Intelligent auxiliary method and device for command system

Abstract

The invention provides an intelligent auxiliary method and device for a command system, and belongs to the technical field of emergency command. According to the embodiment of the invention, the method comprises the steps: carrying out the emergency process sampling of a command system, obtaining a plurality of emergency processes, obtaining the process evaluation values corresponding to the plurality of emergency processes based on the scheme evaluation values of a plurality of emergency schemes corresponding to each emergency process determined by the command system, and carrying out the calculation of the process evaluation values. The method comprises the following steps: acquiring a plurality of emergency processes, screening out a plurality of optimal emergency processes from the plurality of emergency processes, and determining a target emergency process from the plurality of optimal emergency processes based on the target emergency demand information after determining the target emergency demand information of a target emergency event; or inputting the target emergency demand information into a process generation model obtained based on training of a plurality of optimal emergency processes, and outputting to obtain a target emergency process. Therefore, the end-to-end intelligent auxiliary method is realized, and the actual application effect of the target emergency process is effectively improved while the emergency response speed of the emergency command system is improved.

Description

Technical Field

[0001] The present application relates to the field of emergency command technology, and in particular to an intelligent auxiliary method and device for a command system. Background Art

[0002] A mature emergency command system can quickly respond to emergencies, reduce the losses caused by emergencies, achieve cross-departmental and cross-regional coordination, and coordinate and deploy various emergency resources, including personnel, materials, and transportation. The capability level of the emergency command system is crucial and directly related to whether the safety of people's lives and property can be protected to the greatest extent.

[0003] The command processes in different emergency demand scenarios vary greatly. When facing different emergency dispatch needs, we must first formulate macro processes as needed, that is, which specific command nodes to use for emergency command, and then formulate micro processes, that is, which functional services each command node needs to be configured with.

[0004] In related technologies, the corresponding emergency process is usually formulated by template matching, that is, considering various possible emergency needs and situations, various emergency plan templates are prepared in advance. When encountering actual emergency tasks, experienced emergency personnel need to select the corresponding emergency template for emergency handling. However, emergency plan templates are only a limited number of fixed templates, which usually cannot fully cover all emergency demand scenarios. At the same time, they are overly dependent on the command experience of the commander, which leads to insufficient emergency response speed of the emergency command system and difficulty in ensuring the actual application effect of the emergency process. Summary of the invention

[0005] The present application provides an intelligent auxiliary method and device for a command system to solve the problem in the related art that the emergency command system has an insufficient emergency response speed and is difficult to ensure the actual application effect of the emergency process.

[0006] In order to solve the above problems, this application adopts the following technical solutions: In a first aspect, an embodiment of the present application provides an intelligent assistance method for a command system, the method comprising: Sampling emergency processes of the command system to obtain multiple emergency processes; wherein the emergency processes include selected command nodes at each level of the command system and selected emergency parameters of the selected command nodes; Sending the plurality of emergency processes to the command system, so that the command system generates a plurality of emergency plans for any emergency process based on the plurality of process branches of the emergency process, and determines the plan evaluation values ​​corresponding to the plurality of emergency plans; wherein different emergency plans are generated by selecting the command node at the end of different process branches; Receiving the solution evaluation values ​​of the multiple emergency solutions sent by the command system for any emergency process, and determining the process evaluation value of the emergency process based on the solution evaluation values ​​of the multiple emergency solutions, so as to obtain the process evaluation values ​​corresponding to each of the multiple emergency processes; Based on the process evaluation values ​​corresponding to the multiple emergency processes and the solution evaluation values ​​of the multiple emergency solutions, the multiple emergency processes are screened to obtain multiple preferred emergency processes; Determine the target emergency demand information of the target emergency event; Based on the target emergency demand information, determine the target emergency process among the multiple preferred emergency processes; or input the target emergency demand information into a process generation model, and output the target emergency process; wherein the process generation model is trained based on multiple preferred emergency processes.

[0007] In an embodiment of the present application, the step of sampling the emergency process of the command system to obtain multiple emergency processes includes: Constructing a knowledge graph of the command system; the knowledge graph includes a node graph and a function graph of the command system; the node graph represents command nodes at each level of the command system and connection relationships between command nodes at each level, and the function graph represents at least one business software of the command system, at least one micro-application under the business software, at least one micro-service under the micro-application, at least one function algorithm cluster under the micro-service, and at least one algorithm under the function algorithm cluster; Sampling the node graph in the knowledge graph to obtain selected command nodes at each level, and sampling the function graph in the knowledge graph to obtain selected emergency parameters of each selected command node; Based on the selected command nodes at each level and the selected emergency parameters of each selected command node, a plurality of emergency processes are obtained.

[0008] In an embodiment of the present application, the method further includes: Based on the updated information of the command system, the node graph and / or the function graph is updated to obtain a reconstructed knowledge graph.

[0009] In one embodiment of the present application, the step of determining the process evaluation value of the emergency process based on the solution evaluation values ​​of the multiple emergency solutions includes: For any emergency plan, determining a weight parameter of the emergency plan based on the importance of the process branch that generates the emergency plan; Based on the solution evaluation values ​​and weight parameters corresponding to each of the multiple emergency solutions, the process evaluation value of the emergency process is determined.

[0010] In an embodiment of the present application, the method further includes: Inputting a plurality of the emergency plans into a pre-built digital simulation evaluation system to obtain simulation results output by the digital simulation evaluation system; Based on the simulation result, a process evaluation value of the emergency process is determined.

[0011] In one embodiment of the present application, the step of screening the multiple emergency processes based on the process evaluation values ​​corresponding to the multiple emergency processes and the solution evaluation values ​​of the multiple emergency solutions to obtain multiple preferred emergency processes includes: For any emergency process, when the process evaluation value of the emergency process is greater than a first threshold, or the solution evaluation values ​​of multiple emergency solutions of the emergency process are all greater than a second threshold, the emergency process is determined as the preferred emergency process.

[0012] In an embodiment of the present application, the step of determining target emergency demand information of a target emergency event includes: Obtaining emergency reporting information of the target emergency event; The emergency reporting information and the preset prompt engineering text are input into the large language model, and the target emergency demand information in the preset format is output; the prompt engineering text includes emergency auxiliary information, emergency data text and format constraint text for indicating the preset format, and the emergency data text includes an emergency plan library, a standard emergency manual and a historical emergency plan.

[0013] In one embodiment of the present application, the plurality of preferred emergency processes include respective corresponding preset emergency demand information; The step of determining a target emergency process among the plurality of preferred emergency processes based on the target emergency demand information includes: Determining characteristic distances between a plurality of preset emergency demand information and the target emergency demand information; The preferred emergency process corresponding to the preset emergency demand information with the closest characteristic distance is determined as the target emergency process.

