A scheduling method, device and storage medium for semi-physical simulation resources

Abstract

This application relates to the field of hardware-in-the-loop simulation, and particularly to a scheduling method, device, and storage medium for hardware-in-the-loop simulation resources. The method includes: obtaining a number of simulation models, a number of simulation devices, and multiple test tasks in the hardware-in-the-loop simulation resource library; combining the simulation models with different simulation devices to obtain hardware-in-the-loop simulation nodes, and generating multiple collaboration combinations to collaboratively execute multiple test tasks to obtain performance test data for each simulation model; analyzing the performance test data of each simulation model to generate a performance parameter set for each simulation model, where the performance parameter set includes a number of scenario-sensitive indicators and a number of static basic indicators; based on the simulation model requirements and the performance parameter set of each simulation model, selecting target simulation models and target simulation devices from the hardware-in-the-loop simulation resource library to execute target simulation tasks. It realizes a high degree of adaptation of the scheduling strategy to the actual operating environment, improving resource utilization and scheduling accuracy.

Description

Technical Field

[0001] This application relates to the field of hardware-in-the-loop simulation, and in particular, to a scheduling method, device, and storage medium for hardware-in-the-loop simulation resources. Background Art

[0002] Hardware-in-the-Loop (HIL) technology provides an efficient and reliable means for the development, testing, and verification of complex systems by organically combining physical devices with digital simulation models. Its application scenarios cover fields such as aerospace, automotive electronics, power systems, and industrial automation, and can significantly reduce R & D costs and shorten the product launch cycle. In the HIL framework, if advanced resource scheduling technology can be combined, the efficiency and effectiveness of simulation can be further improved.

[0003] For example, the invention patent application with the publication number CN116774978A discloses a service-oriented reusable framework and integration method for simulation models. The service-oriented reusable framework for simulation models includes a resource layer, a service component layer, a service layer, and an application layer; the resource layer is used to solve functions such as resource description, organization, and management; the service component layer is used to encapsulate models at different levels, in different fields, of different categories, and with different granularities in the form of components; the service component layer mainly includes component description, component definition, component construction tools, component verification, component configuration, component installation, component reconstruction, and component encapsulation, and its main task is to complete model description, reconstruction, and assembly; the service layer is used to implement the encapsulation of models in the form of cloud services, so as to be able to shield the heterogeneity between simulation service models at different levels, with different granularities, and in different fields. Another example is the invention patent application with the publication number CN119088562A, which discloses a simulation resource pool management and scheduling method and system, a simulation method and system. The simulation resource pool management and scheduling method includes: constructing a device resource pool and a model resource pool, where the device resource pool is configured with relevant information of multiple simulation devices, and the model resource pool is configured with multiple model running state information; establishing a simulation task, and obtaining the running state information of the model corresponding to the simulation task from the model resource pool; calculating the priority value of the matching between the simulation devices in the device resource pool and the model, and based on the optimal priority value, retrieving the simulation devices and allocating the retrieved simulation devices to the corresponding simulation tasks.

[0004] However, in order to achieve standardized management, existing resource scheduling technologies often use unified data encapsulation, which is difficult to meet the joint scheduling requirements of simulation models and physical devices in hardware-in-the-loop simulation, resulting in insufficient resource dynamic allocation and elastic scheduling capabilities, and restricting the efficiency and scalability of the simulation system. Summary of the Invention

[0005] The main purpose of this application is to provide a scheduling method, device, and storage medium for semi-physical simulation resources. To solve the above-mentioned technical problems, this application specifically adopts the following technical solutions:

[0006] In the first aspect of this application, a scheduling method for semi-physical simulation resources is provided. The method includes:

[0007] S101. Obtain a semi-physical simulation resource library and test tasks under multiple different task scenarios. The semi-physical simulation resource library includes several simulation models and several simulation devices;

[0008] S102. Combine each simulation model with different simulation devices to obtain multiple semi-physical simulation nodes. Among them, the simulation models or simulation devices are different among the multiple semi-physical simulation nodes;

[0009] S103. Generate multiple collaboration combinations based on the multiple semi-physical simulation nodes to collaboratively execute multiple test tasks, and obtain performance test data of each simulation model under different simulation devices, different collaboration combinations, and different task scenarios;

[0010] S104. Analyze the performance test data of each simulation model to generate a performance parameter set for each simulation model. The performance parameter set includes several scenario-sensitive indicators and several static basic indicators;

[0011] S105. Based on the task scenario and simulation model requirements of the target simulation task, select a target simulation model and a target simulation device from the semi-physical simulation resource library according to the simulation model requirements and the performance parameter set of each simulation model to execute the target simulation task.

[0012] In some embodiments, the performance test data includes several evaluation indicators; S104 includes: comparing multiple values of the same evaluation indicator in the performance test data of each simulation model to determine the value fluctuation; when the value fluctuation is greater than a preset fluctuation value, determine the corresponding evaluation indicator as the scenario-sensitive indicator, and associate and store different values of the scenario-sensitive indicator with the task scenario and / or simulation device and / or collaboration combination; when the value fluctuation is less than the preset fluctuation value, determine the corresponding evaluation indicator as the static basic indicator.

[0013] In some embodiments, the method further includes: analyzing the change trend of several values of the scenario-sensitive indicator, respectively determining the influence coefficients of the task scenario, simulation device, and collaboration combination on the scenario-sensitive indicator; determining the influencing factors of the scenario-sensitive indicator from the task scenario, simulation device, and collaboration combination according to the influence coefficients, and associating and storing the influencing factors with the scenario-sensitive indicator.

[0014] In some embodiments, the method further includes: storing the static basic metrics as fixed values or range intervals, where the range intervals are determined based on multiple values of the static basic metrics.

[0015] In some embodiments, the simulation model requirements include performance requirements, and the 105 further includes: screening a number of unused simulation models based on the performance requirements and the static basic metrics of the simulation model to obtain a number of alternative simulation models; generating multiple alternative simulation nodes based on the number of alternative simulation models and a number of alternative simulation devices that are not used in the hardware-in-the-loop simulation resource library; and generating multiple alternative cooperation combinations based on the multiple alternative simulation nodes; evaluating the prediction performance and resource occupancy rate of each alternative cooperation combination in the task scenario of the target simulation task based on the scenario-sensitive metrics of the alternative simulation models; selecting the alternative cooperation combination with the prediction performance meeting the performance requirements and the lowest resource occupancy rate as the target cooperation combination, and using the alternative simulation models and alternative simulation devices within the target cooperation combination as the target simulation model and the target simulation device.

[0016] In some embodiments, before the S105, it further includes: obtaining the priority of the simulation task to be executed, and using at least one simulation task with the highest priority as the target simulation task.

[0017] In some embodiments, the method further includes: loading the target simulation model into the corresponding target simulation device to obtain a number of target hardware-in-the-loop simulation nodes; continuously monitoring the task execution quality of the number of target hardware-in-the-loop simulation nodes during the execution of the target simulation task; when the task execution quality is lower than the preset quality, obtaining the measured performance data of the target hardware-in-the-loop simulation node; and isolating the simulation data generated by the target hardware-in-the-loop simulation node if the difference between the measured performance data and the performance test data of the target hardware-in-the-loop simulation node is greater than the preset difference.

[0018] In some embodiments, the method further includes: if the difference between the measured performance data and the performance test data of the target hardware-in-the-loop simulation node is less than the preset difference, selecting a preferred simulation device from a number of simulation devices that are not used in the hardware-in-the-loop simulation resource library based on the scenario-sensitive metrics of the target simulation model; loading the target simulation model into the preferred simulation device to replace the target simulation device, and updating the target hardware-in-the-loop simulation node.

[0019] A second aspect of the present application lies in providing a computer device, the device includes:

[0020] a memory for storing a computer program;

[0021] A processor for executing the computer program and realizing the steps of the method for scheduling semi-physical simulation resources provided in any embodiment of the present application when executing the computer program.

[0022] In a third aspect of the present application, there is also correspondingly provided a computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the steps of the method for scheduling semi-physical simulation resources provided in any embodiment of the present application.

