Multi-dimensional battle field interactive digital simulation engine construction method

Abstract

The invention discloses a multi-dimensional battle field interactive digital simulation engine construction method, which belongs to the technical field of physical simulation, and comprises the following steps: constructing a distributed engine; designing an intelligent agent distributed dynamic deployment mechanism; designing battle field partition management; designing a distributed interaction mechanism; the battle domain is a medium for information exchange and state updating among different intelligent agents. According to the method for constructing the interactive digital simulation engine in the multi-dimensional battle domain, the problem of collaboration between tens of thousands of agents and complex physical field battle domain calculation is solved, calculation precision and real-time performance are both considered, intelligent scheduling of calculation resources and dynamic agent load balancing are achieved, the resource utilization efficiency is improved, and the application prospect is wide. And the accuracy and continuity of the simulation process are ensured.

Description

Technical Field

[0001] This application relates to the field of physical simulation, and particularly to a method for constructing a multi-dimensional engagement domain interactive digital simulation engine. Background Art

[0002] With the evolution of modern warfare forms, the combat environment has become increasingly complex, showing multi-dimensional and multi-domain characteristics. Simulation technology is increasingly widely used in the military field, but the current mainstream technologies have significant limitations: Single physical field simulation systems (such as ANSYS, COMSOL, etc.) use methods such as finite element and finite difference to accurately model specific physical fields, but have problems such as insufficient multi-physical field coupling ability, low computational efficiency, poor interaction ability, and strong system closure, making it difficult to meet the real-time simulation requirements of complex battlefield environments.

[0003] The High-Level Architecture (HLA), as a distributed simulation standard framework, aims to achieve system interoperability, but it has high complexity, poor real-time performance, limited interoperability, and insufficient support for multi-physical field coupling calculations.

[0004] Multi-Agent System (MAS)-based simulation systems are suitable for simulating group behaviors, but are limited in multi-physical field coupling and large-scale distributed computing.

[0005] Existing technologies have deficiencies in multi-physical field coupling, computational efficiency, real-time interaction, and system scalability, making it difficult to meet the simulation requirements of modern complex battlefield environments. Summary of the Invention

[0006] This application aims to provide a method for constructing a multi-dimensional engagement domain interactive digital simulation engine to solve the deficiencies in multi-physical field engagement domain coupling, computational efficiency, real-time interaction, and system scalability.

[0007] To achieve the above object, the technical solution of this application is: A method for constructing a multi-dimensional engagement domain interactive digital simulation engine, including, Step S1: Construct a distributed engine; Step S2: Design an agent distributed dynamic deployment mechanism; Step S3: Design engagement domain partition management; Step S4: Design a distributed interaction mechanism; Wherein, the engagement domain is a medium for information exchange and status update between different agents.

[0008] Optionally, in the step S1, it includes: Step S11: Define a standardized interface for agents, including the basic attributes and behaviors of agents; Step S12: Establish the basic framework of the engagement domain and implement the spatial and computational grid division of the engagement domain; Step S13: Design a distributed scheduling method to coordinate the operation timings of components on each server node.

[0009] Optionally, in step S2, it includes: Step S21: Establish a server performance evaluation model; Step S22: Implement agent load balancing using the server performance evaluation model.

[0010] Optionally, in step S3, it includes: Step S31: According to the characteristics of large-scale scenarios, divide the entire scene space into multiple computational regions, and each region is responsible for by an independent engagement domain; Step S32: Design a boundary processing mechanism between regions to ensure the continuity and consistency of the engagement domain at the region boundary; Step S33: Design a cooperation mechanism between engagement domains to handle cross-region impacts.

[0011] Optionally, in step S4, it includes: Step S41: Initialize the interaction event, and the agent sends a signal, which contains the detailed information of the agent; Step S42: Update the state of the engagement domain. After the engagement domain related to the agent that sends the signal receives the signal, first determine the affected area range and calculate the computational grid that needs to be updated; Step S43: Calculate the signal propagation, including: the propagation path and intensity attenuation of the signal; Step S44: Superimpose the engagement domain effects, including: superimposing the impacts of other agents or the environment existing in the scene to obtain the updated distribution state of the engagement domain; Step S45: Evaluate the impact of the target agent, and based on the updated distribution state of the engagement domain, obtain the evaluation result; Step S46: Feedback the distributed event, send the evaluation result to the relevant agent, and trigger the corresponding response action; Step S47: Maintain state consistency. Adopt a distributed state synchronization mechanism, and between each engagement domain, through regular state verification and incremental update, maintain the consistency of the engagement domain values.

[0012] Optionally, after step S4, it further includes: Step S5: Optimize the distributed interaction mechanism. In view of the characteristics of combat domain calculation under the distributed interaction mechanism, implement a multi-level optimized distributed interaction mechanism; adopt the adaptive grid technology, use a finer grid division in key areas and a coarser grid in distant areas to improve the calculation efficiency; decompose the large-scale combat domain numerical calculation tasks into multiple subtasks and allocate them to different computing nodes in the combat domain for parallel processing to achieve dynamic allocation of calculation tasks; establish a distributed caching mechanism to cache static or slowly changing scene features, reduce cross-node data transmission and repeated calculations, and improve the combat domain numerical calculation efficiency.

[0013] Optionally, after the step S5, the following steps are further included: Step S6: Evaluate and feedback the distributed effect. Implement the evaluation and feedback mechanism under the distributed interaction mechanism. After all combat domains in the scene are calculated to obtain the evaluation results, evaluate the impact on the agents locally; notify the relevant agents of the evaluation results to trigger corresponding response actions; through the distributed state synchronization mechanism, integrate the evaluation results of all combat domains in the scene, and achieve full-scene-level response and state synchronization through the distributed event bus to ensure that the scene states in all areas are consistent.

[0014] Optionally, set the operator to perform real-time intervention and control on the agents in the scene. After the step S13, the following steps are further included: design an interaction control method to process the operator's real-time instructions; after the step S22, the following steps are further included: the operator adds or removes agents in real time through the control interface to dynamically adjust the scale of the multi-dimensional acoustic combat domain interactive digital simulation.

[0015] Optionally, after the step S12, the following steps are further included: adopt the combat domain coupling mechanism to establish an interaction interface for the combat domain. The interaction interface adopts an asynchronous communication mechanism based on a message queue, and realizes data synchronization through timestamps and version numbers to achieve two-way coupling of the combat domain; when the combat domain changes, update the spatial relationship data through atomic operations.

[0016] Optionally, after the step S5, the following steps are further included: design a parameter optimization mechanism, construct a beam allocation optimization model to achieve dynamic allocation of beam resources; establish a data fusion parameter optimization mechanism, adopt an online learning method based on gradient descent, construct an optimization model including key parameters such as data association threshold, filtering gain, and feature weight, and achieve adaptive adjustment of parameters by minimizing the fusion error; use a long short-term memory network to predict the performance of the agents.

[0017] Optionally, after step S12, the method further includes: designing a field coupling calculator, establishing an interaction model between physical field engagement domains, using the spatial engagement domain as the basic support to provide a unified spatial reference and target state information for other physical field engagement domains; the field coupling calculator adopts a hierarchical coupling architecture, including a physical effect layer, a data exchange layer, and a control and coordination layer. The physical effect layer realizes the interaction calculation between physical field engagement domains; the data exchange layer is responsible for data transmission between different engagement domains, and adopts a combination of shared memory and message queue to ensure the efficient exchange of large-scale data; the control and coordination layer realizes the timing synchronization and calculation coordination between engagement domains, and ensures the causal consistency of multi-physical field calculations through a token passing mechanism.