[0014] In one embodiment of the present application, the process generation model is trained by the following steps: For any preferred emergency process, a sample data group corresponding to the preferred emergency process is generated; the sample data group includes n consecutive sample input data and process label data corresponding to each sample input data, where n is a positive integer greater than 1; Determine the first sample input data as the target data, input the target data into the initial recurrent neural network model, output a predicted process, and calculate the loss function value based on the predicted process and the process label data corresponding to the target data; Determine the next sample input data of the first sample input data as the target data, perform the steps of inputting the target data into the initial recurrent neural network model, outputting a prediction process, and calculating the loss function value based on the prediction process and the process label data corresponding to the target data, until the nth loss function value corresponding to the nth sample input data is calculated; Based on the n loss function values ​​corresponding to a preset number of the preferred emergency processes, the target loss function value is determined, and the network parameters of the initial recurrent neural network model are updated based on the target loss function value until the initial recurrent neural network model meets the training cutoff condition, thereby obtaining the process generation model.

[0015] In a second aspect, based on the same inventive concept, an embodiment of the present application provides an intelligent auxiliary device for a command system, the device comprising: A process sampling module is used to sample the emergency process of the command system to obtain multiple emergency processes; wherein the emergency process includes selected command nodes at each level of the command system and selected emergency parameters of the selected command nodes; A process sending module, used for sending the plurality of emergency processes to the command system, so that the command system generates a plurality of emergency plans for any emergency process based on the plurality of process branches of the emergency process, and determines the plan evaluation values ​​corresponding to the plurality of emergency plans; wherein different emergency plans are generated by selecting a command node at the end of different process branches; A process evaluation module, used for receiving the solution evaluation values ​​of the multiple emergency solutions sent by the command system for any emergency process, and determining the process evaluation value of the emergency process based on the solution evaluation values ​​of the multiple emergency solutions, so as to obtain the process evaluation values ​​corresponding to each of the multiple emergency processes; A process screening module, used to screen the multiple emergency processes based on the process evaluation values ​​corresponding to the multiple emergency processes and the solution evaluation values ​​of the multiple emergency solutions, to obtain multiple preferred emergency processes; An information determination module, used to determine target emergency demand information of a target emergency event; A process determination module is used to determine a target emergency process among multiple preferred emergency processes based on the target emergency demand information; or, input the target emergency demand information into a process generation model and output a target emergency process; wherein the process generation model is obtained by training based on multiple preferred emergency processes.

[0016] In one embodiment of the present application, the process sampling module includes: A knowledge graph construction submodule is used to construct a knowledge graph of the command system; the knowledge graph includes a node graph and a function graph of the command system; the node graph represents the command nodes of each level of the command system and the connection relationship between the command nodes of each level, and the function graph represents at least one business software of the command system, at least one micro application under the business software, at least one micro service under the micro application, at least one function algorithm cluster under the micro service, and at least one algorithm under the function algorithm cluster; A graph sampling submodule, used to sample the node graph in the knowledge graph to obtain the selected command nodes at each level, and to sample the function graph in the knowledge graph to obtain the selected emergency parameters of each selected command node; The emergency process determination submodule is used to obtain a plurality of emergency processes based on the selected command nodes at each level and the selected emergency parameters of each selected command node.

[0017] In one embodiment of the present application, the intelligent auxiliary device of the command system further includes: The knowledge graph updating module is used to update the node graph and / or the function graph based on the update information of the command system to obtain a reconstructed knowledge graph.

[0018] In one embodiment of the present application, the process evaluation module includes: A weight parameter determination submodule, for determining a weight parameter of any emergency plan based on the importance of a process branch for generating the emergency plan; The process evaluation value determination submodule is used to determine the process evaluation value of the emergency process based on the solution evaluation values ​​and weight parameters corresponding to each of the multiple emergency solutions.

[0019] In one embodiment of the present application, the intelligent auxiliary device of the command system further includes: A process simulation module, used for inputting a plurality of the emergency plans into a pre-built digital simulation evaluation system to obtain simulation results output by the digital simulation evaluation system; A simulation evaluation module is used to determine a process evaluation value of the emergency process based on the simulation result.

[0020] In one embodiment of the present application, the process screening module includes: The threshold comparison submodule is used to determine any emergency process as the preferred emergency process when the process evaluation value of the emergency process is greater than a first threshold, or the solution evaluation values ​​of multiple emergency solutions of the emergency process are all greater than a second threshold.

[0021] In one embodiment of the present application, the information determination module includes: An information acquisition submodule is used to acquire emergency reporting information of the target emergency event; The model identification submodule is used to input the emergency reporting information and the preset prompt engineering text into the large language model, and output the target emergency demand information in a preset format; the prompt engineering text includes emergency auxiliary information, emergency data text and format constraint text for indicating the preset format, and the emergency data text includes an emergency plan library, a standard emergency manual and a historical emergency plan.

[0022] In an embodiment of the present application, the plurality of preferred emergency processes include respective corresponding preset emergency demand information; the process determination module includes: A distance determination submodule, used to determine characteristic distances between a plurality of preset emergency demand information and the target emergency demand information; The process determination submodule is used to determine the preferred emergency process corresponding to the preset emergency demand information with the closest characteristic distance as the target emergency process.

[0023] In one embodiment of the present application, the intelligent auxiliary device of the command system further includes: A sample generation module, used for generating a sample data group corresponding to any preferred emergency process, wherein the sample data group includes n consecutive sample input data and process label data corresponding to each sample input data, where n is a positive integer greater than 1; A first sample input module, used to determine the first sample input data as target data, and input the target data into the initial recurrent neural network model, output a prediction process, and calculate a loss function value based on the prediction process and process label data corresponding to the target data; A second sample input module is used to determine the next sample input data of the first sample input data as the target data, perform the steps of inputting the target data into the initial recurrent neural network model, outputting a prediction process, and calculating the loss function value based on the prediction process and the process label data corresponding to the target data, until the nth loss function value corresponding to the nth sample input data is calculated; A parameter updating module is used to determine a target loss function value based on n loss function values ​​corresponding to a preset number of the preferred emergency processes, and to update the network parameters of the initial recurrent neural network model based on the target loss function value until the initial recurrent neural network model meets the training cutoff condition, thereby obtaining the process generation model.

[0024] Compared with the prior art, this application has the following advantages: The intelligent auxiliary method of a command system provided by the embodiment of the present application first samples the emergency process of the command system to obtain multiple emergency processes, and sends the multiple emergency processes to the command system, so that the command system generates multiple emergency plans for any emergency process based on multiple process branches of the emergency process, and determines the corresponding plan evaluation values ​​of the multiple emergency plans; then based on the plan evaluation values ​​of the multiple emergency plans corresponding to each emergency process, obtain the process evaluation values ​​corresponding to the multiple emergency processes, and screen multiple preferred emergency processes from the multiple emergency processes. After determining the target emergency demand information of the target emergency event, the target emergency process can be determined in multiple preferred emergency processes based on the target emergency demand information; or, the target emergency demand information is input into the process generation model obtained by training based on multiple preferred emergency processes, and the target emergency process is output. The embodiment of the present application pre-screens multiple preferred emergency processes. Once an emergency event is encountered, the appropriate target emergency process can be quickly and accurately determined by information matching or process generation model, and emergency assistance is performed on the emergency commanders of each command node. In this way, while improving the emergency response speed of the emergency command system, the actual application effect of the target emergency process can be effectively improved, thereby realizing an end-to-end intelligent auxiliary method and greatly improving the effectiveness of the command system. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0026] Figure 1 It is a step flow chart of an intelligent assistance method of a command system in one embodiment of the present application.

