A flight area simulation method, apparatus, equipment, and medium based on an AI model.

Abstract

This application provides a flight area simulation method, apparatus, equipment, and medium based on an AI model. It includes: modeling each object to be executed based on the pending business, and determining the corresponding business state and preset state conditions; determining the target state of the object to be executed based on the real-time state and logical rules of the modeled object, as well as the business state and preset state conditions; and for objects to be executed with the same target state, determining the execution urgency value of the object to be executed based on the AI ​​model, the state information, attribute information, and preset urgency information of each object to be executed; determining the priority of each object to be executed based on the execution urgency value, and simulating the object with the highest priority to obtain the simulation result; and adjusting the AI ​​model based on the corresponding indicator results and feedback results of the simulation result to obtain the adjusted target AI model. This improves the efficiency and accuracy of the simulation.

Description

Technical Field

[0001] This application relates to the field of flight area operation simulation technology, and in particular to a flight area simulation method, device, equipment and medium based on an AI model. Background Technology

[0002] The flight zone is the ground harbor for aircraft operations and is related to the safety and efficiency of the air transport industry. Simulating the operation of aircraft, vehicles, and support operations in the flight zone and predicting future operational indicators are of great significance for improving the operational level of the flight zone.

[0003] Existing flight area simulation methods mainly employ classical algorithms to simulate specific scenarios, such as airport operational capacity assessment and runway / taxiway intersection congestion analysis. These methods use simulated operational data and classical algorithms for simulation analysis. Alternatively, digital twin technology can be used, leveraging physical models and particle models within a 3D engine for simulation.

[0004] However, existing flight area simulation methods suffer from low efficiency and accuracy. Summary of the Invention

[0005] This application provides a flight area simulation method, apparatus, equipment, and medium based on an AI model to address the problems of low efficiency and accuracy in existing flight area simulation methods.

[0006] Firstly, this application provides a flight area simulation method based on an AI model, the method comprising:

[0007] Based on the business to be executed, model each object to be executed, and determine the corresponding business status and preset status conditions;

[0008] Based on the real-time status and logical rules of the modeled objects to be executed, as well as the business status and preset status conditions, the target status of the objects to be executed is determined. For objects to be executed with the same target status, the execution urgency value of the objects to be executed is determined based on the AI ​​model, the status information, attribute information and preset urgency information of each object to be executed.

[0009] Based on the execution urgency value, the priority of each object to be executed is determined, and the highest priority object to be executed is simulated to obtain the simulation results;

[0010] Based on the simulation results and corresponding index results and feedback results, the AI ​​model is adjusted to obtain the adjusted target AI model.

[0011] In some embodiments of this application, the target state of the object to be executed is determined based on the real-time state and logical rules of the modeled object, as well as the business state and preset state conditions. For objects to be executed with the same target state, the execution urgency value of the object to be executed is determined based on the AI ​​model, the state information, attribute information, and preset urgency information of each object to be executed, including:

[0012] Determine the logical rules corresponding to each object to be executed, and the preset state conditions corresponding to each real-time state;

[0013] When the logic rule is a serial rule, determine whether each real-time state satisfies the corresponding preset state condition.

[0014] When the logical rule is a parallel rule, determine whether there is at least one real-time state that satisfies the corresponding preset state condition.

[0015] If so, then the target state of the object to be executed is determined to be the business state;

[0016] If not, then it is determined that the target state of the object to be executed does not exist;

[0017] For objects to be executed with the same target state, the execution urgency value of the objects to be executed is determined based on the AI ​​model, the state information, attribute information and preset urgency information of each object to be executed.

[0018] In some embodiments of this application, for objects to be executed with the same target state, the execution urgency value of the objects to be executed is determined based on the AI ​​model, the state information, attribute information, and preset urgency information of each object to be executed, including:

[0019] Compare the target states of each object to be executed to see if they are the same, and obtain the comparison results;

[0020] Based on the comparison results, determine the objects to be executed that have the same target state;

[0021] For each object to be executed, the execution urgency value is determined based on the AI ​​model, the status information, attribute information, and preset urgency information of each object to be executed.

[0022] In some embodiments of this application, for an object to be executed, the execution urgency value of the object to be executed is determined based on the AI ​​model, the state information and attribute information of each object to be executed, and preset urgency information, including:

[0023] Determine the preset emergency information and check whether the status information of the object to be executed includes the preset emergency information;

[0024] If so, the object to be executed is determined to be an urgent object to be executed, and the weight value corresponding to the preset urgent information included in the status information is determined.

[0025] Based on the AI ​​model, the weight values ​​corresponding to the urgent objects to be executed, and the preset urgency information, the execution urgency value of the urgent objects to be executed is determined.

[0026] If not, the execution urgency value is determined based on the status and attribute information of the object to be executed.

[0027] In some embodiments of this application, before determining that the object to be executed is an urgent object to be executed and determining the weight value corresponding to the preset urgent information included in the status information, if yes, the method further includes:

[0028] Multiple preset emergency messages are identified, and the urgency of each preset emergency message is ranked based on user feedback.

[0029] Input the sorted preset emergency information into the AI ​​model, and obtain the weight value assigned by the AI ​​model to each preset emergency information.

[0030] In some embodiments of this application, multiple preset emergency messages are determined, and the urgency of each preset emergency message is ranked according to user feedback, including:

[0031] Based on web scraping, several preset emergency messages were identified;

[0032] Based on the feedback, the urgency of each preset emergency message is sorted.

[0033] In some embodiments of this application, the execution urgency value of the urgent object to be executed is determined based on the AI ​​model, the weight value corresponding to the urgent object to be executed, and preset urgency information, including:

[0034] Determine the weight values ​​corresponding to preset emergency information;

[0035] Input preset emergency information into the AI ​​model, obtain the information value corresponding to the preset emergency information output by the AI ​​model, and determine the emergency value corresponding to the preset emergency information based on the product of the weight value and the information value. The information value is the overall value of the information output by the AI ​​model after weighting and assigning values ​​to the various data included in the input information, which is used to characterize the numerical quantification index of the information.

[0036] The execution emergency value is determined by the sum of all emergency values ​​corresponding to the urgent pending objects.

[0037] In some embodiments of this application, before inputting preset emergency information into the AI ​​model, obtaining the information value corresponding to the preset emergency information output by the AI ​​model, and determining the emergency value corresponding to the preset emergency information based on the product of the weight value and the information value, the method further includes:

[0038] The system identifies all real-time data included in the target emergency information and the corresponding target emergency values ​​as sample cases. Multiple sample cases are then input into the AI ​​model to obtain the trained AI model. The AI ​​model is able to perform weighted assignment processing on the data included in the input information and generate the corresponding information values.

[0039] In some embodiments of this application, if not, the execution urgency value is determined based on the state information and attribute information of the object to be executed, including:

[0040] If not, input each attribute information and state information into the AI ​​model to obtain the corresponding information value, and add the multiplication of the information value and weight value corresponding to each attribute information to obtain the total emergency value corresponding to all attribute information, and add the multiplication of the information value and weight value corresponding to each state information to obtain the total emergency value corresponding to all state information.

[0041] The emergency value is determined based on the total emergency value corresponding to the attribute information and status information, as well as multiple preset emergency ranges.

[0042] In some embodiments of this application, before determining the execution emergency value based on the total emergency value corresponding to the attribute information and the status information, and multiple preset emergency ranges, the method further includes:

[0043] Determine multiple target urgency values ​​and sort them according to their numerical values.

[0044] Based on any two adjacent target urgency values, determine multiple preset urgency ranges and the weight value corresponding to each preset urgency range.

[0045] In some embodiments of this application, an execution emergency value is determined based on the total emergency value corresponding to the attribute information and the status information, as well as multiple preset emergency ranges, including:

[0046] Based on multiple preset emergency ranges, the target ranges corresponding to the total emergency values ​​of attribute information and status information are determined respectively; and the weight value of the target range corresponding to the attribute information is determined as the first target weight value, and the weight value of the target range corresponding to the status information is determined as the second target weight value.