[0018] Optionally, after step S22, the method further includes: designing an inter-domain communication mechanism, adopting a hierarchical message routing strategy, establishing an intra-engagement domain routing layer responsible for direct communication between nodes within this engagement domain; establishing an inter-engagement domain routing layer to realize message forwarding between different domains through boundary routing nodes; establishing a global routing layer to maintain the inter-domain routing table and optimize the routing path; dynamically adjusting the routing strategy according to the communication frequency, establishing a direct point-to-point communication link for frequently interacting nodes to reduce routing overhead.

[0019] The method for constructing a multi-dimensional engagement domain interactive digital simulation engine provided by this application realizes the efficient coupling calculation of multiple physical fields such as electromagnetic, acoustic, and optical through the engagement domain concept and sub-region management mechanism, and supports the interaction of multiple effects in complex scenarios. By adopting a distributed engine core architecture and an agent dynamic deployment mechanism, it solves the coordination problem of tens of thousands of agents and complex physical field engagement domain calculations, taking into account both calculation accuracy and real-time performance. Through a distributed interaction mechanism and state consistency maintenance, it realizes millisecond-level human-computer interaction response, supports real-time intervention by operators and adjustment of simulation parameters. It supports flexible expansion of new physical field models and interaction mechanisms through standardized interfaces. Using distributed computing optimization and a multi-level parallel computing framework, it realizes intelligent scheduling and dynamic load balancing of heterogeneous computing resources, improving resource utilization efficiency. Through engagement domain partition management and boundary data exchange mechanism, it realizes the efficient management and coordination of multiple engagement domains in a large-scale space, ensuring the accuracy and continuity of the simulation.

[0020] To make the above features and advantages of the application more obvious and understandable, specific embodiments are hereinafter given and described in detail in conjunction with the accompanying drawings as follows. Description of the Drawings

[0021] Figure 1 It is a flowchart of the method for constructing a multi-dimensional engagement domain interactive digital simulation engine of this application. Detailed Embodiments

[0022] To make the objectives and technical solutions of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings of the embodiments of this application. Obviously, the described embodiments are only a part rather than all of the embodiments of this application. All other embodiments obtained by those of ordinary skill in the art based on the described embodiments of this application without creative efforts shall fall within the scope of protection of this application.

[0023] Please refer to Figure 1 , Figure 1 which is a flowchart of a method for constructing a multi-dimensional combat domain interactive digital simulation engine of this application. Among them, the combat domain is an abstract physical field concept and serves as a medium for information exchange and status update between different intelligent agents; the multi-physical fields include multiple physical fields such as acoustics, optics, heat, electricity, force, magnetism, etc., as well as their interactions and coupling effects. Each physical field has its unique propagation characteristics and action laws. For example, the electromagnetic domain is used to simulate the propagation of electromagnetic interference, and the acoustic domain is used to simulate the propagation of sound waves, etc. This application provides a method for constructing a multi-dimensional combat domain interactive digital simulation engine. Introducing the multi-dimensional combat domain realizes the decoupled design of the interaction between intelligent agents, including: Step S1 to Step S6.

[0024] Step S1: Construct a distributed engine.

[0025] Step S2: Design a distributed dynamic deployment mechanism for intelligent agents.

[0026] Step S3: Design the partition management of the combat domain.

[0027] Step S4: Design a distributed interaction mechanism.

[0028] Step S5: Optimize the distributed interaction mechanism.

[0029] Step S6: Evaluate and feedback the distributed effect.

[0030] As an example, in Step S1, to support large-scale simulation scenarios, a distributed engine is constructed to deploy intelligent agents distributively. According to the computing capabilities of different servers, the intelligent agents are flexibly allocated to multiple servers. When there are interaction areas of multiple intelligent agents in the scenario, multiple combat domains are deployed, and each combat domain is responsible for the field value calculation in a specific spatial area, thereby improving the parallel processing ability of the multi-dimensional combat domain interactive digital simulation engine.

[0031] Specifically, Step S1: Construct a distributed engine, including: Step S11 to Step S13.

[0032] Step S11: Define the standardized interface of the agent, including the basic attributes and behaviors of the agent.

[0033] Step S12: Establish the basic framework of the engagement domain, and implement the spatial and computational grid partitioning of the engagement domain.

[0034] Step S13: Design a distributed scheduling method to coordinate the operation timings of components on the nodes of each server.

[0035] As an example, in Step S11, the third version of the FMI standard is adopted to implement the standardized interface of any agent.

[0036] As an example, in Step S12, the establishment of the basic framework of the engagement domain includes: constructing the physical model of the engagement domain, and based on the physical model of the engagement domain, establishing the numerical calculation method of the engagement domain.

[0037] The implementation of the spatial and computational grid partitioning of the engagement domain includes: using an octree structure to recursively partition the engagement domain into any number of subspaces, and performing computational grid partitioning for each subspace according to the computational accuracy requirements. The computational grid partitioning includes: adopting non-uniform grid technology, using finer grid partitioning in key areas, such as near interference sources or target agents in the scenario, and using coarser grids in areas far from interference sources or target agents in the scenario, so as to balance computational accuracy and efficiency.

[0038] As an example, in Step S13, the nodes of the server are divided into master nodes and slave nodes. The master node is responsible for global clock synchronization and task scheduling, and the slave node is responsible for component management local to the server.

[0039] As an example, the master node maintains a global virtual clock and keeps millisecond-level clock synchronization with each slave node through the Network Time Protocol (NTP). In terms of task scheduling, the distributed scheduling method adopts an event-based scheduling mechanism, and divides the multi-dimensional engagement domain interactive digital simulation process into discrete time steps. Within each time step, the master node first broadcasts the task list of the current time step to each slave node, and the slave node immediately executes the calculation of local components after receiving the task. After the calculation is completed, the slave node returns the result to the master node. After the master node collects the calculation results of all slave nodes, it performs data merging and consistency checking, and then advances to the next time step.

[0040] Furthermore, in order to handle node failures, the distributed scheduling method also includes: a fault detection and recovery mechanism. The master node monitors the status of slave nodes through heartbeat detection. When a node failure is found, the tasks of this node are dynamically migrated to other normal nodes. Through the distributed scheduling method, the synchronous execution and state consistency of components under the distributed engine are ensured.

[0041] After designing a distributed scheduling method to coordinate the running timings of components on nodes of each server, an intelligent agent distributed dynamic deployment mechanism is designed based on scenario requirements and the computing capabilities of the servers to achieve the distributed dynamic deployment of satellite intelligent agents. Specifically, step S2: Design an intelligent agent distributed dynamic deployment mechanism, including: steps S21 to S13.

[0042] Step S21: Establish a server performance evaluation model.

[0043] Step S22: Use the server performance evaluation model to achieve load balancing of intelligent agents.

[0044] As an example, in step S21, the server performance evaluation model includes metrics such as computing power, memory capacity, network bandwidth, etc., where the computing power is the CPU utilization rate, and the server performance evaluation results are divided into three levels: high, medium, and low.