[0027] Figure 2 It is a training diagram of the initial recurrent neural network model in one embodiment of the present application.

[0028] Figure 3 It is a module schematic diagram of an intelligent auxiliary device of a command system in one embodiment of the present application. DETAILED DESCRIPTION

[0029] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0030] It should be noted that the command processes in different emergency demand scenarios vary greatly. When facing different emergency dispatch needs, emergency commanders must first formulate macro processes as needed to determine the macro business organization, including what kind of command mode to adopt, what specific "command nodes" there are (such as different departments, command centers distributed in different locations), each "command node"'s operating seat (such as intelligence summary seat, resource demand estimation seat, etc.) and operating tasks; in addition to specifying the macro process, each emergency "command node" and seat has many optional functional services, data services, and algorithm models.

[0031] In order to further improve the emergency command capability, researchers are actively developing intelligent empowerment technologies for command systems, and continuously adding intelligent algorithms to the core functional modules of the command system. The same function may be implemented by different algorithms or combinations of algorithms under different application conditions. Different algorithms have different inputs and constraints. Therefore, for specific emergency tasks issued by superiors, each emergency "command node" needs to further formulate more microscopic business processes based on tasks and intentions, including what software each emergency "command node" and seat is equipped with, which functional modules and algorithms are selected in each software, and the inputs, constraints, optimization goals and data required to execute this functional module or algorithm. There are many choices at each level, and the planning is complex.

[0032] In related technologies, the corresponding emergency process is usually formulated by template matching, that is, taking into account various possible emergency needs and situations, various emergency plan templates are prepared in advance. The plan template generally includes applicable conditions such as the emergency needs and resources, the macro process of the emergency command system, and the functional services and algorithms called by each seat. When encountering actual emergency tasks, experienced emergency personnel (or teams) are required to coordinate and discuss which emergency template to choose. Since the emergency plan template is only a limited number of fixed templates, it is usually unable to fully cover all emergency demand scenarios. At the same time, it is overly dependent on the command experience of the commander, which leads to insufficient emergency response speed of the emergency command system, and it is difficult to ensure the actual application effect of the emergency process.

[0033] In view of the problem that the emergency response speed of the emergency command system in the related technology is insufficient and it is difficult to ensure the actual application effect of the emergency process, this application aims to provide an intelligent auxiliary method for the command system. First, the command system samples the emergency process to obtain multiple emergency processes, and then uses the command system to generate multiple emergency plans based on multiple process branches of the emergency process for any emergency process, and obtains the process evaluation values ​​corresponding to the multiple emergency processes according to the plan evaluation values ​​corresponding to each of the multiple emergency plans, and then screens multiple preferred emergency processes from the multiple emergency processes. After determining the target emergency demand information of the target emergency event, the target emergency process can be determined based on the target emergency demand information in multiple preferred emergency processes; or, the target emergency demand information is input into the process generation model obtained by training based on multiple preferred emergency processes, and the target emergency process is output. In this way, the appropriate target emergency process can be determined quickly and accurately, and emergency assistance can be provided to the emergency commanders of each command node. While improving the emergency response speed of the emergency command system, the actual application effect of the target emergency process can be effectively improved, thereby realizing an end-to-end intelligent auxiliary method and greatly improving the effectiveness of the emergency command system.

[0034] Reference Figure 1 , shows an intelligent assistance method for a command system of the present application, which may include the following steps: S101: Sample the emergency process of the command system to obtain multiple emergency processes.

[0035] In this embodiment, the emergency process includes selected command nodes at each level of the command system and selected emergency parameters of the selected command nodes.

[0036] It should be noted that the command system usually includes multiple levels of command nodes. For example, the command system includes a first-level command node, which includes multiple second-level command nodes, each of which includes multiple third-level command nodes, and so on, until the last command node. Specifically, each level of command node includes one or more command seats, such as intelligence summary seats, resource demand estimation seats, etc., and the command seats collaborate to form a command node.

[0037] In an emergency command example, the command nodes at various levels of the command system may include: the national emergency command center calls seat a, the provincial emergency command center calls seat b and seat c, etc., the prefecture-level emergency command center calls seat d and seat e, etc., and the district and county-level emergency command center calls seat f, seat g, and seat h.

[0038] It should be noted that the emergency parameters of the command node refer to the emergency parameters used by the command node when conducting emergency command, which may specifically include the target software installed in each emergency command node and seat, the target functional modules and target algorithms selected in each target software, the target input required to execute this functional module or algorithm, target constraints, target optimization targets, and called target data.

[0039] In this embodiment, by sampling the emergency process of the command nodes at each level of the command system and the emergency parameters of the command nodes, various possible emergency processes in the command system can be sampled to achieve comprehensive sampling of the emergency process.

[0040] S102: Sending multiple emergency processes to the command system, so that the command system generates multiple emergency plans for any emergency process based on multiple process branches of the emergency process, and determines the plan evaluation values ​​corresponding to each of the multiple emergency plans.

[0041] It should be noted that different emergency plans are generated by selecting the command node at the end of different process branches.

[0042] It should be further explained that a complete emergency process usually requires the collaboration of multiple departments in different regions. Therefore, the emergency process usually includes multiple process branches, each of which meets the corresponding emergency needs. In other words, the emergency process for different emergency needs is in the form of an inverted tree structure. The command node at the end of each process branch will generate a set of corresponding emergency plans. The combination of multiple emergency plans generated by all the terminal command nodes together meets the emergency needs corresponding to the emergency process.

[0043] In this embodiment, by sending multiple emergency processes to the command system, the actual command system can be driven to generate multiple emergency plans according to multiple process branches of any emergency process, and the command node selected at the end determines the plan evaluation value of the emergency plan it generates.

[0044] In a specific implementation, each terminal command node of the command system is configured with a preset evaluation algorithm, so that the terminal selected command node can evaluate the emergency plan based on the evaluation algorithm and determine the plan evaluation value of the emergency plan. Alternatively, subjective evaluation (such as expert scoring) and objective evaluation (such as quantitative analysis through specific emergency data) can be combined to obtain subjective scores and objective scores; and the plan evaluation value of each emergency plan is determined based on the subjective scores and objective scores.

[0045] S103: Receive the solution evaluation values ​​of multiple emergency solutions sent by the command system for any emergency process, and determine the process evaluation value of the emergency process based on the solution evaluation values ​​of the multiple emergency solutions to obtain the process evaluation values ​​corresponding to each of the multiple emergency processes.