[0023] Beneficial effects:

[0024] The embodiments of the present application provide a method, device, and storage medium for scheduling semi-physical simulation resources. Through diverse heterogeneous resource test combinations, a dynamic-static separation index system is established for the performance parameters of each simulation model, and the model performance is divided into scenario-sensitive indicators and static basic indicators and stored in a customized manner to manifest the performance boundaries and dynamic influence relationships of the simulation model in different environments. Thus, the strong coupling between semi-physical simulation resources is transformed into an advantage for improving collaboration efficiency, enabling it to be highly adapted to the actual operating environment based on multi-dimensional analysis of the performance parameter set, and improving resource utilization and scheduling accuracy.

[0025] Specifically, by cross-combining simulation models with different simulation devices to generate diverse simulation nodes and executing test tasks in a multi-task scenario, complex scenarios under different test conditions (such as task scenario types, simulation device models, and collaboration combination configurations) are covered, thereby capturing the collaborative efficiency and implicit dependency relationships between digital models and physical hardware, as well as different model combinations. Further, by comparing the performance parameter sets, the indicators significantly affected are identified, the dynamic and static indicators are separated, and the dynamic indicators are associated and stored with their key influencing factors, breaking the "one-size-fits-all" data encapsulation method and providing a basis for dynamic scheduling.

[0026] Furthermore, at the initial stage of task matching, models that do not meet the basic capability requirements are quickly filtered through static indicators, reducing the computational complexity of subsequent dynamic evaluations. Then, the real-time impact of task scenarios, device characteristics, and collaboration modes on model performance is quantified through scenario-sensitive indicators, and a resource scheduling is achieved based on a multi-dimensional (simulation performance, resource occupancy) layer-by-layer screening and overall screening mechanism to cope with performance degradation or complementary effects in a strongly coupled environment, predict and avoid performance traps, break through the limitations of single-resource performance evaluation, and optimize the dynamic response ability of resource scheduling to complex environments. Description of the Drawings

[0027] To more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. In all the drawings, similar elements or parts are generally identified by similar reference numerals. In the drawings, the elements or parts do not necessarily draw in actual proportion. Obviously, the following described drawings are some embodiments of the present application. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained according to these drawings.

[0028] Figure 1 is a schematic flowchart of a method for scheduling semi-physical simulation resources provided by an embodiment of the present application;

[0029] Figure 2 is a schematic diagram of the encapsulation of a semi-physical simulation resource library provided by an embodiment of the present application;

[0030] Figure 3 is a schematic block diagram of the structure of a computer device provided by an embodiment of the present application. Detailed implementation manners

[0031] To make the objectives, technical solutions and advantages of the embodiments of the present application clearer, the following will clearly and completely describe the technical solutions in the embodiments of the present application with reference to the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are some, but not all, of the embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present application.

[0032] In this document, suffixes such as "module", "component" or "unit" used to represent elements are only for the convenience of the description of the present application, and have no specific meaning in itself. Therefore, "module", "component" or "unit" can be used interchangeably.

[0033] In this document, terms such as "upper", "lower", "inner", "outer", "front", "rear", "one end", "the other end", etc. indicate the orientation or positional relationship based on the orientation or positional relationship shown in the drawings. It is only for the convenience of describing the present application and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be construed as a limitation of the present application. In addition, the terms "first" and "second" are only used for descriptive purposes and cannot be construed as indicating or implying relative importance.

[0034] In this document, "and / or" includes any and all combinations of one or more of the listed related items.

[0035] As used herein, "a plurality of" means two or more, i.e., it includes two, three, four, five, etc.

[0036] It should be noted that, as used herein, the terms "comprising", "including" or any other variants thereof are intended to cover non-exclusive inclusion, such that a process, method, article or device comprising a series of elements not only includes those elements but also includes other elements not expressly listed, or further includes elements inherent to such process, method, article or device. Without further limitation, an element defined by the statement "comprising one..." does not exclude the presence of additional identical elements in the process, method, article or device comprising that element.

[0037] Hardware-in-the-loop simulation is a technology that combines actual physical hardware with computer-generated digital models for system testing. It verifies the performance and reliability of the system by running real hardware and virtual models in a closed-loop environment.

[0038] As used herein, a simulation model refers to a mathematical or logical model used to simulate system behavior, i.e., the virtual model in hardware-in-the-loop simulation. These models can be dynamic models based on physical laws, control system algorithms, or signal processing flows, etc. They are usually implemented in simulation software and can respond to input data and generate output results.

[0039] As used herein, simulation devices refer to those actual physical components directly involved in the experimental process, such as sensors, actuators, controllers, etc., i.e., the real hardware devices in hardware-in-the-loop simulation. These hardware devices receive data from the simulation model as input and convert it into real physical actions or state changes; at the same time, they can also feedback the results of actual operations to the simulation model.

[0040] In a hardware-in-the-loop simulation system, there is a close connection between the simulation model and the simulation device. On the one hand, the simulation model provides control instructions or environmental parameters for the physical device; on the other hand, the physical device performs corresponding operations according to the received information and feedbacks the results to the simulation model to update its state. This interaction requires a high degree of consistency and synchronization between the two. The strong coupling between the simulation model and the physical device also leads to their mutual influence. Moreover, as the system scale expands and the application scenarios become diversified, simulation models in different fields often need to work together jointly to form a heterogeneous model co-simulation, resulting in mutual influence between the models. For example, a drone model completes an observation task and a radar model completes a detection task. The two simulation models cooperate with each other to complete the task, and the performance of the two simulation models has a coupling relationship.

[0041] Existing resource scheduling technologies tend to adopt a unified data encapsulation method to standardize data formats for easy storage and retrieval. However, they expose serious limitations when facing cross-device and cross-domain scenarios. Due to the lack of a flexible adaptation mechanism, different types of data are difficult to seamlessly connect, easily forming information silos. Even with a rich resource library, it is impossible to accurately achieve dynamic resource allocation and elastic scheduling, restricting the efficiency and scalability of the simulation system.

[0042] Based on this, the embodiments of this application provide a scheduling method, device, and storage medium for semi-physical simulation resources. Through diverse heterogeneous resource test combinations, a static-dynamic separation index system is established for the performance parameters of each simulation model. The model performance is divided into scenario-sensitive indicators and static basic indicators and stored in a customized manner to manifest the performance boundaries and dynamic influence relationships of the simulation model in different environments. Among them, the static basic indicators quantify the inherent performance of the model and are used to screen simulation models with qualified basic capabilities. The scenario-sensitive indicators dynamically reflect the performance fluctuations and adaptability of the model in the actual deployment environment and are used for performance evaluation of joint simulation. Thus, the strong coupling between semi-physical simulation resources is transformed into an advantage for improving collaborative efficiency, enabling it to flexibly match task requirements with resource characteristics based on multi-dimensional analysis of the performance parameter set, and predicting and avoiding performance pitfalls.

[0043] The following will describe in detail some embodiments of this application with reference to the accompanying drawings. Without conflict, the following embodiments and the features in the embodiments can be combined with each other. Please refer to Figure 1 , Figure 1 is a schematic flowchart of a scheduling method for semi-physical simulation resources provided by the embodiments of this application. As Figure 1 shown, the embodiments of this application provide a scheduling method for semi-physical simulation resources, and the method includes S101 to S105.

[0044] S101. Obtain a semi-physical simulation resource library and test tasks under multiple different task scenarios. The semi-physical simulation resource library includes several simulation models and several simulation devices.

[0045] Among them, the semi-physical simulation resource library refers to a comprehensive resource pool storing various simulation models (such as aircraft models, radar models, UAV models, satellite models, etc.) and various simulation devices (such as real-time simulators, five-axis flight turntables, torque load simulators, flight control computers, external task computers, GPS satellite navigation generators, etc.).

[0046] Specifically, multi-dimensional task scenarios are defined to cover combinations of different task types (such as detection tasks, collaborative tasks) and different task background environments (such as mountainous terrain, adverse weather conditions, urban dense building clusters, electromagnetic interference). Each scenario corresponds to specific simulation conditions and constraints. Correspondingly, test tasks are specific verification requirements designed for different task scenarios. For example, testing the communication stability of an unmanned aerial vehicle formation in a rainy environment, or evaluating the radar detection accuracy in an urban canyon scenario. Exemplarily, a limited number of representative task scenarios can be preset according to actual needs. On the one hand, it comprehensively covers the collaborative adaptability of the model and equipment under diverse conditions. On the other hand, it avoids the additional pre-test burden caused by redundant or ineffective test tasks, providing accurate and reliable basic data for subsequent performance analysis. The specific task scenarios and test tasks are not limited here.