[0047] The product of the total urgency value of the attribute information and the first target weight value is determined as the first target multiplier value, and the product of the total urgency value of the status information and the second target weight value is determined as the second target multiplier value. Based on the sum of the first target multiplier value and the second target multiplier value, the execution urgency value is determined.

[0048] In some embodiments of this application, the AI ​​model is adjusted based on the corresponding index results and feedback results from the simulation results to obtain an adjusted target AI model, including:

[0049] Determine the simulation results for each indicator, their corresponding preset threshold values, and the feedback results;

[0050] Compare the results of each indicator with the corresponding preset indicator thresholds to obtain the comparison results, and determine the indicator results that are less than the preset indicator thresholds based on the comparison results.

[0051] Based on the results of the indicators to be adjusted and the feedback results, the model parameters of the AI ​​model are adjusted to obtain the target AI model.

[0052] Secondly, this application provides a flight area simulation device based on an AI model, the device comprising:

[0053] The modeling module is used to model each object to be executed based on the business to be executed, and to determine the corresponding business state and preset state conditions.

[0054] The state determination module is used to determine the target state of the object to be executed based on the real-time state and logical rules of the modeled object to be executed, as well as the business state and preset state conditions. For objects to be executed with the same target state, the module determines the execution urgency value of the object to be executed based on the AI ​​model, the state information, attribute information and preset urgency information of each object to be executed.

[0055] The simulation module is used to determine the priority of each object to be executed based on the execution urgency value, and to simulate the object with the highest priority to be executed to obtain the simulation results;

[0056] The adjustment module is used to adjust the AI ​​model based on the corresponding index results and feedback results of the simulation results, so as to obtain the adjusted target AI model.

[0057] Thirdly, this application provides an apparatus, including: a processor, and a memory communicatively connected to the processor;

[0058] The memory stores the instructions that the computer executes;

[0059] The processor executes computer execution instructions stored in memory to implement the method of this application.

[0060] Fourthly, this application provides a computer-readable storage medium storing program code, which, when executed by a processor, is used to implement the method of this application.

[0061] This application provides a flight area simulation method, apparatus, equipment, and medium based on an AI model. The method involves modeling each object to be executed based on the pending business, and determining the corresponding business state and preset state conditions. Based on the real-time state and logical rules of the modeled objects, as well as the business state and preset state conditions, the target state of the objects is determined. For objects with the same target state, the execution urgency value is determined based on the AI ​​model, the state information, attribute information, and preset urgency information of each object. The priority of each object is determined based on the execution urgency value, and the highest-priority object is simulated to obtain the simulation results. Based on the corresponding indicator results and feedback results from the simulation, the AI ​​model is adjusted to obtain the adjusted target AI model.

[0062] In this way, prominent issues in flight area operation simulation, such as simulation cycle, efficiency, local refinement, and simulation logic arrangement, are resolved, effectively improving the speed and accuracy of flight area operation simulation. Through simulation agent modeling, the simulation process can be visualized and arranged, and through continuous refinement, local refined modeling of the business scenarios of interest can be achieved. Through the analysis of simulation index data such as resource demand, resource planning, and resource requests, an evaluation index model is established for the simulation status of vehicles, personnel, and equipment in the flight area, guiding the rational planning and scheduling of resources, avoiding problems such as blind investment and potential operational risks, and improving the efficiency and level of support. Attached Figure Description

[0063] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0064] Figure 1 A flowchart illustrating a flight area simulation method based on an AI model, provided for an embodiment of this application;

[0065] Figure 2 A schematic diagram illustrating a flight area simulation method based on an AI model, provided in an embodiment of this application;

[0066] Figure 3 A simulation diagram illustrating a flight area simulation method based on an AI model, provided as an embodiment of this application;

[0067] Figure 4 A schematic diagram illustrating a flight area simulation method based on an AI model, provided as an embodiment of this application;

[0068] Figure 5 An optimized schematic diagram of a flight area simulation method based on an AI model provided in this application embodiment;

[0069] Figure 6 A schematic diagram of the structure of a flight area simulation device based on an AI model provided in this application embodiment;

[0070] Figure 7 This is a structural block diagram of an apparatus for performing a flight area simulation method based on an AI model according to an embodiment of this application. Detailed Implementation

[0071] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the claims.

[0072] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will now be described with reference to the accompanying drawings.

[0073] Figure 1 This is a flowchart illustrating a flight area simulation method based on an AI model, provided as an embodiment of this application. Figure 1 As shown, this AI model-based flight area simulation method may include the following steps:

[0074] S110. Based on the business to be executed, model each object to be executed and determine the corresponding business status and preset status conditions.

[0075] Among them, pending business operations are business requirements that are waiting to be executed. For example, these could be various business requirements that may occur within the flight area, such as refueling the aircraft, using a trailer, or cleaning the aircraft.

[0076] The object to be executed is the target object that is waiting to be executed and is associated with the business to be executed. For example, if the business to be executed is that an aircraft needs to use a trailer, then the object to be executed can be a series of target objects related to this business, such as the aircraft and the trailer.

[0077] Modeling refers to abstracting various participating entities (such as aircraft, vehicles, service processes, etc.) involved in the operation of the flight area into digital models containing information in specific dimensions. By clarifying their key characteristics such as attributes, states, and behaviors, they can be identified, calculated, and interacted with by the simulation system.

[0078] The business status is the target status corresponding to the business to be executed, so as to determine whether the current status of the object to be executed can meet the execution requirements of the business to be executed. That is, the business requirements of the business to be executed can only be realized when the current status of the object to be executed is the business status. For example, the business to be executed may be that an aircraft needs to use the runway to take off. The corresponding business status is that the aircraft has been pushed back. Only when the real-time status of the aircraft is pushed back can the execution requirements of takeoff be met.

[0079] The preset state conditions are the state requirements of the object to be executed for the business to be executed. They are used to determine whether the object to be executed can execute the business to be executed, thereby determining whether the state of the object to be executed is a business state. For example, the business to be executed may be that an aircraft needs to use the runway to take off. The business state is that the aircraft has been pushed back. The corresponding preset state condition is whether the aircraft has been pushed back. If the aircraft meets the preset state conditions, that is, the aircraft has been pushed back, then the real-time state of the current aircraft is the business state, and it can execute the business to be executed.

[0080] Based on this, by identifying several pending tasks within the flight area, each pending task object is modeled, and the corresponding task status and preset status conditions are determined. In order to determine whether the pending task object meets the preset status conditions, if it does, the real-time status of the current pending task object is the corresponding task status, which can be used to execute the pending task, so that the task object can be further executed in the future.

[0081] S120. Based on the real-time status and logical rules of the modeled objects to be executed, as well as the business status and preset status conditions, determine the target status of the objects to be executed. For objects to be executed with the same target status, determine the execution urgency value of the objects to be executed based on the AI ​​model, the status information, attribute information and preset urgency information of each object to be executed.

[0082] Among them, the logical rules are logical operation rules, which are calculated by the activation function based on the state of the object to be executed, thereby determining whether the object to be executed meets the preset state conditions; the logical rules include serial rules and parallel rules. Serial rules represent logical AND operations, and parallel rules represent logical OR operations. Logical AND operations refer to the operation method in which multiple conditions must be met simultaneously to trigger a certain logical result, while logical OR operations refer to the operation method in which a certain logical result can be triggered as long as one of the multiple conditions is met.

[0083] The target state is the state of the object to be executed that meets the preset state conditions.

[0084] AI models are AI intelligent models. Through deep learning, AI intelligent models have significantly improved in terms of efficiency, resource requirements, and general problem-solving capabilities, exceeding the results given by traditional programming algorithms. They can be large models like Qwen3-32b, enabling AI decision-making, including AI dialogue APIs, decision-making APIs, and diagnostic APIs.