[0046] It should be noted that, if the solution evaluation value of the emergency solution is higher, it means that the efficiency of the process branch used to generate the emergency solution is higher.

[0047] In this implementation, by comprehensively evaluating the solution evaluation values ​​of multiple emergency solutions of the emergency process, the process evaluation value of the emergency process can be calculated, thereby achieving a comprehensive and accurate evaluation of the emergency process.

[0048] S104: Based on the process evaluation values ​​corresponding to the multiple emergency processes and the solution evaluation values ​​of the multiple emergency solutions, the multiple emergency processes are screened to obtain multiple preferred emergency processes.

[0049] In this embodiment, due to the randomness of emergency process sampling, a certain emergency process obtained by sampling may be a relatively high-efficiency process or a relatively low-efficiency process. Therefore, it is necessary to screen multiple emergency processes to obtain multiple preferred emergency processes with better emergency effects.

[0050] In a specific implementation, for any emergency process, if the process evaluation value of the emergency process is greater than a first threshold, or the solution evaluation values ​​of multiple emergency solutions of the emergency process are greater than a second threshold, the emergency process can be determined as a preferred emergency process.

[0051] In this embodiment, by comprehensively considering the process evaluation value of the emergency process and the solution evaluation values ​​of multiple emergency solutions of the emergency process, it is possible to effectively screen the preferred emergency process, thereby retaining the emergency process with better emergency effect as comprehensively as possible.

[0052] S105: Determine target emergency demand information of the target emergency event.

[0053] It should be noted that, when a target emergency event occurs, the local emergency department will upload the emergency reporting information of the target emergency event. The emergency reporting information is used to reflect the event information of the target emergency event.

[0054] In this embodiment, when facing different emergency events, different emergency departments may upload emergency reporting information of different forms and contents. Therefore, in order to quickly extract and analyze key information, after receiving the emergency reporting information of the target emergency event, the emergency reporting information can be analyzed and processed to obtain the target emergency demand information.

[0055] S106: Based on the target emergency demand information, determine the target emergency process from multiple preferred emergency processes; or, input the target emergency demand information into a process generation model, and output the target emergency process.

[0056] It should be noted that the process generation model is obtained by training based on multiple optimal emergency processes.

[0057] In a specific implementation, for any preferred emergency process, the preset emergency demand information corresponding to the preferred emergency process can be input into the initial neural network model, and a predicted process can be output. Based on the predicted process and the preferred emergency process, the loss function value is calculated, and the network parameters of the initial neural network model are updated according to the loss function value until the initial neural network model meets the training cutoff condition to obtain a process generation model.

[0058] In this embodiment, after determining the target emergency demand information, the target emergency process can be matched in multiple preferred emergency processes by information matching; the target emergency demand information can also be input into the process generation model, and the target emergency process can be obtained by outputting the process generation model. In this way, in the actual application stage, an end-to-end intelligent assistance method can be implemented to quickly and accurately determine the appropriate target emergency process, and provide emergency assistance to the emergency commanders of each command node. In an emergency, the command system can be automatically operated without the need for process personnel to issue process instructions to emergency commanders at all levels, thereby effectively improving the actual application effect of the target emergency process and the effectiveness of the emergency command system while improving the emergency response speed of the emergency command system.

[0059] In a feasible implementation, S101 may specifically include the following sub-steps: S101-1: Construct a knowledge graph of the command system.

[0060] In this embodiment, the knowledge graph includes a node graph and a function graph of the command system. The node graph represents the command nodes at each level of the command system and the connection relationship between the command nodes at each level; the function graph represents at least one business software of the command system, at least one micro application under the business software, at least one micro service under the micro application, at least one function algorithm cluster under the micro service, and at least one algorithm under the function algorithm cluster.

[0061] It should be noted that due to factors such as regional restrictions and functional division, the command nodes in the command system may not be able to establish connections with all other command nodes. Therefore, by constructing a node map, the connection relationship between command nodes at all levels can be clarified. In this way, when sampling the emergency process of the command system, it is effectively avoided to combine two command nodes that cannot actually cooperate and affect the accuracy of the process evaluation.

[0062] It should be further explained that the functional map includes business software, micro-applications, micro-services, functional algorithm clusters and algorithms in order of decreasing granularity. Among them, several algorithms are linked to each other to realize functional algorithm clusters, several functional algorithm clusters are linked to realize micro-services, several micro-services are linked to realize micro-applications, several micro-applications are linked to realize business software, and several business software are linked to realize the belonging and function of business processes. Specifically, the functional map also includes the functions, inputs, outputs, constraints, optimization targets and their corresponding numerical types and ranges of the above elements, as well as other elements related to them. The functional maps together constitute the various functional services that the entire command system can provide.

[0063] In this embodiment, the knowledge graph is constructed based on the command node architecture of the command system and the corresponding functional services. Therefore, when the command system changes, that is, the command node architecture of the command system or the functions, inputs, outputs, constraints, optimization goals and their corresponding numerical types and ranges of the command system's business software, micro-applications, micro-services, functional algorithm clusters, algorithms, etc. change, the node graph and / or functional graph will be updated based on the update information of the command system to obtain a reconstructed knowledge graph.

[0064] In the specific implementation, in order to improve the update efficiency, the knowledge graph can be constructed based on the underlying shared form, shared data dictionary and other strong association methods. In this way, when the target part in the knowledge graph is updated based on the update information of the command system, the associated part of the target part can be adaptively updated, so that the commander only needs to input the update information of the command system to realize the automatic update of the entire knowledge graph.

[0065] S101-2: Sampling the node graph in the knowledge graph to obtain the selected command nodes at each level, and sampling the function graph in the knowledge graph to obtain the selected emergency parameters of each selected command node.

[0066] In this embodiment, by performing a top-down node sampling on the node map, a group of interrelated selected command nodes can be sampled, and the group of selected command nodes includes selected command nodes at each level; by sampling the functional map, corresponding selected emergency parameters can be assigned to each selected command node, and then the corresponding selected command node can be used to implement the corresponding functional service.

[0067] In the specific implementation, according to the constructed knowledge graph, the total number of emergency processes is first estimated by multiplication and addition. For example, the selected emergency parameters in the function graph are sampled according to the preset sampling interval. Among them, continuous values, such as [0-10), can be sampled at fixed equal intervals of 1, and the samples are 0, 1, 2, 3, 4, 5, 6, 7, 8, 9.

[0068] It should be noted that if the total number of emergency processes after the budget is acceptable, no modification will be made, the process sampling will be completed, and the command system will be driven to execute the corresponding process; if the total number of emergency processes after the budget is too small, the sampling interval can be reduced to increase the number of consecutive numerical sampling; if the total number of emergency processes after the budget is too large, the sampling interval can be increased to reduce the number of consecutive numerical sampling.