[0047] It should be understood that by simulating the complexity of real scenarios, the adaptability and collaborative efficiency of simulation resources in heterogeneous environments are systematically verified. For example, detection tasks can test the attenuation of the perception accuracy of the model under extreme weather, and collaborative interaction tasks can test the fluctuation of the communication delay of multiple devices under electromagnetic interference. This multi-dimensional test can obtain fine-grained dynamic performance data, providing a basis for subsequent resource scheduling for differential scenario adaptation and avoiding performance evaluation biases caused by single-scenario testing.

[0048] S102. Combine each simulation model with different simulation devices to obtain multiple hardware-in-the-loop simulation nodes, where the simulation models or simulation devices among the multiple hardware-in-the-loop simulation nodes are different.

[0049] Specifically, a hardware-in-the-loop simulation node refers to the smallest execution unit jointly composed of a simulation model and at least one simulation device. The simulation model is responsible for simulating the digital logic of the system under test, while the simulation device provides a physical interface or a hardware environment. The two cooperate to complete a specific simulation task. By fully cross-combining the simulation model and the simulation device, diverse hardware-in-the-loop simulation nodes are generated. For example, the aircraft model A can be combined with a flight control device and a five-axis flight turntable respectively to obtain two hardware-in-the-loop simulation nodes.

[0050] It should be understood that multiple hardware-in-the-loop simulation nodes cover the diverse possibilities in the hardware-software collaboration of the model and the device, forming a complete set of "model-device" pairings. By testing the adaptability of the model on different devices through the differences between the nodes, it provides a basis for subsequent analysis of the hardware-software collaboration efficiency. Exemplarily, according to the "model-device" pairings in the hardware-in-the-loop simulation nodes, after loading the simulation model into the simulation device, it can be used to execute test tasks or simulation tasks.

[0051] In some embodiments, the simulation device can be a single hardware device or a hardware system integrated by multiple hardware devices. That is to say, multiple hardware devices can be integrated within each simulation device, and according to the depth and breadth that the simulation device can simulate, the simulation devices are divided into lightweight nodes and full-quantization nodes. Among them, the lightweight nodes can be used to simulate or replace the functions of some small controllers or sensors in the system, and the full-quantization nodes are closer to the complete hardware system in actual applications, providing more comprehensive and accurate function simulation than the lightweight nodes or directly operating as a part of a more reliable system.

[0052] Exemplarily, in the simulation of unmanned aerial vehicles, the lightweight nodes can include a real-time simulator, a flight control computer, and an external task computer, which are used for the rapid verification of core functions and meet the lightweight and low-cost simulation requirements; the full-quantization nodes can include a real-time simulator, a five-axis flight turntable, a torque load simulator, a flight control computer, an external task computer, a GPS satellite navigation generator, etc. The configuration of the full-quantization nodes is more comprehensive and can more accurately simulate the execution of tasks such as flight and detection in the real environment. Among them, the real-time simulator is the core computing unit that runs the corresponding simulation model, environment model, and control algorithm; the flight control computer is used to execute flight control logic (such as attitude control, navigation algorithm); the external task computer is used to manage the simulation of task loads (such as sensors, weapon systems); the five-axis flight turntable is used to simulate the attitude changes of the aircraft (such as pitch, roll, yaw, X / Y / Z axis displacement); the torque load simulator is used to simulate the dynamic response of the aircraft under aerodynamic and mechanical loads.

[0053] It should be understood that in practical applications, hardware-in-the-loop simulation also uses various integrated hardware systems corresponding to lightweight nodes and full-quantization nodes. The device combination rules can be preset based on actual usage requirements. If the simulation device is a single hardware device, multiple hardware devices can be combined into simulation devices in various hardware system forms according to the device combination rules.

[0054] Thus, each combination form of the simulation model and the simulation device is both diverse and streamlined, ultimately forming a controllable number of hardware-in-the-loop simulation nodes. On the one hand, it is more in line with the actual configuration of the hardware-in-the-loop simulation nodes in practical applications, improving the effectiveness of the hardware-in-the-loop simulation nodes used in the test, and further enhancing the availability of the subsequent performance test data. On the other hand, there is no need to perform cumbersome and complex high-precision pairing between the simulation model and the simulation device according to other factors such as the performance and task scenarios to be tested by the simulation model, avoiding the interference test data generated by redundant or invalid hardware-in-the-loop simulation nodes, and effectively reducing the complexity of the pairing of hardware-in-the-loop simulation nodes and the difficulty and computational amount of subsequent tests.

[0055] S103. Generate multiple cooperation combinations based on the multiple hardware-in-the-loop simulation nodes to cooperatively execute multiple test tasks, and obtain the performance test data of each simulation model under different simulation devices, different cooperation combinations, and different task scenarios.

[0056] Specifically, a cooperation combination refers to a collaborative working unit composed of two or more hardware-in-the-loop simulation nodes, which is used to simulate the scenario of multiple devices jointly executing tasks during the real task process. For each test task, several hardware-in-the-loop simulation nodes can be selected from multiple hardware-in-the-loop simulation nodes according to the simulation model requirements of the test task to generate multiple cooperation combinations, and the corresponding hardware-in-the-loop simulation nodes are called through different cooperation combinations to execute multiple times to verify the performance of each simulation model under different simulation devices, different cooperation combinations, and different task scenarios, and record the performance test data of each simulation model. That is to say, the performance test data is the model performance data of the simulation model under specific conditions. For example, radar model B is respectively loaded into the full quantization node and the lightweight node, and the detection accuracy, calculation time-consuming for single inference, etc. in the task scenarios of electromagnetic interference or sunny days are obtained.

[0057] For example, in the simulation of an unmanned aerial vehicle, the cooperation combination can be a cluster or a team in the cluster responsible for executing a specific task. The cooperation combination can include the first hardware-in-the-loop simulation node "aircraft model A + full quantization node", the second hardware-in-the-loop simulation node "radar model B + lightweight node", and the third hardware-in-the-loop simulation node "satellite model C + lightweight node".

[0058] For example, when the test task is a detection task, the detection task requires the use of capabilities such as search, tracking, and time synchronization of the unmanned aerial vehicle, and the simulation model requirement of the test task is 3 unmanned aerial vehicle models. Multiple hardware-in-the-loop simulation nodes corresponding to the unmanned aerial vehicle models are screened out, and based on these hardware-in-the-loop simulation nodes, multiple cooperation combinations with 3 hardware-in-the-loop simulation nodes in a group are generated to execute the test task respectively.

[0059] In some embodiments, different or the same hardware-in-the-loop simulation nodes can be combined into cooperation combinations to verify the overall efficiency when homogeneous or heterogeneous resources work together. That is to say, several hardware-in-the-loop simulation nodes in the cooperation combination can be completely the same, partially the same, or completely different, which is not limited herein.

[0060] In some embodiments, both the test task and the target simulation task to be executed include several elements such as the usage requirements of simulation objects, task objectives, environmental parameters, time constraints, and interaction rules. Among them, the task type can be determined according to the task objective, and the general task scenario can be predicted by combining some environmental parameters that are public information, such as the weather conditions and terrain conditions at the task location. Further, the simulation model requirements can be comprehensively analyzed based on multiple elements of the task. The simulation model requirements can include the quantity requirements, type requirements, and performance requirements of the simulation model for the task, etc.

[0061] For example, if the usage requirement of the simulation object is that 3 drones cooperate to perform the tracking task of multiple target objects, it can be determined that the usage requirement of the simulation model is 3 (i.e., the quantity requirement) drone models (i.e., the type requirement), and it is required that the communication delay index of the drone is less than or equal to 50 ms (i.e., the performance requirement). Another example, if the usage requirement of the simulation object is that drones cooperate to perform the tracking task of multiple target objects, it can be determined that the usage requirement of the simulation model is a drone model (i.e., the type requirement), and it is required that the communication delay index of the drone is less than or equal to 50 ms (i.e., the performance requirement).

[0062] Exemplarily, key performance requirements are extracted based on the task type, task background, and usage requirements of simulation objects. For example, detection-type tasks require high-precision sensing, collaborative tasks require low-latency communication, and mountainous scenarios require strong anti-interference capabilities, etc. Then, specific thresholds are set in combination with the task execution conditions (such as real-time requirements and resource budget limitations) to obtain comprehensive performance requirements. It should be noted that the task scenarios and simulation model requirements of both the test task and the target simulation task to be executed can be determined and modified by the user, or obtained with reference to related technologies, which are not limited herein.