[0085] The status information is the real-time status information of the object to be executed. For example, if the object to be executed is an aircraft, the status information can be various real-time information such as the current number of passengers, remaining fuel, and the time since the aircraft missed its flight, which is used to characterize the real-time status of the object to be executed. The attribute information is the basic information of the object to be executed, such as the aircraft type and fixed fuel consumption, which are fixed information that do not change with the real-time status of the aircraft. In other words, the attribute information characterizes the static information of the object to be executed. It is an inherent basic parameter that does not change during the simulation cycle. It is usually defined during initialization and is only read and not modified during runtime. The status information, on the other hand, characterizes the real-time dynamic information of the object to be executed. It is temporary data that changes in real time during the simulation process and is dynamically updated with events, time, or interactions.

[0086] Preset emergency information is information that is pre-set and related to emergency scenarios. It is used to quickly trigger corresponding simulation decisions or behaviors when sudden emergencies (such as aircraft malfunctions, emergency medical needs, etc.) occur, ensuring that the simulation process conforms to the actual emergency logic. For example, high-risk emergency events such as fires and personnel injuries need to be dealt with first.

[0087] The execution urgency value is a quantitative indicator used to determine the execution priority among multiple pending objects. This ensures that in complex scenarios involving resource conflicts or emergencies, the execution priority is determined based on the execution urgency value of each pending object. This allows for accurate execution of business tasks based on the real-time status of the pending objects, better meeting the actual execution needs of the current scenario and breaking through the limitations of conventional decision-making logic. For example, in two flights that need to be pushed back for takeoff, one is an international flight that has already missed its flight by 10 minutes, and the other is a medical emergency flight that needs to take off as soon as possible. Through an AI model, pre-set emergency information, attribute information, and status information are integrated to accurately allocate resources according to the actual scenario, avoiding the limitations of traditional classical algorithms.

[0088] Based on this, by analyzing the real-time status and corresponding logical rules of each object to be executed, the real-time status of the object to be executed is determined according to the logical rules and preset state conditions. This determines the target state of the object to be executed. For objects to be executed with the same target state, i.e., objects that all need to perform the same business, such as several aircraft that need to be pushed back for takeoff, the AI ​​model comprehensively evaluates preset emergency information, attribute information, and status information to determine the execution urgency value of the object to be executed. This allows the execution priority of each object to be executed to be determined based on the execution urgency value, thereby accurately allocating resources.

[0089] S130. Based on the execution urgency value, determine the priority of each object to be executed, and simulate the object with the highest priority to be executed to obtain the simulation results.

[0090] Simulation refers to the technical means of constructing digital models to reproduce the operation process and interaction logic of various entities (such as aircraft, vehicles, and personnel) in the flight area, and to dynamically deduce their behavior and state changes in a computer environment.

[0091] Based on this, for objects to be executed with the same target state, under the condition of limited resources, the execution priority is determined by determining the execution urgency value of each object to be executed, so that the object with the highest priority is executed first, and the corresponding simulation results are obtained.

[0092] S140. Based on the simulation results and corresponding index results and feedback results, adjust the AI ​​model to obtain the adjusted target AI model.

[0093] Among them, the indicator results refer to the quantitative data calculated and output through the simulation process, which are used to evaluate the operational efficiency, resource utilization and potential problems of the flight area. They are the core basis for measuring the simulation effect and guiding actual operations. For example, they can be multiple key indicators such as the resource utilization rate, resource tension, resource fatigue, delay rate and delay rate of the object to be executed.

[0094] Feedback results refer to human feedback on simulation results. This is used to determine the user's feedback on the simulation so that the AI ​​model can be adjusted based on the user's feedback, making subsequent simulations more closely reflect the user's actual needs.

[0095] Based on this, after the simulation ends, simulation records can be generated according to the various data collected during the simulation process. By analyzing the indicator results in the simulation records and the user's feedback on the simulation results, the AI ​​model can be adjusted in a timely manner to obtain a target AI model that better meets the user's actual needs.

[0096] Based on the feasible implementation of S120 described above, this application further provides the following steps: determining the target state of the object to be executed based on the real-time state and logical rules of the modeled object to be executed, as well as the business state and preset state conditions; and for objects to be executed with the same target state, determining the execution urgency value of the object to be executed based on the AI ​​model, the state information, attribute information, and preset urgency information of each object to be executed.

[0097] Determine the logical rules corresponding to each object to be executed, and the preset state conditions corresponding to each real-time state;

[0098] When the logic rule is a serial rule, determine whether each real-time state satisfies the corresponding preset state condition.

[0099] When the logical rule is a parallel rule, determine whether there is at least one real-time state that satisfies the corresponding preset state condition.

[0100] If so, then the target state of the object to be executed is determined to be the business state;

[0101] If not, then it is determined that the target state of the object to be executed does not exist;

[0102] For objects to be executed with the same target state, the execution urgency value of the objects to be executed is determined based on the AI ​​model, the state information, attribute information and preset urgency information of each object to be executed.

[0103] Based on this, by determining whether the logical rule corresponding to the object to be executed is a serial rule or a parallel rule, the real-time state and preset state conditions of the object to be executed are judged according to different logical rules. This is to determine whether the object to be executed meets the preset state conditions. When the conditions are met, the target state of the object to be executed is determined as the business state, which indicates that the object to be executed can be used to execute business. For example, if the business to be executed is that an aircraft needs to push back for takeoff, the object to be executed is the aircraft, and the preset state conditions include "passengers have arrived," "aircraft fuel level is normal," and "crew is ready," then the object to be executed can correspond to the serial rule. That is, the aircraft needs to ensure that all of the above preset state conditions are met before the target state of the aircraft can be determined as the business state, that is, the aircraft can be used to execute the business to be executed. Correspondingly, when the logical rule for a certain object to be executed is a parallel rule, it is only necessary to ensure that the object to be executed meets any one of the preset state conditions to determine the target state.

[0104] Furthermore, in practical applications, the real-time state of the object to be executed can be determined using single-state logic units to determine whether it meets preset state conditions. A single-state logic unit consists of a series of logic activation functions, and the logical result is the calculation result of the activation function based on the state. The system divides logic transport into two types: parallel structure and serial structure. The serial structure represents logical AND operations, and the parallel structure represents logical OR operations. A depth-first search algorithm (DFS) is used to calculate the logic units, and through a logic computing network, associated logic nodes are distributed to more logic computing units.

[0105] Based on the feasible implementation of S120 described above, this application further provides a method for determining the execution urgency value of objects to be executed that have the same target state, based on an AI model, the state information, attribute information, and preset urgency information of each object to be executed. This includes the following steps:

[0106] Compare the target states of each object to be executed to see if they are the same, and obtain the comparison results;

[0107] Based on the comparison results, determine the objects to be executed that have the same target state;

[0108] For each object to be executed, the execution urgency value is determined based on the AI ​​model, the status information, attribute information, and preset urgency information of each object to be executed.

[0109] Based on this, after determining the target state of each object to be executed, the objects to be executed with the same target state are identified by comparing the target states, such as aircraft that all need to be pushed back for takeoff, so that the execution priority of each object to be executed can be determined according to the AI ​​model.

[0110] Based on the feasible implementation of S120 described above, this application further provides a method for determining the execution urgency value of an object to be executed based on an AI model, the state information, attribute information, and preset urgency information of each object to be executed, including the following steps:

[0111] Determine the preset emergency information and check whether the status information of the object to be executed includes the preset emergency information;

[0112] If so, the object to be executed is determined to be an urgent object to be executed, and the weight value corresponding to the preset urgent information included in the status information is determined.

[0113] Based on the AI ​​model, the weight values ​​corresponding to the urgent objects to be executed, and the preset urgency information, the execution urgency value of the urgent objects to be executed is determined.

[0114] If not, the execution urgency value is determined based on the status and attribute information of the object to be executed.

[0115] Among them, the urgent pending objects are those whose status information includes preset urgent information. This indicates that the pending objects are urgent objects, have the highest execution priority, and the business of these objects should be executed first.

[0116] Weight values ​​are numerical values ​​that measure the relative importance of a factor, indicator, or variable in overall evaluation, decision-making, or calculation. They reflect the degree of influence on the final result by assigning different numerical proportions to different objects.