[0069] It should be further explained that the data and ranges available for some sampling links are discrete. For example, if a functional algorithm cluster includes algorithm 1, algorithm 2, and algorithm 3, the number of samples for the algorithm is 3, corresponding to the above three algorithms; for example, if the optimization goals of an algorithm include the fastest emergency response time and the highest emergency consumption ratio, the number of samples for the optimization goal is 2, corresponding to the above two optimization goals. When estimating the total number of emergency processes, the sampling of the above discrete values ​​is usually not reduced, but the continuous values ​​are adjusted first to meet the quantity requirements of the emergency processes.

[0070] S101 - 3: Based on the selected command nodes at each level and the selected emergency parameters of each selected command node, multiple emergency processes are obtained.

[0071] In this embodiment, by performing sampling according to the above sampling strategy, a plurality of emergency processes with sufficient quantity and wide adaptability can be obtained, thereby achieving comprehensive sampling of emergency processes.

[0072] In a feasible implementation, S103 may specifically include the following sub-steps: S103-1: For any emergency plan, determine a weight parameter of the emergency plan based on the importance of the process branch for generating the emergency plan.

[0073] In this embodiment, considering that different process branches of the same emergency process often serve different functional services, different process branches often have different importance. For example, multiple process branches may include a main process branch and multiple auxiliary process branches, then the importance of the main process branch is higher than the importance of multiple auxiliary process branches, and the importance of multiple auxiliary process branches may also vary to a certain extent. Therefore, in order to achieve an accurate evaluation of the entire emergency process, the importance of each process branch will be comprehensively considered to determine the weight parameters of each emergency plan.

[0074] In a specific implementation, the importance of a process branch can be determined based on a preset importance level or importance score of the process branch; the higher the importance of a process branch, the greater the weight parameter of the emergency plan generated by the process branch.

[0075] S103-2: Determine a process evaluation value of the emergency process based on the solution evaluation values ​​and weight parameters corresponding to each of the multiple emergency solutions.

[0076] In a specific implementation, based on the scheme evaluation values ​​and weight parameters corresponding to each of the multiple emergency schemes, the process evaluation value of the emergency process can be calculated by weighted summation.

[0077] In this implementation, by comprehensively considering the importance of each process branch of the emergency process, a comprehensive and accurate evaluation of the emergency process can be achieved simply and efficiently.

[0078] In a feasible implementation manner, the intelligent assistance method of the command system may further include the following steps: S201: Input a plurality of emergency plans into a pre-built digital simulation evaluation system to obtain simulation results output by the digital simulation evaluation system.

[0079] In this embodiment, by pre-building a digital simulation evaluation system, after obtaining multiple emergency plans for the emergency process, the digital simulation evaluation system is used to perform simulation evaluation to obtain simulation results for the emergency process.

[0080] It should be noted that after the digital simulation evaluation system simulates multiple emergency plans according to the emergency process, it can be evaluated from multiple dimensions, for example, it can be evaluated from multiple aspects such as effectiveness, timeliness, resource utilization, safety, etc., and obtain simulation results in multiple dimensions.

[0081] S202: Determine a process evaluation value of the emergency process based on the simulation result.

[0082] In this embodiment, based on the simulation results of multiple dimensions, the process evaluation value of the emergency process can be determined by weighted summation, wherein the simulation results of different dimensions can correspond to different weight values ​​according to the emergency needs.

[0083] It should be noted that, compared with the evaluation methods of S103-1~S103-2, the process evaluation value calculated by system simulation is more accurate, but it is necessary to establish a digital simulation evaluation system, which is more costly and time-consuming. Therefore, according to the actual situation, the appropriate evaluation method can be selected from the above two evaluation methods to conduct process evaluation. If conditions permit, the evaluation results of the two evaluation methods can also be comprehensively considered to further improve the accuracy of the process evaluation value of the emergency process.

[0084] In a feasible implementation, S105 may specifically include the following sub-steps: S105-1: Obtain emergency reporting information of the target emergency event.

[0085] In this embodiment, considering that different emergency departments may upload emergency reporting information of different forms and contents when facing different emergency events, in order to improve the efficiency and accuracy of information matching and facilitate the process generation model to extract and recognize features, a large language model will be used to identify the emergency reporting information and convert the emergency reporting information into target emergency demand information in a preset format.

[0086] S105-2: Input the emergency reporting information and the preset prompt engineering text into the large language model, and output the target emergency demand information in a preset format.

[0087] In this embodiment, the prompt engineering text includes emergency auxiliary information, emergency data text and format constraint text for indicating a preset format, and the emergency data text includes an emergency plan library, a standard emergency manual and a historical emergency plan.

[0088] It should be noted that the emergency auxiliary information is emergency information related to the target emergency event. For example, if the target emergency event is a fire in a certain area, the emergency auxiliary information may include weather information, geographical location information and / or terrain features of the area.

[0089] In this embodiment, the format constraint text may be set according to input acceptable to the command system.

[0090] In one example, the chain of command has the following four acceptable inputs: 1. xx materials need to be allocated to xx place.

[0091] 2. xx location needs to dispatch xx personnel and xx materials.

[0092] 3. Rapid evacuation is required near xx.

[0093] 4. A certain type of emergency personnel needs to be deployed quickly to area xx.

[0094] Then the format constraint text can be set as: Please output according to one of the strict descriptions in 1, 2, 3, and 4. In this way, the prompt engineering text can be set as: Please output the target emergency demand information in the preset format that meets the above format constraint text instructions according to the situation, combined with the emergency auxiliary information and emergency material text (emergency plan library, standard emergency manual and historical emergency plan).

[0095] In a feasible implementation, the plurality of preferred emergency processes include respective corresponding preset emergency demand information; in S106, based on the target emergency demand information, the step of determining the target emergency process in the plurality of preferred emergency processes may specifically include the following sub-steps: S106-1: Determine characteristic distances between a plurality of preset emergency demand information and target emergency demand information.

[0096] It should be noted that when sampling the emergency process of the command system, the preset emergency demand information corresponding to each emergency process will be sampled from the preset multiple preset emergency demand information, so that the command system can generate multiple emergency plans for any emergency process using the multiple process branches of the emergency process to meet the preset emergency demand information corresponding to the emergency process.

[0097] In the specific implementation, after determining the target emergency demand information, the target emergency demand information will be feature extracted to obtain the target feature vector corresponding to the target emergency demand information, and the target feature vector will be measured with the feature vectors of multiple preset emergency demand information to obtain the characteristic distance between each preset emergency demand information and the target emergency demand information.

[0098] S106-2: Determine the preferred emergency process corresponding to the preset emergency demand information with the shortest characteristic distance as the target emergency process.

[0099] It should be noted that the closer the characteristic distance between the target emergency demand information and a certain preset emergency demand information, the higher the matching degree between the target emergency demand information and the preset emergency demand information. Therefore, by selecting the preferred emergency process corresponding to the preset emergency demand information with the closest characteristic distance as the target emergency process, it can be ensured that the multiple emergency plans generated by the target emergency process can effectively meet the emergency needs of the target emergency event.