[0063] It should be understood that for the test task, its environmental parameters can be associated and stored with the task scenario and the corresponding scenario-sensitive indicators, and its simulation model requirements only consider the quantity requirements and type requirements of the simulation model for the task. For the target simulation task, only some basic information of its environmental parameters is public information, and it can be gradually disclosed as the simulation task progresses (such as the environmental data obtained after performing a detection task). It needs to comprehensively consider the quantity requirements, type requirements, and performance requirements of the simulation model for the task, etc., and the performance requirements can be used to match the simulation model that meets the requirements based on the performance parameter set.

[0064] S104. Analyze the performance test data of each simulation model to generate a performance parameter set for each simulation model, where the performance parameter set includes several scenario-sensitive indicators and several static basic indicators.

[0065] Specifically, statistical analysis is performed on all the collected performance test data. By comparing the value changes of the same evaluation index under multiple conditions, the stability of the performance index under different task scenarios, simulation device types, or collaboration combinations is determined. Then, different evaluation indexes are stored differentially according to the stability, generating a performance parameter set that comprehensively reflects its performance characteristics, and systematically describing the performance of each simulation model under various operating conditions.

[0066] Among them, the static basic index is the stable model performance that is inherent and relatively stable in the model, reflecting the basic efficiency of a single simulation model. For example, characteristics at the technical level such as the operating frequency, signal processing ability, and data processing ability of the simulation model remain relatively stable under different conditions. Such indexes are less affected by the external environment and have small fluctuations under different simulation devices, task scenarios, or collaboration combinations. For example, the "minimum detection distance" of a certain radar model is stable at about 5 meters in multiple tests.

[0067] Among them, the scenario-sensitive index is the dynamic model performance that changes significantly with the external environment or collaboration conditions. For example, characteristics at the specific implementation level such as the action distance, coverage range, detection accuracy, and bit error rate of the radar model may vary greatly with changes in the environment (such as an increase in electromagnetic interference intensity or an increase in terrain complexity). Such indexes have dynamic performance fluctuations under specific task scenarios, simulation devices, or collaboration combinations. For example, although simulation model D has high accuracy on simulation device X, its collaboration with simulation model E may cause a sharp increase in communication delay due to resource competition; another example is that the data packet loss rate of the communication model may increase from 0.1% in the normal environment to 5% in the electromagnetic interference scenario, or the path planning success rate of the navigation model may decrease in the lightweight nodes due to insufficient computing resources in the complex terrain.

[0068] In some embodiments, the scenario-sensitive indexes include: collaboration indexes, which are used to characterize the overall efficiency of multi-node collaborative work, such as the data synchronization delay between nodes, the success rate of task collaborative execution, and the resource competition conflicts during multi-model concurrent operation; scenario indexes, which are used to characterize the impact of environmental dynamics on performance, such as the terrain matching error in the mountainous scenario, the communication bit error rate in strong electromagnetic interference, and the number of replanning times of the path planning algorithm in the multi-obstacle environment.

[0069] It should be understood that the performance parameter set provides the basic performance framework of the simulation model through static basic indicators, and also reflects how the simulation model responds to external factor influences in a complex and changeable actual operation environment through scenario-sensitive indicators. The finally formed performance parameter set decouples the static capabilities and dynamic adaptability of the simulation model, provides a multi-dimensional decision-making basis for subsequent scheduling, and ensures that resources and requirements can be accurately matched during dynamic scheduling. For example, in the performance parameter set of the "visual recognition model", the static basic indicators include "image processing resolution 1080p", while the scenario-sensitive indicators include "the recognition accuracy drops from 92% to 78% under low-light conditions and can be increased to 85% when linked with the infrared sensor device E", thus providing a quantitative basis that takes into account both stability and flexibility for subsequent resource scheduling.

[0070] In some embodiments, the performance parameter set of each simulation model is associated with a type label of the simulation model, such as a radar model, an aircraft model, a drone model, etc., for quickly screening out the required type of simulation model and its performance parameter set according to the simulation model requirements subsequently.

[0071] In some embodiments, the performance test data includes a number of evaluation indicators. Among them, the evaluation indicator is a dimension used to quantify the performance performance of the simulation model under specific test conditions (such as task scenario type, simulation device model, cooperation combination configuration), and the evaluation indicator is reflected by the specific value of the performance test data. Different types of simulation models can use different evaluation indicators. These evaluation indicators include the inherent static characteristics of the model and the dynamic characteristics significantly affected by the external environment.

[0072] In some embodiments, S104 includes: comparing multiple values of the same evaluation indicator in the performance test data of each simulation model to determine the value fluctuation; when the value fluctuation is greater than the preset fluctuation value, determining the corresponding evaluation indicator as the scenario-sensitive indicator, and associating and storing different values of the scenario-sensitive indicator with the task scenario and / or the simulation device and / or the cooperation combination; when the value fluctuation is less than the preset fluctuation value, determining the corresponding evaluation indicator as the static basic indicator.

[0073] Among them, the value fluctuation refers to the numerical change range of the same evaluation indicator under different task scenarios, simulation devices or cooperation combinations, and the value fluctuation can be quantified by the standard deviation, range or relative change rate, which is not limited here. Correspondingly, a preset fluctuation value can be set to determine the stability degree of the evaluation indicator, and its specific value can be set according to actual requirements.

[0074] Specifically, the types of evaluation indicators are distinguished by quantitatively analyzing the range of value changes of the same evaluation indicator in the execution of multiple test tasks in the performance test data. If the value fluctuation of a certain evaluation indicator is greater than the preset fluctuation value, indicating that its performance is significantly affected by the external environment, it is determined to be a scene-sensitive indicator, and its different values ​​are associated with specific influencing factors and stored, such as a specific type of task scene, simulation equipment model, and collaborative combination configuration, to form a dynamic performance mapping relationship. For example, the "bit error rate of 3.5%" of simulation model D is associated with "simulation equipment X" and "mountain scene". If the value fluctuation of a certain evaluation indicator is less than the preset fluctuation value, it is determined to be a static basic indicator, and its mean or stable interval is taken as a fixed parameter.

[0075] For example, the calculation time of a simulation model on three simulation devices is 15ms, 18ms, and 16ms respectively, the value fluctuation is 3ms, and the preset upper limit of the fluctuation value is 20% of the median. That is, the fluctuation within the range of 3.2ms above and below the median 16ms is a static basic indicator, and the range exceeding 3.2ms is a scene-sensitive indicator. At this time, it can be determined that the calculation time of a simulation model is an evaluation indicator as a static basic indicator.

[0076] Exemplarily, the standard deviation of the performance test data across scenarios for each evaluation indicator is calculated. If the standard deviation exceeds the preset fluctuation value, the scenario-sensitive flag is triggered and the associated metadata (such as simulation device ID, collaboration combination number) is extracted to construct a performance parameter set that has both stability description and dynamic adaptation relationship.

[0077] In some embodiments, it is characterized in that the method also includes: analyzing the changing trends of several values ​​of scene-sensitive indicators, and respectively determining the influence coefficients of task scenarios, simulation equipment, and collaborative combinations on the scene-sensitive indicators; determining the influence factors of the scene-sensitive indicators from the task scenarios, simulation equipment, and collaborative combinations according to the influence coefficients, and storing the influence factors in association with the scene-sensitive indicators.

[0078] Specifically, the change trend refers to the numerical change law of the same indicator under different test conditions (such as task scenario type, simulation equipment model, and collaborative combination configuration). For example, the "data packet loss rate" of the communication model increases linearly with the increase of electromagnetic interference intensity. According to the change trend, the independent influence of the three factors of task scenario, simulation equipment, and collaborative combination on the indicator is quantified, that is, the influence coefficient. Correspondingly, a preset coefficient can be set to determine the standard for a factor to have a significant impact on the indicator. The specific value can be flexibly set according to actual needs. When the influence coefficient of a factor is greater than the preset coefficient, the factor is marked as an influence factor and stored in association with the scenario-sensitive indicator.