[0117] Based on this, when faced with multiple objects to be executed with the same target state, it is necessary to evaluate the execution priority so that the most needed object business is executed first when resources are limited. Therefore, by determining whether the state information includes preset urgency information, the objects to be executed with preset urgency are executed first. Furthermore, if there is only one urgent execution object among the multiple objects to be executed, the business of the only urgent execution object is executed first. If there is more than one urgent execution object, it is necessary to calculate the execution urgency value according to the preset urgency information and corresponding weight value of each urgent execution object, so as to determine the most urgent object to be executed and the one that needs to be executed most urgently, so as to execute the corresponding business in a timely manner.

[0118] Based on the feasible implementation of S120 described above, this application further provides the following steps before determining that the object to be executed is an urgent object to be executed and determining the weight value corresponding to the preset urgent information included in the status information:

[0119] Multiple preset emergency messages are identified, and the urgency of each preset emergency message is ranked based on user feedback.

[0120] Input the sorted preset emergency information into the AI ​​model, and obtain the weight value assigned by the AI ​​model to each preset emergency information.

[0121] The urgency level is a pre-set classification standard based on factors such as the urgency, scope of impact, and potential risks of the events associated with the information. The purpose is to quickly identify information priorities and ensure that high-urgency information receives priority response and processing.

[0122] Based on this, before practical application, multiple pre-defined emergency information items can be identified and sorted according to their respective urgency levels. This sorted information is then input into the AI ​​model, allowing the model to assign corresponding weights based on the urgency level of each item. The AI ​​model possesses self-learning and adaptive capabilities. If the relationship between urgency level and weight is dynamic or complex (e.g., the weight of "Level 1 Emergency" may differ in different scenarios, such as 0.4 in a medical setting and 0.3 in a fire setting), the model can learn this difference through a large amount of labeled data ("urgency level - weight" samples from different scenarios). This eliminates the need for manual pre-setting of rules for all scenarios. Even if new emergency levels (e.g., "Level 4 Emergency") are added later, only a small amount of data related to that level and its corresponding weight needs to be added. With appropriate weights, the model can autonomously adjust its internal parameters to incorporate new levels into the weight allocation system—this is a manifestation of self-learning ability, iterating and learning from data rather than relying on fixed code. AI models can also autonomously discover patterns from data without requiring manual rule writing. When the sorting rules for emergency information change (e.g., from "Special Level > Level 1" to "Level 1 > Special Level"), or the weight allocation logic is updated (e.g., from linear weight allocation to exponential weight allocation), the model doesn't need to redesign its architecture. It only needs "fine-tuning" with a small amount of new data to adapt to the new rules and output the required weight values. For unseen sub-scenarios (e.g., "Special Level Emergency - Chemical Leak"), the model can automatically match the corresponding weights based on the learned common patterns of "Special Level Emergency" (rather than failing due to being "unseen"). This also demonstrates adaptive ability, transferring existing experience to new scenarios.

[0123] Based on the feasible implementation of S120 described above, this application further provides the steps of determining multiple preset emergency messages and ranking the urgency of each preset emergency message according to user feedback, including:

[0124] Based on web scraping, several preset emergency messages were identified;

[0125] Based on the feedback, the urgency of each preset emergency message is sorted.

[0126] Web crawling is a technique that uses automated programs (web crawlers) to capture and collect web page data from the Internet according to certain rules. It can extract and store information (such as text, images, links, table data, etc.) scattered in web pages in batches and efficiently for subsequent analysis, processing or application.

[0127] Based on this, a large amount of pre-set emergency information is obtained through web crawling, so as to obtain enough samples for training the AI ​​model.

[0128] Based on the feasible implementation of S120 described above, this application further provides a method for determining the execution urgency value of an urgent object based on an AI model, the weight value corresponding to the urgent object to be executed, and preset urgency information, including the following steps:

[0129] Determine the weight values ​​corresponding to preset emergency information;

[0130] Input preset emergency information into the AI ​​model, obtain the information value corresponding to the preset emergency information output by the AI ​​model, and determine the emergency value corresponding to the preset emergency information based on the product of the weight value and the information value. The information value is the overall value of the information output by the AI ​​model after weighting and assigning values ​​to the various data included in the input information, which is used to characterize the numerical quantification index of the information.

[0131] The execution emergency value is determined by the sum of all emergency values ​​corresponding to the urgent pending objects.

[0132] The information value is the overall numerical value output by the AI ​​model after weighting and assigning values ​​to the various data included in the input information. It is used to represent the numerical quantitative indicator corresponding to the information. The preset emergency information can be aircraft malfunction, and the data included can be: aircraft status: aircraft type, flight number, number of passengers (including crew), remaining fuel, location of the malfunction (e.g., left engine failure, brake system failure); location information: whether it is on the runway / taxiway (specific coordinates), altitude (in case of in-flight malfunction), whether it has come to a complete stop; danger level: whether there is a fire (fire intensity), whether there is a fuel leak (leakage amount), whether... There is an explosion risk; personnel status: are there any injured persons (number of seriously injured / minorly injured), is an emergency evacuation required; scope of impact: is the runway / taxiway occupied (estimated runway closure time), will it affect other flight takeoffs and landings; at this time, it is necessary to comprehensively assess the impact of various data to determine the most urgent tasks to be executed. The AI ​​model determines the real-time data included in the preset emergency information and performs weighted value assignment on the data to comprehensively determine the emergency value corresponding to the preset emergency information of each emergency execution object, so as to realize the quantitative index processing of the preset emergency information based on the emergency value.

[0133] Weighted assignment refers to the process of assigning different weights (numerical values) to multiple influencing factors (such as data, indicators, and information) according to their importance, and then obtaining a comprehensive result through mathematical calculations (such as weighted summation and weighted average). Its core is to make important factors have a greater impact on the final result.

[0134] Based on this, when there are multiple emergency execution objects, an AI model can be used to perform weighted assignment on the real-time data of the preset emergency information of each emergency execution object, thereby obtaining information values ​​of quantitative indicators that can represent the preset emergency information. In order to determine the emergency value corresponding to the preset emergency information based on the weight value and the information value, the corresponding execution emergency value can be obtained based on the sum of the emergency values ​​of each emergency execution object.

[0135] Based on the feasible implementation of S120 described above, this application further provides the following steps before inputting preset emergency information into the AI ​​model, obtaining the information value corresponding to the preset emergency information output by the AI ​​model, and determining the emergency value corresponding to the preset emergency information based on the product of the weight value and the information value corresponding to the preset emergency information:

[0136] The system identifies all real-time data included in the target emergency information and the corresponding target emergency values ​​as sample cases. Multiple sample cases are then input into the AI ​​model to obtain the trained AI model. The AI ​​model is able to perform weighted assignment processing on the data included in the input information and generate the corresponding information values.

[0137] The target urgency information is sample information used for training the AI ​​model. The target urgency value is a pre-determined specific numerical value corresponding to the target urgency information. The target urgency information includes multiple real-time data. The AI ​​model can process the data values ​​of each real-time data according to the given target urgency value and multiple real-time data, and determine the corresponding weight values. In this way, the given target urgency value can be obtained based on the data values ​​and weight values. That is, by determining several target urgency information and corresponding target urgency values, they are used as sample cases for model training. The AI ​​model can then perform weighted value assignment on the sample cases based on its self-learning and adaptive capabilities to obtain the trained AI model. In practical applications, the AI ​​model can perform corresponding weighted value assignment on the data included in the preset urgency information of the object to be executed, so as to obtain the delivery urgency value corresponding to the preset urgency information.

[0138] Based on the feasible implementation of S120 described above, this application further provides a method for determining an execution urgency value based on the state information and attribute information of the object to be executed if the condition is not met, including the following steps:

[0139] If not, input each attribute information and state information into the AI ​​model to obtain the corresponding information value, and add the multiplication of the information value and weight value corresponding to each attribute information to obtain the total emergency value corresponding to all attribute information, and add the multiplication of the information value and weight value corresponding to each state information to obtain the total emergency value corresponding to all state information.