[0100] In this embodiment, in order to improve the effectiveness and accuracy of the target emergency process, the process evaluation value of the preferred emergency process corresponding to each preset emergency demand information can also be comprehensively considered based on the characteristic distance between multiple preset emergency demand information and the target emergency demand information.

[0101] Specifically, based on the process evaluation values ​​of multiple preferred emergency processes, the characteristic distances of the preset emergency demand information of multiple preferred emergency processes can be corrected to obtain the corrected characteristic distance of each preset emergency demand information; and then the preferred emergency process corresponding to the preset emergency demand information with the shortest corrected characteristic distance is determined as the target emergency process.

[0102] In a feasible implementation, the process generation model is trained by the following steps: S301: For any preferred emergency process, generate a sample data group corresponding to the preferred emergency process.

[0103] In this implementation, considering that the emergency process has obvious sequence characteristics, that is, the decision of the command node at the current level depends on the decision of the command node at the previous level, and the decision of the command node at the next level depends on the decision of the command node at the current level. Therefore, in order to improve the application effect of the process generation model, the process generation model can be constructed and trained based on a recurrent neural network (RNN).

[0104] It should be noted that recurrent neural networks are good at processing sequence data and can effectively understand and analyze the temporal relationship and dependency between demands. They have the ability to remember past information and can store historical information through the recurrent connection of hidden states and use it for subsequent predictions.

[0105] In this embodiment, in the model application stage, the input of the process generation model includes the state information of the previous step of the emergency process and the output information of the previous step. The state information of the previous step includes the command seat of the selected command node, business software, micro-applications, micro-services, functional algorithm clusters and algorithms, and the output information of the previous step is the output of the algorithm of the previous step. The output of the process generation model includes the selection of the next step of the emergency process and the auxiliary input suggestions of the algorithm. The selection of the next step includes the command seat of the selected command node, business software, micro-applications, micro-services, functional algorithm clusters, and algorithms, wherein the auxiliary input suggestions of the algorithm include the constraints and optimization goals of the algorithm.

[0106] In this embodiment, during the model training phase, a sample data group corresponding to the preferred emergency process can be generated according to the actual hierarchical distribution of the command nodes of the command system. The sample data group includes n continuous sample input data and process label data corresponding to each sample input data, where n is a positive integer greater than 1.

[0107] Specifically, the sample data set can be represented as {x1, y1, x2, y2, ..., ..., x n ,y n}. Among them, x1, x2, ..., x n Represents n consecutive sample input data; y1, y2, ..., y n Represent x1, x2, ..., x n The corresponding process label data.

[0108] It should be further explained that x1, as the first sample input data, is the preset emergency demand information corresponding to the preferred emergency process. The preset emergency demand information is the demand information that the top-level command node needs to meet. y1 represents the command node seat and corresponding functional services that should be used to meet the preset emergency demand information corresponding to x1. x2, as the next sample input data of the first sample input data, represents the emergency demand information for the next level after the task is performed using the seat and functional services corresponding to y1. y2 represents the command node seat and corresponding functional services that should be used by the next level to meet the emergency demand information corresponding to x2. And so on. n As the last sample input data, it is expressed as using y n-1 After the corresponding seats and functional services perform their tasks, they face the emergency demand information at the lowest level, y n Then the bottom layer satisfies x n The command node seats and corresponding functional services that should be used for the corresponding emergency demand information. In other words, based on the sample data group, a complete optimal emergency process can be extracted, and a complete optimal emergency process can be split into the corresponding sample data group.

[0109] S302: Determine the first sample input data as the target data, input the target data into the initial recurrent neural network model, output the predicted process, and calculate the loss function value based on the predicted process and the process label data corresponding to the target data.

[0110] It should be noted that the initial recurrent neural network model is based on the target data, and the output prediction process may be significantly different from the ideal process (i.e., the process indicated by the process label data). Therefore, by calculating the loss function value between the predicted process and the process label data, the degree of difference between the two schemes can be determined so as to subsequently update the parameters of the initial recurrent neural network model.

[0111] S303: Determine the next sample input data of the first sample input data as the target data, input the target data into the initial recurrent neural network model, output the prediction process, and calculate the loss function value based on the prediction process and the process label data corresponding to the target data, until the nth loss function value corresponding to the nth sample input data is calculated.

[0112] It should be noted that in the model training stage, if the predicted process is used as the target data for the next round of training, it may lead to a larger and larger process deviation. Therefore, in each round of training, the sample input data will be input into the initial recurrent neural network model as the target data in the order indicated by the sample data group, and the loss function value corresponding to each sample input data will be calculated, and then n loss function values ​​can be calculated for each sample data group. In the process of model application, the predicted process obtained by the current step prediction of the process generation model is used as the input data for the next step prediction, until the last step prediction obtains the complete target emergency process.

[0113] In one example, the sample data set is {x1, y1, x2, y2, x3, y3}, then based on the sample data set, refer to Figure 2 , shows a schematic diagram of training the initial recurrent neural network model, and the training process of the initial recurrent neural network model specifically includes the following steps: S1: x1 and h0 (default zero vector) are input into the initial recurrent neural network model, and y1' and h1 are output. The loss function is calculated by y1' and y1.

[0114] Among them, h represents the hidden state of the recurrent neural network model, h0 is the preset zero vector, h1 represents the hidden state containing the feature information of the sample input data x1 in the previous step, and y1' represents the prediction process.

[0115] It should be noted that the hidden state of the recurrent neural network model contains feature information of the previous sequence, which is used to store the input information of the current time step and the information of the previous time step. It allows the RNN to remember the important previous content in the sequence, so as to better understand the relationship between the current input and historical information.

[0116] S2: x2 and h1 are input into the initial recurrent neural network model together, and y2' and h2 are output. The loss function is calculated by y2' and y2.

[0117] S2: x3 and h2 are input into the network model together, and y3' and h3 are output. y3' and y3 are used to calculate the loss function.

[0118] In this embodiment, the sequence reasoning ability of the initial recurrent neural network model is utilized so that the initial recurrent neural network model can well extract the sequence features in the sample data group, and then combine the context information of different time steps to better complete the corresponding reasoning tasks in each process link.

[0119] S304: Based on the n loss function values ​​corresponding to a preset number of preferred emergency processes, determine the target loss function value, and update the network parameters of the initial recurrent neural network model based on the target loss function value until the initial recurrent neural network model meets the training cutoff condition to obtain a process generation model.

[0120] In this embodiment, the preset number can be set to 60, that is, each time the initial recurrent neural network model is trained using the sample data groups corresponding to the 60 preferred emergency processes, the target loss function value will be determined based on the 60×n loss function values ​​calculated during the training process, and then the network parameters of the initial recurrent neural network model will be updated using the target loss function value until the initial recurrent neural network model meets the training cutoff condition and completes the training. Among them, the training cutoff condition can be set to: the fluctuation amplitude of the target loss function value reaches less than the amplitude threshold or the training rounds of the initial recurrent neural network model reach the round threshold.