[0079] Taking the "detection range" evaluation index of a certain radar model as an example, when using simulation device A, it is on average 30% shorter than when using simulation device B. Then the influence coefficient of the simulation device type is 0.3; the task scenario of "mountainous terrain" causes the same index to decrease by 20%, so the influence coefficient of this task scenario type is 0.2; and in different collaboration combinations, the same index only fluctuates up and down by 8%, so the influence coefficient of this collaboration combination type is only 0.08. If the preset coefficient is 0.1, then the "collaboration combination" is not included in the influencing factor, and the influencing factors of the scenario-sensitive index of "detection range" are "simulation device" and "task scenario". Correspondingly, the "detection range" of a certain radar model is stored in association with the "simulation device" and "task scenario".

[0080] Exemplarily, the identification and storage of influencing factors can be achieved through an automated data analysis tool, and the influencing factors can also be quantified by using correlation analysis algorithms such as regression analysis or sensitivity analysis. It is also possible to identify the influencing factors of the same index by controlling variables. For example, under the same task scenario and the same simulation device, analyze the influence of the collaboration combination. The specific algorithm is not limited here.

[0081] In some embodiments, it is characterized in that the method further includes: storing the static basic index as a fixed value or a range interval, where the range interval is determined based on multiple values of the static basic index.

[0082] Specifically, for the static basic index, a differentiated strategy is adopted according to its data stability: if the index has extremely small fluctuations and highly consistent performance in cross-scenario tests, it is stored as a fixed value, such as "memory occupancy rate 20.3%"; if there are slight fluctuations within a controllable range, by statistically analyzing multiple values in the performance test data, calculate its extreme value range or confidence interval. For example, the calculation time of the path planning model fluctuates between 10ms and 12ms, then based on the statistical distribution, it is stored as a range interval, such as "time-consuming interval [10ms, 12ms]". Another example is that the communication time of the communication model is 15ms, 15.5ms, and 14.8ms respectively in three collaboration combinations, and its range interval is defined as [14.5ms, 15.6ms].

[0083] Thus, it not only simplifies the query efficiency of the basic performance, but also retains the necessary details through the upper and lower limits of the interval, provides an elastic matching space for subsequent scheduling, and at the same time avoids the mis-screening problem caused by the excessive strictness of a single fixed value.

[0084] It should be understood that the finally formed parameter set decouples static capabilities from dynamic adaptability. For example, in the parameter set of "visual recognition model D", the static basic metrics include "image processing resolution 1080p", while the scene-sensitive metrics are recorded in combination with the test environment context as "recognition accuracy drops from 92% to 78% under low light conditions and can be increased to 85% when linked with infrared sensor device E", thus providing a quantitative basis for subsequent resource scheduling.

[0085] In some embodiments, before S105, it further includes: obtaining the priorities of the simulation tasks to be executed, and taking at least one simulation task with the highest priority as the target simulation task.

[0086] Specifically, obtain the priorities of all the simulation tasks to be processed. The priorities can be defined by the user or automatically generated according to the urgency of the tasks, resource requirements, and deadlines. For example, the priority of a rescue task is higher than that of a patrol task, which is specifically represented as a numerical value or a level. Subsequently, based on the priority tags, screen out one or more tasks with the highest priority as the "target simulation tasks" to ensure that high-timeliness or high-value critical tasks can obtain resource matching first in a resource competition scenario, and avoid delays in the core process due to chaotic task sequences.

[0087] S105. Based on the task scenario and simulation model requirements of the target simulation task, select a target simulation model and target simulation equipment from the semi-physical simulation resource library according to the simulation model requirements and the performance parameter sets of each simulation model to execute the target simulation task.

[0088] Specifically, for the target simulation task to be executed, the task type can be determined according to the task objective, and the task scenario predicted by combining the currently disclosed environmental parameters can be considered, such as the weather conditions and terrain conditions at the task location. Further, the simulation model requirements can be determined according to the usage requirements of the simulation object. For example, the performance requirements and type requirements of the simulation model for the target simulation task. First, screen out the corresponding type of simulation model based on the type requirements, and then, with the known task scenario of the target simulation task as the limiting item and the simulation equipment as the optional item, select a simulation model whose evaluation indicators meet the performance requirements. If the simulation model requirements have a clear quantity requirement for a certain type of simulation model, finally, according to the quantity requirement of the target simulation task for the simulation model, the corresponding quantity of target simulation models and target simulation equipment with a matching relationship can be determined.

[0089] In some embodiments, the performance parameter sets of each simulation model are screened layer by layer according to the simulation model requirements. First, the simulation models that do not meet the performance requirements are quickly filtered through static basic indicators; for the qualified simulation models that meet the static basic indicators, combined with the task scenario of the target simulation task, with the determined task scenario in the target simulation task as the limit, according to the task scenarios and simulation devices associated with the scenario-sensitive indicators of the qualified simulation models, flexibly analyze the specific values of the scenario-sensitive indicators that the simulation models may present under different task scenarios and simulation devices, and select the combination of the simulation model and the simulation device whose lower limit value meets the performance requirements, which is the qualified simulation node, to achieve the strong binding screening between the simulation model and the simulation device, ensuring that both meet the static performance benchmark and can adapt to the dynamic environment requirements when collaborating.

[0090] Thus, regarding the simulation model and the simulation device as an overall execution unit for joint evaluation fundamentally solves the problem of unqualified performance caused by the decoupling of hardware and algorithms in traditional scheduling, and effectively avoids performance degradation combinations. For example, although a certain simulation model meets the static indicators, if it is combined with a low-performance device, it will cause real-time crashes. The binding screening can directly exclude such combinations, avoid generating invalid cooperation combinations subsequently, and reduce the computational amount and efficiency of resource scheduling.

[0091] Furthermore, based on several qualified simulation nodes that meet the performance requirements, several qualified cooperation combinations are generated. Based on the cooperation combinations associated with the scenario-sensitive indicators of the simulation models, evaluate the cooperation between several simulation models. For example, some combinations may have better or worse performance when cooperating. At this time, the simulation models and simulation devices in the qualified cooperation combination with the best cooperation can be preferentially selected as the target simulation model and the target simulation device.

[0092] Thus, first exclude the simulation models that do not meet the static basic indicators, and then use the simulation model and the simulation device as a bound execution unit for secondary screening to exclude the simulation nodes that do not meet the scenario-sensitive indicators. Finally, introduce the cooperation combination for screening the optimal cooperation efficiency. By screening layer by layer, the redundant computational amount caused by unqualified and inefficient simulation nodes and cooperation combinations is effectively reduced, and the efficiency and response speed of resource scheduling are improved. It should be noted that for different cooperation combinations, the performance of the simulation model may show a small downward fluctuation due to resource competition and other reasons, but in actual applications, more often than not, better model efficiency is generated through mutual cooperation. Therefore, even if the screening of the cooperation combination is postponed, based on the principle of selecting the best, it is possible to avoid selecting the cooperation combination with downward fluctuations, ensuring that the finally selected target simulation model and target simulation device can still meet the performance requirements.

[0093] It should be understood that for the performance parameters of each simulation model, an index system that separates static and dynamic aspects is established, breaking the "one-size-fits-all" resource description mode. Static basic indicators, which are independent of environmental changes, are used to reflect the inherent performance of the model. At the initial stage of task matching, models that do not meet the basic capability requirements are quickly filtered through static indicators, significantly reducing the computational complexity of subsequent dynamic evaluations. Scenario-sensitive indicators, by associating influencing factors such as task scenarios and device models, quantify the performance boundaries of the model in real deployment. At the resource binding stage, the model-device combination with the optimal scenario-sensitive indicators is preferentially selected in combination with the current task scenario. Thus, while retaining the efficiency advantages of traditional resource scheduling, the dynamic response ability to complex environments is also enhanced.

[0094] In some embodiments, since the task scenario can have diverse parameters, and the target simulation task generally discloses or predicts some task scenarios rather than all of them. For known task scenarios, scenario-limiting parameters (such as weather conditions, temperature conditions) can be determined, and for unknown task scenarios, scenario-selectable parameters can be determined. In the subsequent model selection process, the scenario-sensitive indicators can be pre-screened according to the parameter values of the scenario-limiting parameters, and the scenario-sensitive indicators whose associated stored scenario-limiting parameters match the scenario-limiting parameters of the target simulation task (for example, the weather condition of the target simulation task is cloudy, and the weather condition bound to a certain scenario-sensitive indicator is also cloudy) are retained, and the scenario-sensitive indicators whose associated stored scenario-limiting parameters do not match the scenario-limiting parameters of the target simulation task (for example, the limiting parameter of the target simulation task is 20 °C, and the weather condition bound to a certain scenario-sensitive indicator is 30 °C - 40 °C) are set to a hidden state. Thus, with the known task scenarios of the target simulation task as the limiting items, it is ensured that the performance parameters that do not match the task scenario of the target simulation task are not used in the resource allocation process of this target simulation task, further improving the resource allocation efficiency.