[0140] The emergency value is determined based on the total emergency value corresponding to the attribute information and status information, as well as multiple preset emergency ranges.

[0141] The preset emergency range is a pre-defined range of values ​​corresponding to the total emergency value. For example, it could be a range of 10 to 20 or a range of 20 to 30.

[0142] Therefore, if the state information of multiple objects to be executed with the same target state does not include preset urgency information, meaning there are no urgent execution objects among the objects to be executed, then the AI ​​model can be used to determine the information values ​​corresponding to the attribute information and state information of each object to be executed. These information values ​​are then multiplied by weight values ​​and summed to obtain the total urgency value corresponding to the attribute information and the total urgency value corresponding to the state information of the objects to be executed. This allows for the subsequent determination of the preset urgency range within multiple preset urgency ranges based on the total urgency values ​​corresponding to the attribute information and state information, thereby determining the weight values ​​corresponding to the attribute information and state information. This allows for a further weighted assignment of the total urgency values ​​corresponding to the attribute information and state information to obtain the final execution urgency value, ensuring that the execution urgency value comprehensively considers the attributes and state information. The different weightings of attribute information and status information allow for a better comprehensive assessment of the actual situation of the object to be executed. In practical applications, attribute information represents the static information of the object to be executed, while status information represents its real-time dynamic information. The two correspond to the specific information of the object to be executed under different circumstances. Therefore, in order to better evaluate the actual situation of the object to be executed and to better allocate limited resources to the objects that need to be executed immediately, attribute information and status information need to be assigned separate weights to reflect their differences in importance. Generally, status information can be given a higher weight, thereby increasing the proportion of the total urgency value corresponding to status information in the execution urgency value, so as to better combine with the actual scenario data and thus evaluate the execution priority of each object to be executed.

[0143] Based on the feasible implementation of S120 described above, this application further provides steps including, before determining the execution of an emergency value based on the total emergency value corresponding to the attribute information and the state information, and multiple preset emergency ranges:

[0144] Determine multiple target urgency values ​​and sort them according to their numerical values.

[0145] Based on any two adjacent target urgency values, determine multiple preset urgency ranges and the weight value corresponding to each preset urgency range.

[0146] The target emergency value is a predetermined value used to determine the preset emergency range. For example, the target emergency value can be determined to be 10, 20, or 30. Then the preset emergency range 1 can be 10~20, and the preset emergency range 2 can be 20~30.

[0147] Based on this, by pre-determining multiple target urgency values ​​and sorting them according to their numerical values, a series of target urgency values ​​that satisfy the numerical sorting rules are obtained. Thus, a numerical range is determined based on any two adjacent target urgency values. In order to determine multiple preset urgency ranges based on multiple target urgency values, and further determine the weight value corresponding to each preset urgency range, so as to determine the preset urgency range in which the total urgency value of each object belongs based on the total urgency value corresponding to the attribute information and status information of the object to be executed, thereby determining the weight value corresponding to the preset urgency range.

[0148] Based on the feasible implementation of S120 described above, this application further provides a method for determining an emergency value based on the total emergency value corresponding to the attribute information and the status information, as well as multiple preset emergency ranges, including the following steps:

[0149] Based on multiple preset emergency ranges, the target ranges corresponding to the total emergency values ​​of attribute information and status information are determined respectively; and the weight value of the target range corresponding to the attribute information is determined as the first target weight value, and the weight value of the target range corresponding to the status information is determined as the second target weight value.

[0150] The product of the total urgency value of the attribute information and the first target weight value is determined as the first target multiplier value, and the product of the total urgency value of the status information and the second target weight value is determined as the second target multiplier value. Based on the sum of the first target multiplier value and the second target multiplier value, the execution urgency value is determined.

[0151] Based on this, by determining the preset emergency range corresponding to the total emergency value of the attribute information of the object to be executed, and defining this preset emergency range as the target range, the weight value corresponding to the target range is determined as the first target weight value corresponding to the attribute information. Similarly, the second target weight value corresponding to the state information is determined. Thus, based on the total emergency value and target weight value corresponding to the attribute information and state information, the corresponding multiplication values ​​are obtained and added together to obtain the execution emergency value corresponding to the object to be executed. By determining the weight value corresponding to the preset emergency range, the weight values ​​corresponding to the attribute information and state information can be determined in real time based on the total emergency value determined in each simulation. This solves the limitation of pre-set fixed weight values, improves the flexibility of determining weight values, and makes the determined weight values ​​more in line with the needs of actual scenarios, thereby improving the accuracy of calculation and making the priority determination more accurate.

[0152] Based on the feasible implementation of S120 described above, this application further provides steps for adjusting the AI ​​model according to the index results and feedback results corresponding to the simulation results to obtain the adjusted target AI model, including:

[0153] Determine the simulation results for each indicator, their corresponding preset threshold values, and the feedback results;

[0154] Compare the results of each indicator with the corresponding preset indicator thresholds to obtain the comparison results, and determine the indicator results that are less than the preset indicator thresholds based on the comparison results.

[0155] Based on the results of the indicators to be adjusted and the feedback results, the model parameters of the AI ​​model are adjusted to obtain the target AI model.

[0156] Among them, the preset indicator threshold refers to the preset threshold corresponding to each indicator, which is used to determine whether the corresponding indicator meets the requirements.

[0157] Based on this, by comparing the results of each indicator recorded in the simulation results with the corresponding preset indicator thresholds, the indicator results that are lower than the preset indicator thresholds in each simulation are determined, and the user's feedback results for this simulation are determined. Based on the indicator results to be adjusted and the feedback results, the parameters of the AI ​​model are adjusted accordingly so that the target AI model obtained after adjustment can better meet the actual needs in subsequent simulations.

[0158] In practical applications, the system introduces AI intelligent decision-making + interface to build a decision knowledge base and rule base. It performs AI intelligent decision-making for simulation under multiple states, avoiding the stacking of traditional classic algorithms. Traditional classic algorithms rely on "preset priority rules + strict sorting" (such as "always sorted in a fixed order of "flight type - departure time - delay duration"). However, the AI ​​model of this solution has the ability to solve general problems through deep learning: when the rules do not cover complex scenarios (such as "international flights are delayed but must give way to emergency flights"), the AI ​​model can break through the fixed sorting logic and generate flexible decisions based on the potential rule (implied in the knowledge base) that "emergency flights have the highest priority" and the real-time state.

[0159] Please refer to Figure 2 , Figure 2 A schematic diagram illustrating a flight area simulation method based on an AI model, provided in this application embodiment; as shown. Figure 2As shown, a simulation role is defined by three dimensions: attributes, state, and behavior, which can be further refined locally according to the simulation needs. Attributes are static information describing the simulation role, such as text, numbers, time, coordinates, attachments, etc. State changes with actions and is the core data driving the simulation. Behaviors establish relationships between different roles, continuously activating the simulation process through the initiation, execution, and completion of behaviors until all behaviors are completed.

[0160] Please refer to Figure 3 , Figure 3 A simulation diagram illustrating a flight area simulation method based on an AI model, provided in this application embodiment; as shown. Figure 3 As shown, a logic unit consists of a series of logic activation functions, and the logic result is the computational result of the activation functions based on their states. The system divides logic transport into two types: parallel structure and serial structure. The serial structure represents the logical AND operation, and the parallel structure represents the logical OR operation. A depth-first search algorithm (DFS) is used to compute the logic units. Through a logic computing network, associated logic nodes are distributed to more logic computing units.

[0161] Please refer to Figure 4 , Figure 4 A schematic diagram of a flight area simulation method based on an AI model provided in this application embodiment; as shown Figure 4 As shown, a time- and state-based action activation model (NRG) is constructed. Various behaviors of the simulated object are triggered by time and state. When the triggering conditions are not met, the activation failure (Delay) stage is entered, and the simulation plan is updated according to the activation source.