[0121] In this embodiment, by adopting a recurrent neural network model for training, the advantages of the recurrent neural network model in processing sequence data can be fully utilized, so that the trained process generation model has obvious advantages in generating emergency processes based on emergency demand information, which helps to achieve fast, flexible and effective emergency response.

[0122] Second, refer to Figure 3 The embodiment of the present application provides an intelligent auxiliary device 300 of a command system, and the intelligent auxiliary device 300 of the command system includes: The process sampling module 301 is used to sample the emergency process of the command system to obtain multiple emergency processes; wherein the emergency process includes the selected command nodes at each level of the command system and the selected emergency parameters of the selected command nodes; The process sending module 302 is used to send multiple emergency processes to the command system, so that the command system generates multiple emergency plans based on multiple process branches of the emergency process for any emergency process, and determines the plan evaluation values ​​corresponding to each of the multiple emergency plans; wherein different emergency plans are generated by selecting a command node at the end of different process branches; The process evaluation module 303 is used to receive the solution evaluation values ​​of multiple emergency solutions sent by the command system for any emergency process, and determine the process evaluation value of the emergency process based on the solution evaluation values ​​of the multiple emergency solutions, so as to obtain the process evaluation values ​​corresponding to each of the multiple emergency processes; A process screening module 304 is used to screen multiple emergency processes based on the process evaluation values ​​corresponding to the multiple emergency processes and the solution evaluation values ​​of the multiple emergency solutions to obtain multiple preferred emergency processes; Information determination module 305, used to determine target emergency demand information of target emergency events; The process determination module 306 is used to determine the target emergency process among multiple preferred emergency processes based on the target emergency demand information; or, input the target emergency demand information into the process generation model and output the target emergency process; wherein the process generation model is obtained based on training of multiple preferred emergency processes.

[0123] In one embodiment of the present application, the process sampling module 301 includes: The knowledge graph construction submodule is used to construct the knowledge graph of the command system; the knowledge graph includes the node graph and the function graph of the command system; the node graph represents the command nodes of each level of the command system and the connection relationship between the command nodes of each level, and the function graph represents at least one business software of the command system, at least one micro application under the business software, at least one micro service under the micro application, at least one function algorithm cluster under the micro service, and at least one algorithm under the function algorithm cluster; The graph sampling submodule is used to sample the node graph in the knowledge graph to obtain the selected command nodes at each level, and to sample the function graph in the knowledge graph to obtain the selected emergency parameters of each selected command node; The emergency process determination submodule is used to obtain multiple emergency processes based on the selected command nodes at each level and the selected emergency parameters of each selected command node.

[0124] In one embodiment of the present application, the intelligent auxiliary device 300 of the command system further includes: The knowledge graph updating module is used to update the node graph and / or function graph based on the updated information of the command system to obtain a reconstructed knowledge graph.

[0125] In one embodiment of the present application, the process evaluation module 303 includes: A weight parameter determination submodule is used to determine the weight parameter of any emergency plan based on the importance of the process branch for generating the emergency plan; The process evaluation value determination submodule is used to determine the process evaluation value of the emergency process based on the solution evaluation values ​​and weight parameters corresponding to each of the multiple emergency solutions.

[0126] In one embodiment of the present application, the intelligent auxiliary device 300 of the command system further includes: A process simulation module is used to input multiple emergency plans into a pre-built digital simulation evaluation system to obtain simulation results output by the digital simulation evaluation system; The simulation evaluation module is used to determine the process evaluation value of the emergency process based on the simulation results.

[0127] In one embodiment of the present application, the process screening module 304 includes: The threshold comparison submodule is used to determine any emergency process as a preferred emergency process when the process evaluation value of the emergency process is greater than a first threshold, or the solution evaluation values ​​of multiple emergency solutions of the emergency process are all greater than a second threshold.

[0128] In one embodiment of the present application, the information determination module 305 includes: The information acquisition submodule is used to obtain the emergency reporting information of the target emergency event; The model recognition submodule is used to input the emergency reporting information and the preset prompt engineering text into the large language model, and output the target emergency demand information in a preset format; the prompt engineering text includes emergency auxiliary information, emergency data text and format constraint text used to indicate the preset format, and the emergency data text includes the emergency plan library, standard emergency manual and historical emergency plan.

[0129] In an embodiment of the present application, a plurality of preferred emergency processes include respective corresponding preset emergency demand information; the process determination module 306 includes: A distance determination submodule, used to determine characteristic distances between a plurality of preset emergency demand information and target emergency demand information; The process determination submodule is used to determine the preferred emergency process corresponding to the preset emergency demand information with the closest characteristic distance as the target emergency process.

[0130] In one embodiment of the present application, the intelligent auxiliary device 300 of the command system further includes: The sample generation module is used to generate a sample data group corresponding to any preferred emergency process; the sample data group includes n consecutive sample input data and process label data corresponding to each sample input data, where n is a positive integer greater than 1; A first sample input module is used to determine the first sample input data as the target data, and input the target data into the initial recurrent neural network model, output the predicted process, and calculate the loss function value based on the predicted process and the process label data corresponding to the target data; A second sample input module is used to determine the next sample input data of the first sample input data as the target data, input the target data into the initial recurrent neural network model, output a prediction process, and calculate the loss function value based on the prediction process and the process label data corresponding to the target data, until the nth loss function value corresponding to the nth sample input data is calculated; The parameter updating module is used to determine the target loss function value based on the n loss function values ​​corresponding to a preset number of preferred emergency processes, and to update the network parameters of the initial recurrent neural network model based on the target loss function value until the initial recurrent neural network model meets the training cutoff condition to obtain a process generation model.

[0131] It should be noted that the specific implementation of the intelligent assistance device 300 of the command system of the embodiment of the present application refers to the specific implementation of the intelligent assistance method of the command system proposed in the first aspect of the embodiment of the present application, and will not be repeated here.

[0132] It will be appreciated by those skilled in the art that the embodiments of the present invention may be provided as methods, devices, or computer program products. Therefore, the embodiments of the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the embodiments of the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0133] The embodiments of the present invention are described with reference to the flowcharts and / or block diagrams of the methods, terminal devices (systems), and computer program products according to the embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of the processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing terminal device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing terminal device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0134] These computer program instructions may also be stored in a computer readable memory capable of directing a computer or other programmable data processing terminal device to operate in a specific manner, so that the instructions stored in the computer readable memory produce a manufactured product including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0135] These computer program instructions can also be loaded onto a computer or other programmable data processing terminal device so that a series of operating steps are executed on the computer or other programmable terminal device to produce a computer-implemented process, thereby providing instructions for executing on the computer or other programmable terminal device to implement the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0136] Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the embodiments of the present invention.

[0137] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or terminal device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or terminal device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or terminal device including the elements.