[0045] As an example, in step S22, satellite intelligent agents are dynamically allocated according to the evaluation results of the server performance evaluation model. During operation, the load status of each server is continuously monitored. When necessary, the intelligent agent migration mechanism is triggered to transfer some intelligent agents to servers with lighter loads to ensure the load balancing and stable operation of the intelligent agents on the servers.

[0046] Specifically, step S3: Implement war zone partition management, including: steps S31 to S33.

[0047] Step S31: According to the characteristics of large-scale scenarios, divide the entire scenario space into multiple computing regions, and each region is responsible for by an independent war zone.

[0048] Step S32: Design a boundary processing mechanism between regions to ensure the continuity and consistency of the war zones at the region boundaries.

[0049] Step S33: Design a cooperation mechanism between war zones to handle cross-region impacts.

[0050] As an example, in step S32, the boundary processing mechanism includes: at the boundary between regions, the overlapping grid technology is adopted, and an overlapping area of 2 - 3 grid cells is set between adjacent regions. In each time step, after the war zones of adjacent regions respectively construct the physical models of their own war zones and calculate the war zone values, the war zone values in the overlapping area are weighted averaged through an interpolation algorithm to achieve a smooth transition of the war zones.

[0051] As an example, to ensure the stability of numerical calculations, a high-order interpolation method, such as cubic spline interpolation, is adopted in the overlapping area to ensure the continuity of the war zone values and their derivatives.

[0052] Furthermore, when the engagement domain value at the boundary changes drastically, the calculation time step is automatically reduced to improve the calculation accuracy and achieve adaptive time step control.

[0053] As an example, in step S33, the cooperation mechanism includes: when the influence range of the agent spans multiple regions, the engagement domain of the source region first calculates the engagement domain value distribution, conducts data exchange through the boundary processing mechanism, and transmits the engagement domain value to the adjacent region. After receiving the boundary engagement domain value, the engagement domain of the adjacent region uses it as the boundary condition for the calculation in this region and continues to calculate the propagation of the influence of the agent within this region.

[0054] As an example, to ensure the calculation accuracy, an iterative solution strategy is adopted, and multiple boundary data exchanges and engagement domain value updates are performed within each time step until the engagement domain values of adjacent regions converge.

[0055] Furthermore, the engagement domains of each region independently calculate the interference effects on the target agents within their respective regions and then share the evaluation results, finally obtaining a complete interference effect evaluation within the region and implementing a distributed interference effect evaluation mechanism.

[0056] Specifically, step S4: Design a distributed interaction mechanism to implement the interaction mechanism between the agent and the engagement domain under the distributed interaction mechanism, including: steps S41 to S47.

[0057] Step S41: Initialize the interaction event.

[0058] Step S42: Update the engagement domain state.

[0059] Step S43: Calculate the signal propagation.

[0060] Step S44: Superimpose the engagement domain effect.

[0061] Step S45: Evaluate the target impact.

[0062] Step S46: Feedback the distributed event.

[0063] Step S47: Maintain state consistency.

[0064] As an example, in step S41, the agent in the scenario emits a signal that contains detailed information about the agent, including: the agent's position coordinates, motion state, signal characteristics, target agent information, and duration parameters.

[0065] As an example, in step S42, after the engagement domain related to the agent that emits the signal receives the signal, it first determines the affected area range. According to the positional relationship between the interference source and the target agent, combined with the calculation of the numerical distribution in the engagement domain, the calculation grid that needs to be updated is calculated. For cross-region impacts, data exchange is carried out between relevant engagement domains through a boundary processing mechanism to maintain the continuity and consistency of the engagement domain at the regional boundary.

[0066] As an example, in step S43, after determining the affected area range, within the determined affected range, the propagation path and intensity attenuation of a single engagement domain signal are calculated, while also considering the impact of obstacles such as terrain and buildings in the scenario on the signal.

[0067] As an example, in step S44, after calculating the propagation path and intensity attenuation of a single signal, considering the impact of other agents or the environment existing in the scenario, the combined field strength of each other agent is calculated. For multiple agents with similar frequencies, their coherence and intermodulation effects also need to be considered. Finally, the results of the updated engagement domain values are reflected in real time on each node of the calculation grid to obtain the distribution state of the updated engagement domain.

[0068] As an example, in step S45, based on the distribution state of the updated engagement domain, the engagement domain evaluates the impact of the signal on the target agent, while also considering the performance of the target agent to obtain the final evaluation result.

[0069] As an example, in step S46, the engagement domain sends the evaluation result to the relevant agents through a distributed event bus. After receiving these distributed events, the target agent triggers the corresponding response actions.

[0070] As an example, in step S47, to ensure the state consistency of each engagement domain under the distributed interaction mechanism, a distributed state synchronization mechanism is adopted. Through regular state verification and incremental updates between each engagement domain, the consistency of the engagement domain values is maintained. Further, the interaction state between the agent and the engagement domain is also synchronized through the event log and state snapshot mechanism to ensure the coherence and accuracy of the engagement domain interactive digital simulation process in a single region.

[0071] Specifically, step S5: Optimize the distributed interaction mechanism, including: implementing a multi-level optimized distributed interaction mechanism according to the characteristics of combat domain calculation under the distributed interaction mechanism. First, adopt the adaptive grid technology, use finer grid division in key areas and coarser grids in distant areas to improve the calculation efficiency. Further, decompose the large-scale combat domain numerical calculation tasks into multiple subtasks and allocate them to different computing nodes in the combat domain for parallel processing to achieve dynamic allocation of calculation tasks. Finally, establish a distributed caching mechanism to cache static or slowly changing scene features, reduce cross-node data transmission and repeated calculations, and improve the numerical calculation efficiency of the combat domain.

[0072] Specifically, step S6: Evaluate and feedback the distributed effect, and implement the evaluation and feedback mechanism under the distributed interaction mechanism, including: after the evaluation results are obtained for all combat domains in the scene, first evaluate the impact on the intelligent agents locally, then notify the evaluation results to the relevant intelligent agents through the distributed event bus to trigger corresponding response actions, and finally, through the distributed state synchronization mechanism, integrate the evaluation results of all combat domains in the scene and achieve full-scene-level response and state synchronization through the distributed event bus to ensure that the scene states in all regions are consistent.

[0073] In the first specific embodiment of the present application, satellites are used as intelligent agents. Taking the electromagnetic confrontation of a large number of satellite intelligent agents as an example, the satellite intelligent agents exchange information and update their states through the electromagnetic combat domain, demonstrating the construction process of the multi-dimensional electromagnetic combat domain interactive digital simulation engine. Set up a scenario containing tens of thousands of satellite intelligent agents distributed in different orbits, and different servers control the satellite intelligent agents. Among them, some satellite intelligent agents have the ability of electromagnetic interference and can implement electromagnetic interference on the target satellite intelligent agents. At this time, the construction process of the multi-dimensional combat domain interactive digital simulation engine needs to accurately simulate the propagation characteristics of electromagnetic waves, spatial attenuation, multipath effects, and the impact on the communication systems of target satellite intelligent agents.

[0074] As an example, in step S1, a distributed engine is constructed.