[0095] In some embodiments, the simulation model requirements include performance requirements, and the 105 further includes: screening a number of unused simulation models based on the performance requirements and the static basic indicators of the simulation model to obtain a number of alternative simulation models; generating a variety of alternative simulation nodes based on the number of alternative simulation models and a number of alternative simulation devices that are not used in the semi-physical simulation resource library; and obtaining a variety of alternative cooperation combinations based on the variety of alternative simulation node combinations; evaluating the prediction performance and resource occupancy rate of each alternative cooperation combination in the task scenario of the target simulation task based on the scenario-sensitive indicators of the alternative simulation models; selecting the alternative cooperation combination with the prediction performance meeting the performance requirements and the lowest resource occupancy rate as the target cooperation combination, and using the alternative simulation model and alternative simulation device within the target cooperation combination as the target simulation model and the target simulation device.

[0096] Specifically, multi-dimensional screening is carried out on the performance parameter sets of each simulation model according to the simulation model requirements. First, a number of unused simulation models in the hardware-in-the-loop simulation resource library are quickly filtered through static basic indicators, and the unqualified simulation models are removed. For a number of alternative simulation models that meet the static basic indicators, a variety of alternative simulation nodes are generated based on a number of alternative simulation models and a number of alternative simulation devices that are not used in the hardware-in-the-loop simulation resource library. On this basis, a variety of alternative cooperation combinations are obtained by randomly combining a variety of alternative simulation nodes. According to the cooperation combinations, task scenarios, and simulation devices associated with the scenario-sensitive indicators of the alternative simulation models, combined with the task scenario of the target simulation task, the specific values of the scenario-sensitive indicators that the simulation model may present under different alternative cooperation combinations and alternative simulation devices are flexibly analyzed. Furthermore, the comprehensive performance that can be generated by the cooperation of multiple simulation models is evaluated from the macroscopic perspective of the entire cooperation combination, that is, the prediction performance. Among them, the prediction performance refers to the expected comprehensive effectiveness level of the alternative cooperation combination in the target simulation task scenario, such as task completion time, target recognition accuracy, cooperation delay, etc.

[0097] At the same time, the resource occupancy rate can be comprehensively evaluated from dimensions such as the number of simulation models, the number of simulation devices, and the hardware resource consumption required for running the simulation (such as CPU occupancy, memory bandwidth, communication bandwidth) occupied by the alternative cooperation combination. Then, the prediction performance and the resource occupancy rate are converted into a comprehensive score through a weighted algorithm. For example, the target cooperation combination of "meeting the performance standard and having the lowest hardware resource consumption" or "meeting the performance standard and having the fewest simulation models" is preferentially selected. Thus, on the one hand, the actual occupancy rate can be accurately calculated to avoid resource overload. On the other hand, the simulation task can be completed with fewer simulation resources selectively, and the saved simulation resources can be used for other simulation tasks to maximize the resource utilization efficiency.

[0098] For example, for the alternative cooperation combinations including "simulation model F + simulation device X" and "simulation model G + simulation device Y", by retrieving the scenario-sensitive indicator of simulation model F as "the strong electromagnetic interference scenario recognition rate on simulation device X is 85%" and the scenario-sensitive indicator of simulation model G as "the mountain path planning time-consuming on simulation device Y is 18 ms", combined with the static basic indicators of simulation model F and simulation model G as "data synchronization interval (5 ms)", the prediction model is used to predict that the comprehensive recognition rate of this alternative cooperation combination in the "mountain electromagnetic confrontation" scenario of the target simulation task is 82% and the task cycle is 23 ms.

[0099] The resource occupancy rate can be determined by accumulating the static basic indicators of each simulation model within the alternative cooperation combinations and the resource overheads of the simulation devices. For example, the memory occupancy of simulation device X is 1.2 GB, the CPU utilization rate of simulation device Y is 40%, and the cross-device communication bandwidth occupancy is 15%. Calculate the overall resource consumption, and then combine the number of simulation models and the number of simulation devices required by the alternative cooperation combinations to comprehensively determine the resource occupancy rate.

[0100] If it is finally determined that the predicted performances of two alternative cooperation combinations both meet the performance requirements, where the first alternative cooperation combination requires 3 simulation models, while the second alternative cooperation combination only requires 2 simulation models, the second alternative cooperation can be selected as the target cooperation combination at this time.

[0101] In some embodiments, a prediction model is trained based on a labeled dataset of the performance performances actually demonstrated by a large number of cooperation combinations in actual tasks. Thus, based on the prediction model, the predicted performance of the alternative cooperation combinations can be predicted according to the static basic indicators and scenario-sensitive indicators of each alternative simulation model in the alternative cooperation combinations.

[0102] It should be understood that during the execution of the simulation task, if a large amount of resources have been occupied, and it is impossible to select a simulation model that matches the performance requirements by layer-by-layer screening of the performance parameter sets of each simulation model according to the simulation model requirements (that is, pre-screening based on static basic indicators; one-round screening based on the simulation devices and task scenarios associated with the scenario-sensitive indicators of the simulation model, and two-round screening based on the cooperation combinations and task scenarios associated with the scenario-sensitive indicators of the simulation model), a comprehensive screening can be triggered at this time (that is, pre-screening based on static basic indicators; unified screening based on the simulation devices, task scenarios, and cooperation combinations associated with the scenario-sensitive indicators of the simulation model). By comprehensively planning the collaborative efficiency of the model and the device, and the model and the model with greater computing power, a strong binding screening between the simulation model and the simulation device, and between the simulation model and the simulation model can be achieved. Combinations with "insufficient single-body performance but complementary combinations" can be selected when resources are limited to meet the overall performance standards.

[0103] In some embodiments, the simulation model requirements include performance requirements, and the 105 further includes: screening a number of unused simulation models based on the performance requirements and the static basic metrics of the simulation models to obtain a number of qualified simulation models; generating a variety of qualified simulation nodes based on the number of qualified simulation models and a number of unused simulation devices in the hardware-in-the-loop simulation resource library; evaluating the prediction performance of each qualified simulation node based on the scenario-sensitive metrics of the qualified simulation models and the task scenario of the target simulation task; selecting a number of qualified simulation nodes whose prediction performance meets the performance requirements; generating a number of qualified collaboration combinations based on the number of qualified simulation nodes that meet the performance requirements, and selecting the simulation models and simulation devices in the qualified collaboration combination with the best collaboration performance under the task scenario of the target simulation task as the target simulation model and the target simulation device.

[0104] If no qualified simulation node whose prediction performance meets the performance requirements is selected. For example, when a drone model and a visual recognition model are required at this time, only the qualified simulation node corresponding to the drone model that meets the performance requirements is selected, and the qualified simulation node corresponding to the visual recognition model that meets the performance requirements is not selected.

[0105] At this time, screening a number of unused simulation models based on the performance requirements and the static basic metrics of the simulation models to obtain a number of alternative simulation models; generating a variety of alternative simulation nodes based on the number of alternative simulation models and a number of unused alternative simulation devices in the hardware-in-the-loop simulation resource library; obtaining a variety of alternative collaboration combinations based on the combination of a variety of alternative simulation nodes; evaluating the prediction performance of each alternative collaboration combination under the task scenario of the target simulation task based on the scenario-sensitive metrics of the alternative simulation models; selecting the alternative collaboration combination whose prediction performance meets the performance requirements as the target collaboration combination, and using the alternative simulation models and alternative simulation devices within the target collaboration combination as the target simulation model and the target simulation device. At this time, the target collaboration combination may include a qualified simulation node corresponding to a visual recognition model that does not meet the performance requirements in the layer-by-layer screening, and a simulation node corresponding to another infrared sensor device, and the collaboration of the two can meet the performance requirements.

[0106] It should be noted that the alternative simulation models or qualified simulation models obtained by screening a number of unused simulation models based on the performance requirements and the static basic metrics of the simulation models are essentially the same simulation models, and different names are given for the purpose of distinguishing layer-by-layer screening and overall screening.