[0162] Please refer to Figure 5 , Figure 5 An optimized schematic diagram of a flight area simulation method based on an AI model provided in this application embodiment; as shown Figure 5 As shown, users evaluate or provide feedback on the simulation results. When the AI ​​decision-making model receives a good evaluation or positive feedback, the model will continue or strengthen its current output strategy based on the user feedback, and will continue to execute similar strategies in future similar tasks. Conversely, when the AI ​​decision-making model receives a negative evaluation or poor feedback, the model will adjust its internal parameters or strategies based on the feedback content, and will optimize its behavior pattern in the next similar task.

[0163] There are two main methods for flight area operation simulation: 1. Simulating specific scenarios using classical algorithms. Examples include airport operational capacity assessment and runway / taxiway intersection congestion analysis. This involves using simulated operational data and classical algorithms for simulation analysis. The biggest problem with this method is the need to develop corresponding simulation algorithms for specific operations, which generally takes a long time, and the simulation effect depends heavily on the accuracy of the classical algorithms. 2. Using digital twin technology combined with a 3D engine to simulate and extrapolate the overall reality. This method utilizes physical models and particle models within the 3D engine for simulation. The advantage of this method is better visualization, but the biggest problem is the enormous computational load resulting from using a general physical simulation model. When the simulation object reaches a certain scale, it places higher demands on hardware. Furthermore, when further refining a specific simulation element is required, an overall improvement in accuracy is often necessary. However, actual needs are generally focused and not comprehensive and detailed.

[0164] In this embodiment, to avoid the limitations of traditional classical algorithms and address issues such as simulation cycle, efficiency, and effectiveness, the simulation can be infinitely refined for key concerns, achieving local refinement rather than global refinement. This reduces the simulation requirements on underlying devices and enables simulation service orchestration, providing a clear visualization of the simulation thought chain. By modeling related objects to be executed through the pending services, the corresponding service states and preset state conditions are determined. Based on the logical rules corresponding to the pending services, it is determined whether the real-time state of the objects to be executed meets the preset conditions. If it does, the corresponding target state is determined as the service state, meaning the objects to be executed can be used for execution. The pending business logic can further include both sequential and parallel rules. After determining the target state of the objects to be executed as the business state, the target states of each object are compared to identify those with the same target state. Given limited resources, the execution priority of each object needs to be ranked to prioritize the highest-priority object. Specifically, this can be achieved by constructing a rule base and a knowledge base, allowing the AI ​​model to determine preset emergency information. This preset emergency information represents an emergency scenario; objects containing preset emergency information should be executed first. If only one of the multiple objects contains such information... For urgent execution objects containing preset urgency information, the highest execution priority is determined. If there are multiple urgent execution objects, the most urgent first execution object needs to be identified. At this point, the preset urgency information contained in each urgent execution object can be determined. The AI ​​model can output the corresponding information value and weight value of the preset urgency information based on the real-time data corresponding to the input preset urgency information. The AI ​​model has self-learning and adaptive capabilities. Through deep learning, significant improvements have been made in efficiency, resource requirements, and general problem-solving capabilities. During the model training process, the AI ​​model has already undergone… Multiple preset emergency information items are sorted by urgency, and corresponding weight values ​​are output. The model is trained by weighting the data contained in the emergency information based on sample cases composed of multiple given target emergency information items and target emergency values. Therefore, the AI ​​model trained by weighting and weighting the data corresponding to the emergency information in the sample cases can output corresponding information values ​​and weight values ​​based on the input information. Thus, based on the information values ​​and weight values ​​of the preset emergency information output by the AI ​​model, the emergency value of the preset emergency information is determined. Based on the numerical value of the emergency value of each emergency execution object, the emergency execution object with the largest value is determined to have the highest priority.

[0165] For multiple execution objects without preset urgency information (i.e., multiple execution objects without urgent execution objects), the AI ​​model can generate information values ​​and weight values ​​corresponding to the attribute information and state information of each execution object, thereby determining the total urgency value corresponding to the attribute information and the total urgency value corresponding to the state information of each execution object. This allows for further overall weighting of the attribute information and state information. Specifically, by determining the weight value of the preset urgency range corresponding to the total urgency value, the weight values ​​and total urgency value corresponding to the attribute information and state information are determined, thus obtaining the execution urgency value of the execution object. Based on the magnitude of the execution urgency value, the execution object with the highest execution priority is determined, thereby achieving simulation. Based on the simulation results, various indicator results are determined, and by comparison, indicators below preset indicator thresholds that require adjustment are identified, along with user feedback on the simulation results. Based on the indicators requiring adjustment and the feedback results, the AI ​​model is adjusted accordingly to obtain the target AI model.

[0166] In this way, prominent issues in flight area operation simulation, such as simulation cycle, efficiency, local refinement, and simulation logic arrangement, are resolved, effectively improving the speed and accuracy of flight area operation simulation. Based on a knowledge base and rule base, intelligent decision-making under multiple states is achieved through the Qwen3-32b large model. The entire simulation process is simulated through dialogue API, decision API, and diagnostic API, improving the simulation efficiency under the general model. The simulation process is visualized and arranged through simulation intelligent agent modeling and action activation model (NRG), and can be continuously refined to achieve local refined modeling of the business scenarios of concern. Through the analysis of simulation index data such as resource demand, resource planning, and resource request, an evaluation index model is established for the simulation status of vehicles, personnel, and equipment in the flight area, guiding the rational planning and scheduling of resources, avoiding problems such as blind investment and potential operational risks, and improving the efficiency and level of support.

[0167] Figure 6 This is a schematic diagram of the structure of a flight area simulation device 600 based on an AI model, provided as an embodiment of this application. Figure 6 As shown, this AI model-based flight area simulation device 600 includes: a modeling module 610, a state determination module 620, a simulation module 630, and an adjustment module 640; wherein:

[0168] The modeling module 610 is used to model each object to be executed based on the business to be executed, and to determine the corresponding business state and preset state conditions.

[0169] The state determination module 620 is used to determine the target state of the object to be executed based on the real-time state and logical rules of the modeled object to be executed, as well as the business state and preset state conditions. For objects to be executed with the same target state, the execution urgency value of the object to be executed is determined based on the AI ​​model, the state information, attribute information and preset urgency information of each object to be executed.

[0170] The simulation module 630 is used to determine the priority of each object to be executed based on the execution urgency value, and to simulate the object to be executed with the highest priority to obtain the simulation results;

[0171] The adjustment module 640 is used to adjust the AI ​​model based on the index results and feedback results corresponding to the simulation results, so as to obtain the adjusted target AI model.

[0172] In this embodiment of the application, the state determination module 620 can also be specifically used for:

[0173] Determine the logical rules corresponding to each object to be executed, and the preset state conditions corresponding to each real-time state;

[0174] When the logic rule is a serial rule, determine whether each real-time state satisfies the corresponding preset state condition.

[0175] When the logical rule is a parallel rule, determine whether there is at least one real-time state that satisfies the corresponding preset state condition.

[0176] If so, then the target state of the object to be executed is determined to be the business state;

[0177] If not, then it is determined that the target state of the object to be executed does not exist;

[0178] For objects to be executed with the same target state, the execution urgency value of the objects to be executed is determined based on the AI ​​model, the state information, attribute information and preset urgency information of each object to be executed.

[0179] In this embodiment of the application, the state determination module 620 can also be specifically used for:

[0180] Compare the target states of each object to be executed to see if they are the same, and obtain the comparison results;

[0181] Based on the comparison results, determine the objects to be executed that have the same target state;

[0182] For each object to be executed, the execution urgency value is determined based on the AI ​​model, the status information, attribute information, and preset urgency information of each object to be executed.

[0183] In this embodiment of the application, the state determination module 620 can also be specifically used for:

[0184] Determine the preset emergency information and check whether the status information of the object to be executed includes the preset emergency information;

[0185] If so, the object to be executed is determined to be an urgent object to be executed, and the weight value corresponding to the preset urgent information included in the status information is determined.

[0186] Based on the AI ​​model, the weight values ​​corresponding to the urgent objects to be executed, and the preset urgency information, the execution urgency value of the urgent objects to be executed is determined.