[0075] As an example, to support such a large-scale simulation scenario, a distributed engine is constructed for distributed deployment of satellite intelligent agents. According to the computing capabilities of different servers, the satellite intelligent agents are flexibly allocated to multiple servers. For example, a server with stronger performance can deploy 3000 - 5000 satellite intelligent agents, and a server with general performance can deploy 1000 - 2000 satellite intelligent agents. Considering the computationally intensive characteristics of the electromagnetic combat domain, the satellite intelligent agents are independently deployed on dedicated high-performance servers. When there are multiple electromagnetic confrontation areas of satellite intelligent agents in the scenario, multiple electromagnetic combat domains are deployed, and each electromagnetic combat domain is responsible for the electromagnetic field calculation of a specific spatial area, thereby improving the parallel processing ability of the multi-dimensional combat domain interactive digital simulation engine.

[0076] As an example, in step S11, a standardized interface for the satellite agent is defined, including the basic attributes and behaviors of the satellite agent. Among them, the basic attributes include: position, attitude, and communication parameters; the behaviors include: orbital motion, communication, and interference.

[0077] In step S12, a basic framework for the electromagnetic engagement domain is established to achieve the spatial and computational grid division of the electromagnetic engagement domain. First, a physical model of the electromagnetic engagement domain needs to be constructed, including mathematical descriptions such as electromagnetic wave propagation equations, boundary conditions, and initial conditions. Based on these physical models, numerical calculation methods for the electromagnetic engagement domain are established, including the implementation of algorithms such as the finite-difference time-domain method (FDTD) and the method of moments.

[0078] As an example, the implementation of spatial and computational grid division includes: recursively dividing the electromagnetic engagement domain into any number of subspaces using an octree structure, and each subspace is divided into computational grids according to the requirements of computational accuracy.

[0079] As an example, the implementation of computational grid division includes: using non-uniform grid technology, in key areas, such as near interference sources or target satellite agents in the scenario, finer grid division is used, while in areas far from interference sources or target satellite agents in the scenario, coarser grids are used to balance computational accuracy and efficiency.

[0080] In step S13, a distributed scheduling method is designed to coordinate the running timing of components on the nodes of each server. The nodes of the server are divided into master nodes and slave nodes. The master node is responsible for global clock synchronization and task scheduling, and the slave node is responsible for component management on the local server.

[0081] As an example, the master node maintains a global virtual clock and keeps millisecond-level clock synchronization with each slave node through the Network Time Protocol (NTP). In terms of task scheduling, the distributed scheduling method adopts an event-based scheduling mechanism, dividing the multi-dimensional electromagnetic engagement domain interactive digital simulation process into discrete time steps. Within each time step, the master node first broadcasts the task list of the current time step to each slave node. After receiving the task, the slave node immediately executes the calculation of the local component. After completing the calculation, the slave node returns the result to the master node. After the master node collects the calculation results of all slave nodes, it performs data merging and consistency checking, and then advances to the next time step.

[0082] Furthermore, in order to handle node failures, the distributed scheduling method also includes: a fault detection and recovery mechanism. The master node monitors the status of slave nodes through heartbeat detection. When a node failure is found, the tasks of that node are dynamically migrated to other normal nodes. Through the distributed scheduling method, the synchronous execution and state consistency of components under the distributed engine are ensured.

[0083] As an example, in step S2, a distributed dynamic deployment mechanism for intelligent agents is designed.

[0084] After designing a distributed scheduling method to coordinate the operation timing of components on the nodes of each server, based on the scenario requirements and the computing power of the server, a distributed dynamic deployment mechanism for intelligent agents is designed to achieve the distributed dynamic deployment of satellite intelligent agents.

[0085] As an example, in step S21, a server performance evaluation model is established. The server performance evaluation model includes indicators such as computing power, memory capacity, and network bandwidth, where the computing power is the CPU utilization rate. The server performance evaluation results are divided into three levels: high, medium, and low. For low-performance servers (CPU utilization rate < 60%, memory usage rate < 70%, network bandwidth > 10 Gbps), 4000 - 5000 satellite intelligent agents can be allocated; for medium-performance servers (CPU utilization rate 60 - 80%, memory usage rate 70 - 85%, network bandwidth 5 - 10 Gbps), 2000 - 4000 satellite intelligent agents can be allocated; for high-performance servers (CPU utilization rate > 80%, memory usage rate > 85%, network bandwidth < 5 Gbps), 1000 - 2000 satellite intelligent agents can be allocated.

[0086] In step S22, the server performance evaluation model is used to achieve intelligent agent load balancing. According to the evaluation results of the server performance evaluation model, satellite intelligent agents are dynamically allocated, and the load status of each server is continuously monitored during operation. When necessary, the intelligent agent migration mechanism is triggered to transfer some satellite intelligent agents to the server with lighter load, ensuring the load balance of satellite intelligent agents on the server and the stable electromagnetic confrontation performance.

[0087] As an example, in step S3, the management of the engagement domain partition is realized.

[0088] In step S31, according to the characteristics of large-scale scenarios, the entire scenario space is divided into multiple computing regions, and each region is responsible for an independent electromagnetic engagement domain.

[0089] In step S32, a boundary processing mechanism between regions is designed to ensure the continuity and consistency of the electromagnetic engagement domain at the region boundary. At the boundary between regions, the overlapping grid technology is adopted, and an overlapping region of 2 - 3 grid cells is set between adjacent regions. In each time step, after the electromagnetic engagement domains of adjacent regions respectively construct the physical model of the electromagnetic engagement domain of their own regions and calculate the numerical values of the electromagnetic engagement domain, the numerical values of the electromagnetic engagement domain in the overlapping region are weighted and averaged through an interpolation algorithm to achieve a smooth transition of the electromagnetic engagement domain.

[0090] Furthermore, to ensure the stability of numerical calculations, a high-order interpolation method, such as cubic spline interpolation, is adopted in the overlapping region to ensure the continuity of the numerical values and their derivatives in the electromagnetic engagement domain.

[0091] Furthermore, when the numerical values of the electromagnetic engagement domain at the boundary change drastically, the time step of the calculation is automatically reduced to improve the calculation accuracy and achieve adaptive time step control.

[0092] In step S33, a collaborative mechanism between engagement domains is designed to handle the electromagnetic interference effects across regions. When an electromagnetic interference source is located in one region but its influence range spans multiple regions, the electromagnetic engagement domain of the source region where the electromagnetic interference source is located first calculates the radiation field distribution of the electromagnetic interference source, conducts data exchange through the boundary processing mechanism, and transmits the numerical values of the electromagnetic engagement domain to adjacent regions. After receiving the numerical values of the boundary electromagnetic engagement domain, the electromagnetic engagement domains of adjacent regions use them as the boundary conditions for the calculation in their respective regions and continue to calculate the propagation of the interference field within their regions.

[0093] As an example, to ensure the calculation accuracy, an iterative solution strategy is adopted, and multiple boundary data exchanges and updates of the numerical values of the electromagnetic engagement domain are performed within each time step until the numerical values of the electromagnetic engagement domains of adjacent regions converge.

[0094] Furthermore, the electromagnetic engagement domains of each region independently calculate the interference effects on the targets within their respective regions and then share the evaluation results to finally obtain a complete interference effect evaluation within their regions, realizing a distributed interference effect evaluation mechanism.

[0095] As an example, in step 4, a distributed interaction mechanism is designed to realize the interaction mechanism between the intelligent agent and the engagement domain under the distributed interaction mechanism.