[0107] Quantify the real-time impact of task scenarios, device characteristics, and collaboration modes on model performance through scenario-sensitive metrics, and implement resource scheduling based on multi-dimensions (simulation performance, resource occupancy) using a layer-by-layer screening and overall screening mechanism to address performance degradation or complementary effects in strongly coupled environments, predict and avoid performance traps, break through the limitations of single-resource performance evaluation, and optimize the dynamic response ability of resource scheduling to complex environments. For example, when resources are sufficient, prioritize avoiding performance degradation combinations caused by coupling between devices or models; when resources are limited, achieve performance complementarity through overall planning. For example, the single performance of a certain model is insufficient, but it can meet the task requirements after being combined with specific devices or collaboration combinations, thereby maximizing resource utilization, avoiding resource waste caused by "information islands" in traditional scheduling, and achieving dynamic adaptation in complex scenarios, improving the efficiency and scalability of the simulation system.

[0108] In some embodiments, by pre-evaluating the predicted performance and resource occupancy rate of alternative collaboration combinations, it supports rapid switching when resources are tight and avoids the rigidity caused by fixed resource allocation in traditional scheduling.

[0109] In some embodiments, the method further includes: loading the target simulation model into the corresponding target simulation device to obtain a plurality of target hardware-in-the-loop simulation nodes; during the execution of the target simulation task, continuously monitoring the task execution quality of the plurality of target hardware-in-the-loop simulation nodes; when the task execution quality is lower than the preset quality, obtaining the measured performance data of the target hardware-in-the-loop simulation node; if the difference between the measured performance data of the target hardware-in-the-loop simulation node and the performance test data is greater than the preset difference, isolating the simulation data generated by the target hardware-in-the-loop simulation node.

[0110] Specifically, in the simulation task execution stage, load the target simulation model onto the corresponding target simulation device to form a plurality of target hardware-in-the-loop simulation nodes, such as deploying a communication model to a selected communication simulation device. Subsequently, during the task operation, continuously monitor and collect the task execution quality of each target hardware-in-the-loop simulation node in real time. Among them, the task execution quality is used to characterize the actual operation performance of the target hardware-in-the-loop simulation node, including key parameters such as data accuracy (such as target recognition rate), latency (such as communication response time), and resource utilization rate (such as CPU occupancy rate). For example, in the simulation of an unmanned aerial vehicle formation, if a certain target hardware-in-the-loop simulation node is responsible for path planning, its task execution quality may consist of indicators such as "path deviation ≤ 2 meters" and "task completion time ≤ 10 seconds".

[0111] When it is detected through monitoring that the task execution quality of a certain node is lower than the preset quality (such as the path deviation exceeds 3 meters or the task times out), the measured performance data of the target hardware-in-the-loop simulation node will be obtained, that is, the performance parameters actually collected under the current operating state. Subsequently, the measured performance data will be compared with the expected value in the historical performance test data of the target simulation model. If the difference exceeds the preset difference, it is determined that the target hardware-in-the-loop simulation node is abnormal. At this time, the simulation data generated by this node will be immediately isolated, such as cutting off the data flow output to the main system or marking it as invalid data, to prevent abnormal data from affecting the overall simulation result. It should be understood that a series of highly relevant performance test data with the highest similarity to the task scenario type, collaboration combination configuration, and simulation device model of the target simulation task can be selected to compare with the measured performance data.

[0112] Among them, the setting of the preset quality and the difference threshold can be based on the statistical distribution of historical test data. For example, taking the upper limit of the 95% confidence interval as the preset quality can ensure both rapid response to anomalies and avoidance of misjudgment due to short-term fluctuations. The specific numerical values are not limited here.

[0113] It should be understood that by dynamically comparing the measured data with the expected data, the performance degradation of the simulation model and simulation device caused by environmental fluctuations (such as hardware overload, sudden interference) during actual operation can be accurately identified. For example, if a radar model meets the anti-interference ability standard during testing, but its performance drops suddenly during actual operation due to poor equipment heat dissipation, the difference between its measured data and the test data will trigger isolation to avoid polluting the global simulation with false detection results.

[0114] In some embodiments, the method further includes: if the difference between the measured performance data of the target hardware-in-the-loop simulation node and the performance test data is less than the preset difference, based on the scenario-sensitive index of the target simulation model, select an optimal simulation device from several unused simulation devices in the hardware-in-the-loop simulation resource library; load the target simulation model into the optimal simulation device to replace the target simulation device and update the target hardware-in-the-loop simulation node.

[0115] Specifically, during the simulation task execution phase, if it is detected through monitoring that the difference between the measured performance data of the target hardware-in-the-loop simulation node and the historical performance test data does not exceed the preset difference, the current combination of the simulation model and the simulation device still has basic performance guarantee. At this time, further based on the scenario-sensitive index of this model, an optimal simulation device is selected from the alternative simulation devices that have not been used in the resource library. Subsequently, the target simulation model is migrated from the original target simulation device to the optimal simulation device to update the target hardware-in-the-loop simulation node, so as to optimize resource allocation while ensuring the task quality.

[0116] Exemplarily, according to the scenario-sensitive metrics of the current target simulation model, the task scenario it is in, and the collaboration combination, determine the simulation device that has a beneficial effect on the model performance of the target simulation model as the preferred simulation model. That is to say, associated storage is the basis for quickly locating performance fluctuations and eliminating performance fluctuations in resource allocation. The preferred simulation device is a simulation device with better performance in the simulation environment (i.e., task scenario, collaboration combination) where the current target simulation model is located, and its selection goal is to ensure that the model maintains or improves its effectiveness in a dynamic environment.

[0117] It should be understood that when the task scenario suddenly changes (such as the weather changing from sunny to heavy rain) or a minor performance fluctuation is caused by equipment failure, to avoid triggering excessive intervention, the alternative combination can be quickly switched according to the associated relationship of the scenario-sensitive metrics. For example, enable a model-device combination with a higher "heavy rain scenario recognition rate" without the need for a full-scale retest. Thus, utilize the scenario-sensitive metrics to identify the collaborative potential between the simulation model and the simulation device to improve the quality of completing the simulation task.

[0118] Steps such as the unified management of model resources, model dynamic loading technology, and exception handling mechanism in the embodiments of this application. Among them, the construction of the model resource pool can achieve the unified management and scalability of heterogeneous model resources, improving the reusability and utilization rate of the models; the model dynamic loading technology is to carry out the dynamic loading of the hardware-in-the-loop model in the simulation, strengthening the flexible deployment of the hardware-in-the-loop simulation model; the exception handling mechanism can achieve fault tolerance control in the process of hardware-in-the-loop simulation, improving the robustness and stability of the hardware-in-the-loop simulation system.

[0119] In some embodiments, to meet the diverse simulation task requirements, it is necessary to perform unified formal description and functional encapsulation on the simulation models and simulation devices in the hardware-in-the-loop simulation resource library, map the specific resources to logical resources, break the barriers between the resources of the co-simulation nodes, and achieve the unified management of the hardware-in-the-loop simulation resources and the comprehensive scheduling in the simulation process, improving the efficiency of resource scheduling and the system simulation performance.

[0120] Please refer to Figure 2 , Figure 2 which is a schematic diagram of the encapsulation of the hardware-in-the-loop simulation resource library provided by the embodiments of this application. As Figure 2As shown, the simulation resources in the hardware-in-the-loop simulation resource library include simulation models such as aircraft models, UAV models, satellite models, and radar models, as well as simulation devices such as full quantization node 1, full quantization node 2, full quantization node N, lightweight node 1, lightweight node 2, and lightweight node N. Different test tasks in different task scenarios are set in advance according to different task requirements in the actual application scenario. For example, in the case of collaborative task requirements, it includes time-space synchronization tasks, transmission tasks, data update tasks, and network transmission tasks; in the case of detection task requirements, it includes search tasks, target indication tasks, and tracking tasks. Based on this, steps S102 to S104 are executed to obtain scenario-sensitive indicators at the implementation level, such as operating range, coverage, detection accuracy, and bit error rate; and static basic indicators at the technical level, such as operating frequency, signal processing ability, and data processing ability. The simulation models are virtualized and encapsulated from these two levels respectively to obtain a performance parameter set.