[0138] The above is a detailed introduction to the intelligent auxiliary method and device for a command system provided by the present invention. Specific examples are used in this article to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea; at the same time, for general technical personnel in this field, according to the idea of ​​the present invention, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

Claims

1. An intelligent auxiliary method for a command system, characterized in that: The method comprises: Sampling emergency processes of the command system to obtain multiple emergency processes; wherein the emergency processes include selected command nodes at each level of the command system and selected emergency parameters of the selected command nodes; Sending the plurality of emergency processes to the command system, so that the command system generates a plurality of emergency plans for any emergency process based on the plurality of process branches of the emergency process, and determines the plan evaluation values ​​corresponding to the plurality of emergency plans; wherein different emergency plans are generated by selecting the command node at the end of different process branches; Receiving the solution evaluation values ​​of the multiple emergency solutions sent by the command system for any emergency process, and determining the process evaluation value of the emergency process based on the solution evaluation values ​​of the multiple emergency solutions, so as to obtain the process evaluation values ​​corresponding to each of the multiple emergency processes; Based on the process evaluation values ​​corresponding to the multiple emergency processes and the solution evaluation values ​​of the multiple emergency solutions, the multiple emergency processes are screened to obtain multiple preferred emergency processes; Determine the target emergency demand information of the target emergency event; Based on the target emergency demand information, determine the target emergency process among the multiple preferred emergency processes; or input the target emergency demand information into a process generation model, and output the target emergency process; wherein the process generation model is trained based on multiple preferred emergency processes.

2. The intelligent assistance method for a command system according to claim 1, characterized in that: The step of sampling the emergency process of the command system to obtain multiple emergency processes includes: Constructing a knowledge graph of the command system; the knowledge graph includes a node graph and a function graph of the command system; the node graph represents command nodes at each level of the command system and connection relationships between command nodes at each level, and the function graph represents at least one business software of the command system, at least one micro-application under the business software, at least one micro-service under the micro-application, at least one function algorithm cluster under the micro-service, and at least one algorithm under the function algorithm cluster; Sampling the node graph in the knowledge graph to obtain selected command nodes at each level, and sampling the function graph in the knowledge graph to obtain selected emergency parameters of each selected command node; Based on the selected command nodes at each level and the selected emergency parameters of each selected command node, a plurality of emergency processes are obtained.

3. The intelligent assistance method for a command system according to claim 2, characterized in that: The method further comprises: Based on the updated information of the command system, the node graph and / or the function graph is updated to obtain a reconstructed knowledge graph.

4. The intelligent assistance method for a command system according to claim 1, characterized in that: The step of determining the process evaluation value of the emergency process based on the solution evaluation values ​​of the plurality of emergency solutions comprises: For any emergency plan, determining a weight parameter of the emergency plan based on the importance of the process branch that generates the emergency plan; Based on the solution evaluation values ​​and weight parameters corresponding to each of the multiple emergency solutions, the process evaluation value of the emergency process is determined.

5. The intelligent assistance method for a command system according to claim 1, characterized in that: The method further comprises: Inputting a plurality of the emergency plans into a pre-built digital simulation evaluation system to obtain simulation results output by the digital simulation evaluation system; Based on the simulation result, a process evaluation value of the emergency process is determined.

6. The intelligent assistance method for a command system according to claim 1, characterized in that: The step of screening the multiple emergency processes based on the process evaluation values ​​corresponding to the multiple emergency processes and the solution evaluation values ​​of the multiple emergency solutions to obtain multiple preferred emergency processes includes: For any emergency process, when the process evaluation value of the emergency process is greater than a first threshold, or the solution evaluation values ​​of multiple emergency solutions of the emergency process are all greater than a second threshold, the emergency process is determined as the preferred emergency process.

7. The intelligent assistance method for a command system according to claim 1, characterized in that: The step of determining target emergency demand information of a target emergency event includes: Obtaining emergency reporting information of the target emergency event; The emergency reporting information and the preset prompt engineering text are input into the large language model, and the target emergency demand information in the preset format is output; the prompt engineering text includes emergency auxiliary information, emergency data text and format constraint text for indicating the preset format, and the emergency data text includes an emergency plan library, a standard emergency manual and a historical emergency plan.

8. The intelligent assistance method for a command system according to claim 1, characterized in that: The plurality of preferred emergency processes include respective corresponding preset emergency demand information; The step of determining a target emergency process among the plurality of preferred emergency processes based on the target emergency demand information includes: Determining characteristic distances between a plurality of preset emergency demand information and the target emergency demand information; The preferred emergency process corresponding to the preset emergency demand information with the closest characteristic distance is determined as the target emergency process.

9. The intelligent assistance method for a command system according to claim 1, characterized in that: The process generation model is trained by the following steps: For any preferred emergency process, a sample data group corresponding to the preferred emergency process is generated; the sample data group includes n consecutive sample input data and process label data corresponding to each sample input data, where n is a positive integer greater than 1; Determine the first sample input data as the target data, input the target data into the initial recurrent neural network model, output a predicted process, and calculate the loss function value based on the predicted process and the process label data corresponding to the target data; Determine the next sample input data of the first sample input data as the target data, perform the steps of inputting the target data into the initial recurrent neural network model, outputting a prediction process, and calculating the loss function value based on the prediction process and the process label data corresponding to the target data, until the nth loss function value corresponding to the nth sample input data is calculated; Based on the n loss function values ​​corresponding to a preset number of the preferred emergency processes, the target loss function value is determined, and the network parameters of the initial recurrent neural network model are updated based on the target loss function value until the initial recurrent neural network model meets the training cutoff condition, thereby obtaining the process generation model.

10. An intelligent auxiliary device for a command system, characterized in that: The device comprises: A process sampling module is used to sample the emergency process of the command system to obtain multiple emergency processes; wherein the emergency process includes selected command nodes at each level of the command system and selected emergency parameters of the selected command nodes; A process sending module, used for sending the plurality of emergency processes to the command system, so that the command system generates a plurality of emergency plans for any emergency process based on the plurality of process branches of the emergency process, and determines the plan evaluation values ​​corresponding to the plurality of emergency plans; wherein different emergency plans are generated by selecting a command node at the end of different process branches; A process evaluation module, used for receiving the solution evaluation values ​​of the multiple emergency solutions sent by the command system for any emergency process, and determining the process evaluation value of the emergency process based on the solution evaluation values ​​of the multiple emergency solutions, so as to obtain the process evaluation values ​​corresponding to each of the multiple emergency processes; A process screening module, used to screen the multiple emergency processes based on the process evaluation values ​​corresponding to the multiple emergency processes and the solution evaluation values ​​of the multiple emergency solutions, to obtain multiple preferred emergency processes; An information determination module, used to determine target emergency demand information of a target emergency event; A process determination module is used to determine a target emergency process among multiple preferred emergency processes based on the target emergency demand information; or, input the target emergency demand information into a process generation model and output a target emergency process; wherein the process generation model is obtained by training based on multiple preferred emergency processes.

Images