[0096] In step S41, the interaction event is initialized. When satellite intelligent agent A initiates electromagnetic interference, at this time satellite intelligent agent A is the interference source, and it first sends a signal to the electromagnetic engagement domain related to satellite intelligent agent A. The signal contains detailed information about the interference source, including the position coordinates of the interference source, motion state, characteristics of the interference signal (transmission power, operating frequency, modulation method, bandwidth, etc.), information about the target intelligent agent (target position, communication frequency band, etc.), and the interference duration parameter.

[0097] In step S42, the state of the engagement domain is updated. After receiving the signal, the electromagnetic engagement domain related to satellite intelligent agent A first determines the affected area range. According to the positional relationship between the interference source and the target satellite intelligent agent and combining the propagation characteristics of electromagnetic waves, the calculation grid that needs to be updated is calculated. For the interference effects across regions, the relevant electromagnetic engagement domains perform data exchange through the boundary processing mechanism to maintain the continuity and consistency of the electromagnetic engagement domain at the regional boundary.

[0098] In step S43, the signal propagation is calculated. After determining the scope of the affected area, within the determined affected range, the electromagnetic engagement domain calculates the propagation path and intensity attenuation of the interference signal according to the spatial propagation model. Among them, the spatial propagation model includes factors such as free space loss, atmospheric attenuation, and multipath effect, and at the same time combines the influence of terrain, buildings and other obstacles in the scene on the electromagnetic wave propagation. For complex propagation paths, high-precision algorithms such as ray tracing are used for simulation.

[0099] In step S44, the engagement domain effect is superimposed. After calculating the propagation path and intensity attenuation of the interference signal of the single interfering source satellite agent A, the electromagnetic engagement domain needs to consider the influence of other electromagnetic sources existing in the scene. By the principle of electromagnetic field superposition, the combined field strength of each other electromagnetic source is calculated. For multiple interfering sources with similar frequencies, their coherence and intermodulation effects also need to be considered. Finally, the results of the updated electromagnetic engagement domain are reflected in real time on each node of the calculation grid, and the distribution state of the updated engagement domain is obtained.

[0100] In step S45, the target impact is evaluated. Based on the distribution state of the updated electromagnetic engagement domain, the electromagnetic engagement domain evaluates the impact of the interference signal on the communication system of the target satellite agent B. The evaluation content includes key indicators such as the change in signal-to-noise ratio of the received signal, the increase in bit error rate, and the decrease in communication link capacity. At the same time, considering the anti-interference ability of the target satellite agent B, such as characteristics of antenna pattern, receiver dynamic range, etc., the final interference effect evaluation result is obtained.

[0101] In step S46, the distributed event is fed back. The electromagnetic engagement domain sends the evaluation result to the relevant satellite agents through the distributed event bus. Among them, the evaluation result includes detailed information about the interference effect, such as parameters of signal-to-noise ratio, available bandwidth, link quality, etc. of each communication channel. After receiving this information, the target satellite agent B triggers corresponding response actions, such as switching to an alternative communication frequency band, adjusting the transmission power, starting the anti-interference mode, etc.

[0102] In step S47, the state consistency is maintained. To ensure the state consistency of each electromagnetic engagement domain under the distributed interaction mechanism, a distributed state synchronization mechanism is adopted. Among the electromagnetic engagement domains, through regular state verification and incremental update, the consistency of the electromagnetic engagement domain values is maintained. Further, the interaction state between the satellite agent and the electromagnetic engagement domain is also synchronized through the event log and state snapshot mechanism to ensure the coherence and accuracy of the entire multi-dimensional engagement domain interactive digital simulation process.

[0103] As an example, in step S5, the distributed interaction mechanism is optimized.

[0104] As an example, in view of the characteristics of electromagnetic engagement domain calculation under the distributed interaction mechanism, a multi-level optimized distributed interaction mechanism is implemented. First, the adaptive grid technology is adopted. A finer grid is used around the interfering source satellite agent A and the target satellite agent B, while a coarser grid is used in the distant area to improve the calculation efficiency. Further, the numerical calculation tasks of the large-scale electromagnetic engagement domain are decomposed into multiple subtasks and assigned to different computing nodes in the electromagnetic engagement domain for parallel processing to achieve dynamic allocation of the calculation tasks. Finally, a distributed caching mechanism is established to cache the static or slowly changing scene features, reduce cross-node data transmission and repeated calculations, and improve the calculation efficiency.

[0105] As an example, in step S6, the distributed effect is evaluated and feedback is provided.

[0106] As an example, an evaluation and feedback mechanism under the distributed interaction mechanism is implemented. After the evaluation results of all electromagnetic engagement domains in the scene are obtained, first, the impact on satellite agent communication is evaluated locally, including indicators such as signal-to-noise ratio degradation and increased bit error rate. Then, the evaluation results are notified to the relevant satellite agents through the distributed event bus to trigger corresponding response actions, such as switching communication frequency bands and starting anti-jamming modes. Finally, through the distributed state synchronization mechanism, the evaluation results of all electromagnetic engagement domains in the scene are integrated, and full-scene-level response and state synchronization are achieved through the distributed event bus to ensure that the scene states in all regions are consistent.

[0107] As an example, the performance of the electromagnetic engagement domain in this embodiment under the multi-dimensional engagement domain interactive digital simulation engine is shown in Table 1.

[0108] Table 1 Performance table of the multi-dimensional engagement domain interactive digital simulation engine in the first specific embodiment In this embodiment, the electromagnetic engagement domain adopts the multi-dimensional engagement domain interactive digital simulation engine construction method of this application, realizing ultra-large-scale distributed simulation capabilities and supporting the collaborative simulation of tens of thousands of satellite agents. A flexible intelligent agent distributed dynamic deployment mechanism is established, which can optimize the deployment plan according to the characteristics of hardware resources. Efficient engagement domain partition management is achieved, supporting parallel simulation of electromagnetic confrontation by satellite agents in multiple regions. A stable distributed interaction mechanism is provided to ensure data consistency in a large-scale distributed environment. Through multi-level parallel computing and optimization mechanisms, the simulation efficiency of large-scale scenes is significantly improved.

[0109] In the second specific embodiment of the present application, surface ships, submarines, and underwater robots are used as agents. Taking underwater acoustic countermeasure as an example, in the scenario, three types of agents, namely surface ships, submarines, and underwater robots, exchange information and update their states through an acoustic engagement domain, demonstrating the application of the construction method of a multi-dimensional acoustic engagement domain interactive digital simulation engine in a complex marine environment. The multi-dimensional acoustic engagement domain interactive digital simulation process requires accurate simulation of the propagation characteristics of sound waves underwater, including thermocline effects, multi-path propagation, seabed terrain reflection, etc. At the same time, environmental factors such as ocean currents, temperature, and salinity that affect the propagation of sound waves also need to be considered.

[0110] Furthermore, an operator is set to perform real-time intervention and control on the agents in the scenario, observe the scenario through a three-dimensional visualization interface, adjust sonar parameters in real time, or trigger specific tactical actions to enhance the authenticity and interactivity of the multi-dimensional engagement domain interactive digital simulation.

[0111] As an example, the system adopts a distributed deployment architecture, and deploys functional modules such as marine environment simulation, acoustic calculation, and agent behavior control on different servers. Among them, the calculation of the acoustic engagement domain is responsible for by a dedicated high-performance computing cluster, and the computing resources can be dynamically expanded according to needs.

[0112] As an example, in step S1, a distributed engine is constructed.