[0121] It should be understood that in the highly adaptable hardware-in-the-loop simulation scenario, the dynamic adaptability and coupling effect of models and devices in different scenarios and collaborative combinations can be captured, and the implicit dependencies between heterogeneous resources can also be captured in cross-device and cross-domain scenarios, avoiding the formation of "information islands". Further, the real-time impact of quantization task scenarios, device characteristics, and collaboration modes on model performance is quantified, and resource allocation is matched based on multiple dimensions (simulation performance, resource occupancy) to cope with performance degradation or complementary effects in a strongly coupled environment (such as a low-performance model achieving functional compliance through a specific device combination). When resources are limited, the utilization rate is improved through optimal combination, or when a task execution anomaly occurs, the problem is quickly isolated and resources are reconstructed, ultimately resulting in limited scalability and flexibility of the simulation system and difficulty in meeting the high-efficiency collaboration requirements in complex scenarios.

[0122] Please refer to Figure 3 , Figure 3 FIG. is a schematic block diagram of the structure of a computer device provided by an embodiment of the present application. The computer device can be a terminal device or a server.

[0123] Exemplarily, the above method can be implemented in the form of a computer program, and the computer program can run on a computer device as shown in Figure 3 .

[0124] As shown in Figure 3 , the computer device includes a processor, a memory, and a network interface connected through a system bus. Among them, the memory can include a non-volatile storage medium and an internal memory.

[0125] The non-volatile storage medium can store an operating system and a computer program. The computer program includes program instructions, and when the program instructions are executed, the processor can execute any scheduling method for hardware-in-the-loop simulation resources.

[0126] The processor is used to provide computing and control capabilities to support the operation of the entire computer device.

[0127] The internal memory provides an environment for the operation of a computer program in a non-volatile storage medium. When the computer program is executed by the processor, the processor can execute any one of the scheduling methods for semi-physical simulation resources.

[0128] The network interface is used for network communication, such as sending assigned tasks, etc.

[0129] It should be understood that the processor can be a central processing unit (CPU), and the processor can also be other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc.

[0130] Among them, in one embodiment, the processor is used to run a computer program stored in the memory to implement the following steps:

[0131] S101. Obtain a semi-physical simulation resource library and test tasks under multiple different task scenarios. The semi-physical simulation resource library includes several simulation models and several simulation devices;

[0132] S102. Combine each simulation model with different simulation devices to obtain multiple semi-physical simulation nodes. Among them, the simulation models or simulation devices are different among the multiple semi-physical simulation nodes;

[0133] S103. Generate multiple cooperation combinations based on the multiple semi-physical simulation nodes to cooperate in executing multiple test tasks, and obtain performance test data of each simulation model under different simulation devices, different cooperation combinations, and different task scenarios;

[0134] S104. Analyze the performance test data of each simulation model to generate a performance parameter set for each simulation model. The performance parameter set includes several scenario-sensitive indicators and several static basic indicators;

[0135] S105. Based on the task scenario and simulation model requirements of the target simulation task, select a target simulation model and target simulation equipment from the semi-physical simulation resource library according to the simulation model requirements and the performance parameter sets of each simulation model to execute the target simulation task.

[0136] Exemplarily, the processor is used to run the computer program stored in the memory and is also used to implement the steps of the scheduling method for semi-physical simulation resources provided in any embodiment of the present application, which will not be elaborated here.

[0137] An embodiment of the present application also provides a computer-readable storage medium. The computer-readable storage medium stores a computer program, and the computer program includes program instructions. The processor executes the program instructions to implement the steps of the scheduling method for semi-physical simulation resources provided in any one of the embodiments of the present application.

[0138] Among them, the computer-readable storage medium may be an internal storage unit of the computer device described in the foregoing embodiment, such as the hard disk or memory of the computer device. The computer-readable storage medium may also be an external storage device of the computer device, such as a plug-in hard disk, a SmartMedia Card (SMC), a Secure Digital (SD) card, a Flash Card, etc. equipped on the computer device.

[0139] As described above, the above is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present application can easily think of various equivalent modifications or replacements, and these modifications or replacements should be covered within the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.

Claims

1. A scheduling method for semi-physical simulation resources, characterized in that The method includes: S101. Obtain a hardware-in-the-loop simulation resource library and test tasks under multiple different task scenarios. The hardware-in-the-loop simulation resource library includes several simulation models and several simulation devices; S102. Combine each simulation model with different simulation devices to obtain multiple hardware-in-the-loop simulation nodes. Among them, the simulation models or simulation devices are different among the multiple hardware-in-the-loop simulation nodes; S103. Generate multiple cooperation combinations based on the multiple hardware-in-the-loop simulation nodes to cooperate in executing multiple test tasks, and obtain performance test data of each simulation model under different simulation devices, different cooperation combinations, and different task scenarios; the performance test data includes several evaluation indicators; S104. Analyze the performance test data of each simulation model to generate a performance parameter set for each simulation model. The performance parameter set includes several scenario-sensitive indicators and several static basic indicators; S104 includes: comparing multiple values of the same evaluation indicator in the performance test data of each simulation model to determine the value fluctuation; when the value fluctuation is greater than a preset fluctuation value, determine the corresponding evaluation indicator as the scenario-sensitive indicator, and associate and store different values of the scenario-sensitive indicator with the task scenario and / or simulation device and / or cooperation combination; when the value fluctuation is less than the preset fluctuation value, determine the corresponding evaluation indicator as the static basic indicator; S105. Based on the task scenario and simulation model requirements of the target simulation task, select a target simulation model and a target simulation device from the hardware-in-the-loop simulation resource library according to the simulation model requirements and the performance parameter set of each simulation model to execute the target simulation task.

2. The method according to claim 1, wherein The method further includes: Analyze the change trend of several values of the scenario-sensitive indicator, and respectively determine the influence coefficients of the task scenario, simulation device, and cooperation combination on the scenario-sensitive indicator; Determine the influence factors of the scenario-sensitive indicator from the task scenario, simulation device, and cooperation combination according to the influence coefficients, and associate and store the influence factors with the scenario-sensitive indicator.

3. The method according to claim 1, wherein The method further includes: storing the static basic indicator as a fixed value or a range interval, where the range interval is determined based on multiple values of the static basic indicator.

4. The method according to claim 1, characterized in that, The simulation model requirements include performance requirements, and S105 further includes: Based on the performance requirements and the static basic indicators of the simulation model, screen several unused simulation models to obtain several alternative simulation models; Generate multiple alternative simulation nodes based on the several alternative simulation models and several alternative simulation devices not used in the hardware-in-the-loop simulation resource library; and generate multiple alternative cooperation combinations based on the multiple alternative simulation nodes; Evaluate the prediction performance and resource occupancy rate of each alternative cooperation combination under the task scenario of the target simulation task based on the scenario-sensitive indicators of the alternative simulation models. Select the alternative cooperation combination with the predicted performance meeting the performance requirements and the lowest resource occupancy rate as the target cooperation combination, and use the alternative simulation models and alternative simulation devices within the target cooperation combination as the target simulation model and the target simulation device.

5. The method according to claim 1, wherein Before the S105, it further includes: obtaining the priority of the simulation tasks to be executed, and using at least one simulation task with the highest priority as the target simulation task.

6. The method according to claim 1, wherein The method further includes: Loading the target simulation model into the corresponding target simulation device to obtain a number of target hardware-in-the-loop simulation nodes; During the execution of the target simulation task, continuously monitor the task execution quality of the number of target hardware-in-the-loop simulation nodes; When the task execution quality is lower than the preset quality, obtain the measured performance data of the target hardware-in-the-loop simulation node; If the difference between the measured performance data of the target hardware-in-the-loop simulation node and the performance test data is greater than the preset difference, isolate the simulation data generated by the target hardware-in-the-loop simulation node.

7. The method according to claim 6, wherein The method further includes: If the difference between the measured performance data of the target hardware-in-the-loop simulation node and the performance test data is less than the preset difference, select a preferred simulation device from several unused simulation devices in the hardware-in-the-loop simulation resource library based on the scenario-sensitive index of the target simulation model; Load the target simulation model into the preferred simulation device to replace the target simulation device and update the target hardware-in-the-loop simulation node.

8. A computer device, characterized in that, The device includes: A memory for storing computer programs; A processor for executing the computer program and implementing the scheduling method of the hardware-in-the-loop simulation resources as described in any one of claims 1 to 7 when executing the computer program.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, and when the computer program is executed by the processor, the processor is caused to implement the scheduling method of the hardware-in-the-loop simulation resources as described in any one of claims 1 to 7.

Images