[0187] If not, the execution urgency value is determined based on the status and attribute information of the object to be executed.

[0188] In this embodiment of the application, the state determination module 620 can also be specifically used for:

[0189] Multiple preset emergency messages are identified, and the urgency of each preset emergency message is ranked based on user feedback.

[0190] Input the sorted preset emergency information into the AI ​​model, and obtain the weight value assigned by the AI ​​model to each preset emergency information.

[0191] In this embodiment of the application, the state determination module 620 can also be specifically used for:

[0192] Based on web scraping, several preset emergency messages were identified;

[0193] Based on the feedback, the urgency of each preset emergency message is sorted.

[0194] In this embodiment of the application, the state determination module 620 can also be specifically used for:

[0195] Determine the weight values ​​corresponding to preset emergency information;

[0196] Input preset emergency information into the AI ​​model, obtain the information value corresponding to the preset emergency information output by the AI ​​model, and determine the emergency value corresponding to the preset emergency information based on the product of the weight value and the information value. The information value is the overall value of the information output by the AI ​​model after weighting and assigning values ​​to the various data included in the input information, which is used to characterize the numerical quantification index of the information.

[0197] The execution emergency value is determined by the sum of all emergency values ​​corresponding to the urgent pending objects.

[0198] In this embodiment of the application, the state determination module 620 can also be specifically used for:

[0199] The system identifies all real-time data included in the target emergency information and the corresponding target emergency values ​​as sample cases. Multiple sample cases are then input into the AI ​​model to obtain the trained AI model. The AI ​​model is able to perform weighted assignment processing on the data included in the input information and generate the corresponding information values.

[0200] In this embodiment of the application, the state determination module 620 can also be specifically used for:

[0201] If not, input each attribute information and state information into the AI ​​model to obtain the corresponding information value, and add the multiplication of the information value and weight value corresponding to each attribute information to obtain the total emergency value corresponding to all attribute information, and add the multiplication of the information value and weight value corresponding to each state information to obtain the total emergency value corresponding to all state information.

[0202] The emergency value is determined based on the total emergency value corresponding to the attribute information and status information, as well as multiple preset emergency ranges.

[0203] In this embodiment of the application, the state determination module 620 can also be specifically used for:

[0204] Determine multiple target urgency values ​​and sort them according to their numerical values.

[0205] Based on any two adjacent target urgency values, determine multiple preset urgency ranges and the weight value corresponding to each preset urgency range.

[0206] In this embodiment of the application, the state determination module 620 can also be specifically used for:

[0207] Based on multiple preset emergency ranges, the target ranges corresponding to the total emergency values ​​of attribute information and status information are determined respectively; and the weight value of the target range corresponding to the attribute information is determined as the first target weight value, and the weight value of the target range corresponding to the status information is determined as the second target weight value.

[0208] The product of the total urgency value of the attribute information and the first target weight value is determined as the first target multiplier value, and the product of the total urgency value of the status information and the second target weight value is determined as the second target multiplier value. Based on the sum of the first target multiplier value and the second target multiplier value, the execution urgency value is determined.

[0209] In this embodiment of the application, the adjustment module 640 can also be specifically used for:

[0210] Determine the simulation results for each indicator, their corresponding preset threshold values, and the feedback results;

[0211] Compare the results of each indicator with the corresponding preset indicator thresholds to obtain the comparison results, and determine the indicator results that are less than the preset indicator thresholds based on the comparison results.

[0212] Based on the results of the indicators to be adjusted and the feedback results, the model parameters of the AI ​​model are adjusted to obtain the target AI model.

[0213] Figure 7 This is a schematic diagram of the structure of a device for performing a flight area simulation method based on an AI model according to an embodiment of this application. Figure 7 As shown, the device 700 includes:

[0214] The device 700 may include a processor 701 with one or more processing cores, a memory 702 with one or more computer-readable storage media, a communication component 703, and other components. The processor 701, memory 702, and communication component 703 are connected via a bus 704.

[0215] In the specific implementation process, at least one processor 701 executes computer execution instructions stored in memory 702, causing at least one processor 701 to execute the above-mentioned flight area simulation method based on an AI model.

[0216] The specific implementation process of processor 701 can be found in the above method embodiments, and its implementation principle and technical effect are similar. It will not be repeated here.

[0217] Furthermore, the processor can be a Central Processing Unit (CPU), or other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), etc. A general-purpose processor can be a microprocessor or any conventional processor. The steps of the method disclosed in this application can be directly implemented by a hardware processor, or implemented by a combination of hardware and software modules within the processor.

[0218] The memory may include random access memory (RAM) and may also include non-volatile memory (NVM), such as at least one disk storage device.

[0219] The bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. Buses can be categorized as address buses, data buses, control buses, etc. For ease of illustration, the buses shown in the accompanying drawings are not limited to a single bus or a single type of bus.

[0220] In some embodiments, a computer program product is also proposed, including a computer program or instructions that, when executed by a processor, implement the steps in any of the above-described AI model-based flight area simulation methods.

[0221] For details on the implementation of each of the above operations, please refer to the previous examples, which will not be repeated here.

[0222] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be performed by instructions, or by instructions controlling related hardware. These instructions can be stored in a computer-readable storage medium and loaded and executed by a processor.

[0223] Therefore, embodiments of this application provide a computer-readable storage medium storing multiple lines of program code that can be loaded by a processor to execute steps in any of the AI ​​model-based flight area simulation methods provided in embodiments of this application.

[0224] The storage medium may include: read-only memory (ROM), random access memory (RAM), disk or optical disk, etc.

[0225] According to one aspect of this application, a computer program product or computer program is provided, the computer program product or computer program including computer instructions stored in a computer-readable storage medium.

[0226] Since the instructions stored in the storage medium can execute the steps in any of the AI ​​model-based flight area simulation methods provided in the embodiments of this application, the beneficial effects that any of the AI ​​model-based flight area simulation methods provided in the embodiments of this application can achieve can be realized. For details, please refer to the previous embodiments, which will not be repeated here.

[0227] Other embodiments of this application will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of this application that follow the general principles of this application and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of this application are indicated by the appended claims.

[0228] It should be understood that this application is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope.

Claims

1. A flight area simulation method based on an AI model, characterized in that, The method includes: Based on the business to be executed, model each object to be executed, and determine the corresponding business status and preset status conditions; Based on the real-time state and logical rules of the objects to be executed after modeling, as well as the business state and the preset state conditions, the target state of the objects to be executed is determined. For the objects to be executed with the same target state, the execution urgency value of the objects to be executed is determined based on the AI ​​model, the state information, attribute information and preset urgency information of each object to be executed. This includes determining the logical rules corresponding to each object to be executed and the preset state conditions corresponding to each real-time state. When the logic rule is a serial rule, it is determined whether each of the real-time states satisfies the corresponding preset state condition. When the logical rule is a parallel rule, determine whether there is at least one real-time state that satisfies the corresponding preset state condition; If so, then the target state of the object to be executed is determined to be the business state; If not, then it is determined that the target state of the object to be executed does not exist; For the objects to be executed that have the same target state, the execution urgency value of the objects to be executed is determined based on the AI ​​model, the state information, attribute information, and preset urgency information of each object to be executed; Specifically, determining the execution urgency value of each execution object with the same target state, based on the AI ​​model, the state information, attribute information, and preset urgency information of each execution object, includes: Compare the target states of each of the objects to be executed to see if they are the same, and obtain the comparison results; Based on the comparison results, determine the objects to be executed that have the same target state among the objects to be executed; For the object to be executed, the execution urgency value of the object to be executed is determined based on the AI ​​model, the state information, attribute information, and preset urgency information of each object to be executed; Specifically, determining the execution urgency value of the object to be executed based on the AI ​​model, the state information, attribute information, and preset urgency information of each object to be executed includes: Determine the preset emergency information, and determine whether the status information of the object to be executed includes the preset emergency information; If so, the object to be executed is determined to be an urgent object to be executed, and the weight value corresponding to the preset urgent information included in the status information is determined; The execution urgency value of the urgent object to be executed is determined based on the AI ​​model, the weight value corresponding to the urgent object to be executed, and the preset urgency information. If not, then the execution urgency value is determined based on the status information and attribute information of the object to be executed; The step of determining the execution urgency value of the urgent execution object based on the AI ​​model, the weight value corresponding to the urgent execution object, and the preset urgency information includes: Determine the weight value corresponding to the preset emergency information; The preset emergency information is input into the AI ​​model to obtain the information value corresponding to the preset emergency information output by the AI ​​model. The emergency value corresponding to the preset emergency information is determined according to the product of the weight value and the information value. The information value is the overall value corresponding to the information output by the AI ​​model after weighting and assigning values ​​to the various data included in the input information, and is used to characterize the numerical quantification index corresponding to the information. The execution urgency value is determined based on the sum of all urgency values ​​corresponding to the urgent pending objects; Based on the execution urgency value, the priority of each of the objects to be executed is determined, and the object with the highest priority is simulated to obtain the simulation result; Based on the corresponding index results and feedback results of the simulation results, the AI ​​model is adjusted to obtain the adjusted target AI model.

