any one of claims 1 to 6.

Images