[0113] In step S11, standard interfaces for surface ship agents, submarine agents, and underwater robot agents are defined, including the basic attributes and behaviors of surface ship agents, submarine agents, and underwater robot agents. Among them, the basic attributes include: position, speed, acoustic characteristics; the behaviors include: maneuvering, sonar detection, and acoustic countermeasure.

[0114] In step S12, a basic framework of the acoustic engagement domain is established to realize the spatial and computational grid division of the engagement domain. In this embodiment, it also includes modeling the three-dimensional ocean space. The modeling of the three-dimensional ocean space includes: first, constructing an ocean hydrological model, which includes parameters such as temperature stratification, salinity distribution, and ocean current field; then constructing a seabed terrain model, which includes parameters such as terrain undulation and bottom sediment characteristics; finally, constructing an environmental noise model, which includes parameters such as sea surface noise, biological noise, and ship noise. These environmental factors will all affect the propagation characteristics of sound waves in the ocean space, and the operator can adjust the environmental parameters in real time and observe their impact on acoustic countermeasures.

[0115] In step S13, a distributed scheduling method is designed to coordinate the running timings of components on the nodes of each server.

[0116] In the second specific embodiment, after step S13, it also includes designing an interaction control method to process the real-time instructions of the operator.

[0117] As an example, in step S2, design an intelligent agent distributed dynamic deployment mechanism.

[0118] After designing a distributed scheduling method to coordinate the running timings of components on the nodes of each server, based on the scenario requirements and the computing capabilities of the servers, design an intelligent agent distributed dynamic deployment mechanism to achieve the distributed dynamic deployment of surface ship agents, submarine agents, and underwater robot agents.

[0119] In step S21, establish a server performance evaluation model. The server performance evaluation model includes metrics such as computing power, memory capacity, and network bandwidth. The server performance evaluation results are divided into three levels: high-performance servers (computing power > 80%, memory usage rate > 70%, network bandwidth > 90%), medium-performance servers (computing power 50 - 80%, memory usage rate 40 - 70%, network bandwidth 60 - 90%), and low-performance servers (computing power < 50%, memory usage rate < 40%, network bandwidth < 60%).

[0120] In step S22, use the server performance evaluation model to achieve intelligent agent load balancing. Dynamically allocate surface ship agents, submarine agents, and underwater robot agents according to the evaluation results of the server performance evaluation model, and allocate surface ship agents, submarine agents, and underwater robot agents to servers with lighter loads. For high-performance servers, preferentially allocate computationally intensive underwater robot agents, 250 - 300 per node; for medium-performance servers, mainly allocate submarine agents, 150 - 200 per node; for low-performance servers, mainly allocate surface ship agents, 100 - 150 per node.

[0121] In the second specific embodiment, it further includes step S23: The operator can add or remove agents in real time through the control interface to dynamically adjust the scale of the multi-dimensional acoustic engagement domain interactive digital simulation.

[0122] As an example, in step S3, implement engagement domain partition management.

[0123] In step S31, according to the characteristics of large-scale scenarios, divide the entire scenario space into multiple computing regions, and each region is responsible for by an independent engagement domain.

[0124] In step S32, design a boundary processing mechanism between regions to ensure the continuity and consistency of the engagement domain at the region boundaries.

[0125] In step S33, design a cooperation mechanism between engagement domains to handle the electromagnetic interference effects across regions.

[0126] As an example, in step S4, a distributed interaction mechanism is designed to implement the interaction mechanism between the agent and the engagement domain under the distributed interaction mechanism.

[0127] In step S41, the agent in the scenario emits a signal, which contains the detailed information of the agent, including: the position coordinates of the agent, the motion state, the signal characteristics, the information of the target agent, and the duration parameter. The operator can observe the scenario through the 3D visualization interface and can select any agent for control. The 3D visualization interface displays the status information of the surface ship agent, the submarine agent, and the underwater robot agent in real time, including the position, speed, sonar working mode, etc. The operator can adjust the sonar parameters through the interface, such as the transmit power, working frequency, beam direction, etc., and send these control instructions to the corresponding agent.

[0128] In step S42, when the operator triggers the sonar emission or the agent autonomously conducts sonar detection, the system generates an acoustic event. The acoustic event information contains the detailed parameters of the sound source, such as the sound source position, the sound source level, the acoustic wave characteristics (frequency, waveform, etc.), the emission direction, etc., and sends this acoustic event information to the agent related to the acoustic engagement domain. After the engagement domain related to the agent that emits the signal receives the signal, it determines the affected area range and calculates the computational grid that needs to be updated.

[0129] In step S43, after the acoustic engagement domain receives the acoustic event, it first determines the computational area range. Based on the sound source characteristics and environmental conditions, algorithms such as ray tracing are used to calculate the propagation path of the acoustic engagement domain. The process of calculating the propagation path of the acoustic engagement domain takes into account effects such as thermocline refraction, seabed reflection, and multipath propagation, and at the same time combines the real-time state of the ocean environment to ensure the accuracy of the calculation results, and obtains the distribution result of the acoustic engagement domain.

[0130] In step S44, while calculating the distribution of the acoustic engagement domain, the influence of environmental factors is evaluated. The environmental factors include: the influence of ocean currents on the bending of acoustic rays, the influence of temperature stratification on the acoustic propagation loss, the influence of seabed topography on the acoustic reflection, etc. These environmental factor influences will be reflected in the calculation results of the acoustic engagement domain distribution in real time, and the operator can intuitively observe these effects through the 3D visualization interface.

[0131] In step S45, based on the updated distribution state of the acoustic engagement domain, the sound pressure level of the sound wave reaching each agent position is calculated. Combining the sonar performance parameters (such as sensitivity, directivity, etc.) of the surface ship agent, the submarine agent, and the underwater robot agent, it is evaluated whether the sound wave can detect the target agent. At the same time, considering the influence of the background environmental noise, the signal-to-noise ratio is calculated, and the evaluation result is obtained, and these evaluation results are displayed on the operator's 3D visualization interface in real time.

[0132] In step S46, the operator sends tactical instructions to the corresponding surface ship agent, submarine agent, and underwater robot agent through the distributed event system according to the evaluation results. The surface ship agent, submarine agent, and underwater robot agent execute corresponding response actions and feedback the execution results. Among them, the response actions include: adjusting the course to avoid detection, changing the sonar working mode for anti-detection, launching acoustic decoys for countermeasure, etc.

[0133] In step S47, to ensure the state consistency of each electromagnetic engagement domain under the distributed interaction mechanism, a distributed state synchronization mechanism is adopted to maintain the consistency of the scene states of all agents, acoustic engagement domain values, and user interfaces in the scene. Among them, it includes information such as the positions, postures, sonar states, etc. of the surface ship agent, submarine agent, and underwater robot agent, as well as information such as the acoustic engagement domain distribution results and environmental states. As an example, the operator can observe the changes in the scene in real time through the three-dimensional visualization interface and intervene and control as needed.

[0134] As an example, in step S5, the distributed interaction mechanism is optimized.

[0135] As an example, aiming at the characteristics of acoustic engagement domain calculation under the distributed interaction mechanism, a multi-level optimization mechanism is implemented. First, the adaptive grid technology is adopted. In key areas, such as around the sound source and near the thermocline, finer grid divisions are used, and coarser grids are used in distant areas to improve the calculation efficiency. Further, the acoustic engagement domain distribution calculation task is decomposed into multiple subtasks for parallel processing to achieve dynamic allocation of calculation tasks. Finally, a distributed caching mechanism is established to cache the environmental characteristics and sound propagation characteristics, reduce cross-node data transmission and repeated calculations, and improve the calculation efficiency.