2. The method according to claim 1, characterized in that, If so, before determining that the object to be executed is an urgent object to be executed, and before determining the weight value corresponding to the preset urgent information included in the status information, the method further includes: Multiple preset emergency messages are identified, and the urgency of each preset emergency message is ranked according to the user's feedback. The sorted preset emergency information is input into the AI ​​model to obtain the weight value assigned by the AI ​​model to each preset emergency information.

3. The method according to claim 2, characterized in that, The step of determining multiple preset emergency messages and ranking the urgency of each preset emergency message according to user feedback includes: Based on web crawling, multiple preset emergency information messages were determined; Based on the feedback operation, the urgency of each preset emergency message is sorted.

4. The method according to claim 1, characterized in that, Before inputting the preset emergency information into the AI ​​model, obtaining the information value corresponding to the preset emergency information output by the AI ​​model, and determining the emergency value corresponding to the preset emergency information based on the product of the weight value and the information value, the method further includes: The system identifies all real-time data included in the target emergency information and the corresponding target emergency value as sample cases, and inputs multiple sample cases into the AI ​​model to obtain the trained AI model. The AI ​​model can perform weighted assignment processing on the data included in the input information to generate the corresponding information value.

5. The method according to claim 1, characterized in that, If not, then based on the state information and attribute information of the object to be executed, the execution urgency value is determined, including: If not, input each attribute information and state information into the AI ​​model to obtain the corresponding information value, and add the multiplication of the information value and weight value corresponding to each attribute information to obtain the total emergency value corresponding to all the attribute information, and add the multiplication of the information value and weight value corresponding to each state information to obtain the total emergency value corresponding to all the state information. The execution emergency value is determined based on the total emergency value corresponding to the attribute information and the status information, as well as multiple preset emergency ranges.

6. The method according to claim 5, characterized in that, Before determining the execution emergency value based on the total emergency value corresponding to the attribute information and the status information, and multiple preset emergency ranges, the method further includes: Multiple target urgency values ​​are determined, and the multiple target urgency values ​​are sorted according to their numerical values; Based on any two adjacent target urgency values, determine a plurality of corresponding preset urgency ranges, and determine the weight value corresponding to each preset urgency range.

7. The method according to claim 5, characterized in that, The step of determining the execution emergency value based on the total emergency value corresponding to the attribute information and the status information, and multiple preset emergency ranges, includes: Based on multiple preset emergency ranges, target ranges corresponding to the total emergency values ​​of the attribute information and the status information are determined respectively; and the weight value of the target range corresponding to the attribute information is determined to be a first target weight value, and the weight value of the target range corresponding to the status information is determined to be a second target weight value. The product of the total urgency value of the attribute information and the first target weight value is determined as the first target multiplier value, and the product of the total urgency value of the status information and the second target weight value is determined as the second target multiplier value. The execution urgency value is determined based on the sum of the first target multiplier value and the second target multiplier value.

8. The method according to claim 1, characterized in that, The step of adjusting the AI ​​model based on the corresponding indicator results and feedback results of the simulation results to obtain the adjusted target AI model includes: Determine the results of each indicator corresponding to the simulation results, their corresponding preset indicator thresholds, and the feedback results; Compare the results of each indicator with the corresponding preset indicator threshold to obtain a comparison result, and determine the indicator results that are less than the preset indicator threshold based on the comparison result; Based on the results of the indicators to be adjusted and the feedback results, the model parameters of the AI ​​model are adjusted to obtain the target AI model.

9. A flight area simulation device based on an AI model, characterized in that, The device includes: The modeling module is used to model each object to be executed based on the business to be executed, and to determine the corresponding business state and preset state conditions. The state determination module is used to determine the target state of the object to be executed based on the real-time state and logical rules of the modeled object to be executed, as well as the business state and the preset state conditions. For the objects to be executed with the same target state, the module determines the execution urgency value of the object to be executed based on the AI ​​model, the state information, attribute information and preset urgency information of each object to be executed. This includes determining the logical rules corresponding to each object to be executed and the preset state conditions corresponding to each real-time state. When the logic rule is a serial rule, it is determined whether each of the real-time states satisfies the corresponding preset state condition. When the logical rule is a parallel rule, determine whether there is at least one real-time state that satisfies the corresponding preset state condition; If so, then the target state of the object to be executed is determined to be the business state; If not, then it is determined that the target state of the object to be executed does not exist; For the objects to be executed that have the same target state, the execution urgency value of the objects to be executed is determined based on the AI ​​model, the state information, attribute information, and preset urgency information of each object to be executed; Specifically, determining the execution urgency value of each execution object with the same target state, based on the AI ​​model, the state information, attribute information, and preset urgency information of each execution object, includes: Compare the target states of each of the objects to be executed to see if they are the same, and obtain the comparison results; Based on the comparison results, determine the objects to be executed that have the same target state among the objects to be executed; For the object to be executed, the execution urgency value of the object to be executed is determined based on the AI ​​model, the state information, attribute information, and preset urgency information of each object to be executed; Specifically, determining the execution urgency value of the object to be executed based on the AI ​​model, the state information, attribute information, and preset urgency information of each object to be executed includes: Determine the preset emergency information, and determine whether the status information of the object to be executed includes the preset emergency information; If so, the object to be executed is determined to be an urgent object to be executed, and the weight value corresponding to the preset urgent information included in the status information is determined; The execution urgency value of the urgent object to be executed is determined based on the AI ​​model, the weight value corresponding to the urgent object to be executed, and the preset urgency information. If not, then the execution urgency value is determined based on the status information and attribute information of the object to be executed; The step of determining the execution urgency value of the urgent execution object based on the AI ​​model, the weight value corresponding to the urgent execution object, and the preset urgency information includes: Determine the weight value corresponding to the preset emergency information; The preset emergency information is input into the AI ​​model to obtain the information value corresponding to the preset emergency information output by the AI ​​model. The emergency value corresponding to the preset emergency information is determined according to the product of the weight value and the information value. The information value is the overall value corresponding to the information output by the AI ​​model after weighting and assigning values ​​to the various data included in the input information, and is used to characterize the numerical quantification index corresponding to the information. The execution urgency value is determined based on the sum of all urgency values ​​corresponding to the urgent pending objects; The simulation module is used to determine the priority of each of the objects to be executed based on the execution urgency value, and to simulate the object to be executed with the highest priority to obtain the simulation result; The adjustment module is used to adjust the AI ​​model based on the index results and feedback results corresponding to the simulation results, so as to obtain the adjusted target AI model.

10. A flight area simulation device based on an AI model, characterized in that, include: One or more processors; Memory; One or more programs, wherein the one or more programs are stored in memory and configured to be executed by one or more processors, the one or more programs being configured to perform the method as described in any one of claims 1 to 8.

11. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores program code that can be called by a processor to perform the method as described in any one of claims 1 to 8.

Images