[0136] In the second specific embodiment, after step S5, it further includes optimizing the human-computer interaction mechanism.

[0137] As an example, the three-dimensional visualization interface of the operator adopts a multi-view design, including a three-dimensional scene view, a sound field distribution view, an agent state view, etc., and supports various interaction methods, such as mouse clicking, gesture operations, shortcut key controls, etc. Further, a prediction mechanism is adopted to optimize the interaction response speed, predict possible results when the operator executes control, perform calculations in advance, reduce interaction latency, and achieve an efficient human-computer interaction mechanism.

[0138] As an example, in step S6, the distributed effect is evaluated and feedback is provided to implement the evaluation and feedback mechanism under the distributed interaction mechanism.

[0139] As an example, the performance of the electromagnetic engagement domain in this embodiment under the multi-dimensional engagement domain interactive digital simulation engine is shown in Table 2.

[0140] Table 2 Performance Table of the Multidimensional Engagement Domain Interactive Digital Simulation Engine in the Second Specific Embodiment In this embodiment, the acoustic engagement domain is constructed by using the multidimensional engagement domain interactive digital simulation engine construction method of the present application, realizing acoustic countermeasure simulation in a complex marine environment and supporting accurate simulation of various acoustic phenomena. A flexible human-computer interaction mechanism is established to support real-time intervention and control by operators. An efficient distributed computing framework is realized, significantly improving the efficiency of sound field calculation. An intuitive visualization interface is provided, enhancing the usability and practicality of the simulation system. Through a multi-level optimization mechanism, a smooth interaction experience in a large-scale scenario is achieved.

[0141] In the third specific embodiment of the present application, multiple types of detection platforms such as land-based phased array radars, sea-based shipborne radars, and space-based early warning radars are used as agents. Taking the collaborative detection of land-based phased array radars, sea-based shipborne radars, and space-based early warning radars as an example, the application of the construction method of the multidimensional engagement domain interactive digital simulation engine in a complex detection scenario is demonstrated. The scenario includes multiple types of detection platforms such as land-based phased array radar agents, sea-based shipborne radar agents, and space-based early warning radar agents, which collaborate to detect and track air and space targets. The multidimensional engagement domain interactive digital simulation process needs to accurately simulate the propagation characteristics of radar waves in different spatial environments, including atmospheric refraction, ionospheric effects, sea surface reflection, etc., and at the same time consider the influence of factors such as target characteristics, meteorological conditions, and geographical environment on the detection effect.

[0142] To achieve accurate collaborative detection simulation, a hierarchical distributed architecture is adopted. The detection tasks of different types of radar agents are assigned to the computing nodes of dedicated servers. For example, the ionospheric propagation calculation of space-based early warning radars is responsible for by the high-altitude physical simulation server cluster, and the sea surface clutter calculation of sea-based shipborne radars is processed by the ocean environment simulation server cluster. Information sharing and collaborative processing are realized between each engagement domain through a data fusion mechanism.

[0143] As an example, in step S1, a distributed engine is constructed.

[0144] In step S11, the standardized interfaces and a unified coordinate system of land-based phased array radar agents, sea-based shipborne radar agents, and space-based early warning radar agents are defined. The interface design follows the modular principle, supporting the plug-and-play of multiple types of radar agents and improving the scalability of the construction method of the multidimensional engagement domain interactive digital simulation engine.

[0145] In step S12, a physical model of the spatial engagement domain is constructed. First, a fast spatial relationship calculation engine is established. The fast spatial relationship calculation engine adopts an octree spatial index structure and parallel computing technology. The octree spatial index structure recursively divides the detection space into uniform subspaces, and each server node stores the spatial positions and attribute information of target agents. The parallel computing technology distributes the spatial relationship calculation tasks to multiple GPU computing cores to achieve millisecond-level spatial queries and collision detections, ultimately realizing efficient spatial relationship calculations. At the same time, a target motion model is constructed. The target motion model combines the Singer-α-β filter and the interacting multiple model algorithm (IMM) to predict the trajectories of air and space target agents such as aircraft, missiles, and satellites detected by land-based phased array radar agents, sea-based shipborne radar agents, and space-based early warning radar agents. Among them, the Singer-α-β filter is used to handle the maneuvering characteristics of the target, and the IMM algorithm improves the trajectory prediction accuracy through the adaptive switching of multiple motion models (uniform motion, uniformly accelerated motion, coordinated turning motion, etc.). The target motion model automatically selects an appropriate combination of motion models according to the target agent type and motion characteristics to achieve precise tracking of targets with different maneuvering characteristics, providing real-time geometric information support for the electromagnetic engagement domain.

[0146] As an example, a physical model of the electromagnetic engagement domain is constructed to realize the propagation simulation of radar waves in multi-dimensional space based on physical laws. In the near-surface layer, a terrain masking model is established, adopting the improved GTD diffraction theory and combining with the terrain database to calculate the electromagnetic wave diffraction loss; a multipath effect model, based on ray tracing and specular reflection principles, considering the frequency and incident angle dependence of the ground reflection coefficient; a ground clutter model, adopting the improved distribution model (Gamma) and introducing terrain undulation and surface characteristic parameters. In the ocean layer, a sea surface reflection model is constructed, based on the improved two-dimensional spectrum method and considering the influence of the sea wave spectrum and wind speed; a sea clutter model, adopting the composite K distribution model and combining sea conditions and meteorological parameters; an atmospheric duct model, based on the parabolic equation method and considering the dynamic changes of the atmospheric refractive index profile. In the atmosphere layer, an atmospheric refraction model is developed, adopting a piecewise linear refractive index profile to realize tropospheric refraction modeling; an ionospheric dispersion model, based on the improved Appleton-Hartree equation and considering the influence of electron density and geomagnetic field; a space plasma model, adopting magnetohydrodynamic equations to simulate the influence of high-altitude plasma on electromagnetic waves.

[0147] In the third specific embodiment, after step S12, it also includes: adopting the engagement domain coupling mechanism to realize the bidirectional coupling between the space engagement domain and the electromagnetic engagement domain. An interactive interface between the space engagement domain and the electromagnetic engagement domain is established. The interactive interface adopts an asynchronous communication mechanism based on a message queue, and realizes data synchronization through timestamps and version numbers. The interactive interface defines a standardized data exchange format, which contains key information such as spatial position, physical field strength, and target characteristics. Distributed locks and transaction mechanisms are used to ensure data consistency. When the engagement domain changes, the spatial relationship data is updated through atomic operations, and vice versa. At the same time, a cache mechanism is introduced to locally cache frequently accessed field data to reduce cross-domain access overhead. The interactive interface also implements data verification and fault-tolerant processing, ensures the reliability of the calculation process through heartbeat detection and data verification, and ensures the synchronization and consistency of the relationship calculation between the electromagnetic engagement domain calculation and the space engagement domain calculation.

[0148] As an example, based on the design concept of multi-dimensional combat domain, the entire scene is divided into land layer, ocean layer, atmosphere layer and space layer, and corresponding radar agents are deployed in each layer, including land-based phased array radar agents, sea-based shipborne radar agents in the ocean layer and space-based early warning radar agents in the atmosphere layer. A hierarchical grid structure is used to finely model the detection space, and adaptive grid density is used at different levels to balance calculation accuracy and efficiency.

[0149] As an example, the hierarchical grid structure used in this embodiment includes: dividing the entire scene into several large-scale grid units, each unit size is 10km×10km×1km, and then adaptively subdividing the computing grid according to the physical characteristics and computing requirements of different levels. In the ground layer, the adaptive mesh technology (AMR, Adaptive Mesh Refinement) based on quadtree is adopted to dynamically adjust the mesh density according to the complexity of the terrain, and the resolution can be subdivided to 100m in key terrain areas; in the ocean layer, the adaptive mesh technology based on octree automatically adjusts the mesh density in combination with the gradient of sea state changes, and the resolution of key sea areas can reach 500m; in the atmosphere, an adaptive encryption strategy based on physical quantity gradient is adopted to perform local encryption in areas where electromagnetic wave propagation changes dramatically (such as the atmospheric waveguide layer), and the grid size smoothly transitions from 1km at low altitude to 5km at high altitude; in the space layer, due to the relatively uniform physical characteristics, a uniform grid of 20km is adopted.

[0150] In step S13, a distributed scheduling method is designed to coordinate the running sequence of components on the nodes of each server.

[0151] As an example, in step S2, a distributed dynamic deployment mechanism of intelligent agents is designed.

[0152] In step S21, a server performance evaluation model is established.

[0153] In step S22, satellite agents are dynamically allocated according to the evaluation results of the server performance evaluation model. During the operation, the load conditions of each server are continuously monitored, and the computing resource allocation is adaptively adjusted according to the target agent density in the spatial region to ensure the computing efficiency.

[0154] In the third specific embodiment, step S2 further includes: step S23, establishing a multi-dimensional environment database.

[0155] As an example, a multi-dimensional environment database is established. The multi-dimensional environment database includes dynamic environment information such as terrain undulation, sea state changes, and meteorological conditions. The environmental data adopts a distributed storage architecture, supporting real-time updates and fast access, and providing accurate environmental parameters for radar wave propagation calculation.

[0156] In the third specific embodiment, step S2 further includes: step S24, establishing a radar agent model to achieve standardized modeling of different types of radar agents.

[0157] As an example, a land-based phased array radar agent model is established, a digital twin model of the land-based phased array radar agent is constructed to achieve accurate simulation of the antenna array characteristics, and a beamformer model is mainly established, including functional modules such as digital beamforming, adaptive beam control, and space-time signal processing. At the same time, an environmental impact model is constructed, and the environmental impact model includes the influence of factors such as terrain occlusion and ground clutter on the detection performance of the land-based phased array radar agent.

[0158] As an example, a sea-based shipborne radar agent model is established, a dynamic characteristic model of the sea-based shipborne radar agent is constructed, the influence of the movement of the radar agent on the radar performance is mainly considered, a sea surface reflection model and a sea clutter model are established to achieve accurate simulation of various detection modes such as sea, air, and space; at the same time, a ship body attitude compensation model is constructed to improve the detection accuracy.

[0159] As an example, a space-based early warning radar agent model is established, a space warning radar model is constructed, the ionospheric propagation characteristics are mainly concerned, an ionospheric propagation model including factors such as atmospheric attenuation and space debris interference is established to achieve effective detection of detection targets inside and outside the atmosphere; at the same time, an orbital dynamics model is constructed to ensure the continuity of detection coverage.

[0160] In the third specific embodiment, step S2 further includes: step S25, designing a radar agent cooperation mechanism: based on the standardized interface specifications of land-based phased array radar agents, sea-based shipborne radar agents, and space-based early warning radar agents, a task allocation model, a resource scheduling model, and a data sharing model are established to support cooperative detection of multiple radar agents and realize the cooperative working mechanism between different radar agents.

[0161] As an example, in step S25, a collaborative detection mechanism is designed to implement the collaborative working mechanism of multiple radar agents, including steps S251 to S253.

[0162] Step S251: Design a task planning mechanism.

[0163] As an example, a detection task planning mechanism based on task priority is established, a task decomposition model is constructed, and complex detection tasks are decomposed into multiple subtasks.

[0164] Furthermore, a task scheduling algorithm is developed to achieve optimal task allocation according to the performance characteristics and spatial positions of land-based phased array radar agents, sea-based shipborne radar agents, and space-based early warning radar agents. According to the performance parameters such as the detection range, angle coverage, number of beams, and signal processing capabilities of land-based phased array radar agents, sea-based shipborne radar agents, and space-based early warning radar agents, as well as the spatial information such as the geographical location, altitude, and attitude of land-based phased array radar agents, sea-based shipborne radar agents, and space-based early warning radar agents, an improved Hungarian algorithm is used to optimize task allocation. The improved Hungarian algorithm first constructs a task-radar efficiency matrix, and the matrix elements are calculated by weighted fusion of indicators such as detection probability, measurement accuracy, and response time; then, based on the Branch and Bound method, the optimal matching is solved to realize the mapping relationship between tasks and each radar agent. At the same time, a dynamic weight mechanism is introduced to adaptively adjust the weights of each indicator according to factors such as the target threat level and task urgency, ensuring that critical tasks are preferentially assigned to the most suitable radar agent. Furthermore, a task conflict detection and coordination mechanism is established. When multiple tasks compete for the resources of the same radar agent, arbitration is carried out through task priority and timing constraints to achieve optimal task allocation.

[0165] Step S252: Data fusion processing.

[0166] As an example, a fusion processing system for multi-source detection data is constructed. First, a data preprocessing model is established to achieve noise filtering and clutter suppression; further, a spatio-temporal alignment algorithm is developed. Through timestamp alignment, GPS time synchronization and Network Time Protocol (NTP) are used to synchronize the clocks of each radar agent, and the data at different sampling times are interpolated to a unified time reference; coordinate system transformation, a unified spatial reference is established based on the WGS84 coordinate system, and the mapping from the local coordinate systems of different radar agents to the unified coordinate system is realized through a coordinate transformation matrix; attitude correction, using inertial navigation data and attitude measurement device data to compensate for the measurement errors caused by platform movement; geometric registration, adopting feature point matching and image registration technologies to achieve the spatial alignment of observation data from different perspectives and other steps to achieve the unified registration of data from different radar agents; finally, a multi-source data fusion algorithm is adopted. The multi-source data fusion algorithm adopts a three-layer fusion architecture: the first layer is data layer fusion, using an improved Kalman filter to achieve the optimal estimation of measurement data, and introducing an adaptive factor to dynamically adjust the measurement noise covariance; the second layer is feature layer fusion, extracting and fusing target feature information based on a combined method of fuzzy inference and evidence theory, including motion features, scattering features, and electromagnetic features, etc.; the third layer is decision layer fusion, adopting an attention mechanism based on deep learning to perform weighted fusion on the detection results of different sensors, and the weight coefficients are optimized in real time through an online learning method. Finally, unified target tracks and situation information are output.

[0167] Step S253: Cooperative control optimization.

[0168] As an example, cooperative control optimization is designed to achieve dynamic allocation of detection resources. A beam management model is constructed to optimize the beam pointing of multiple radar agents; a working mode switching strategy is established to adapt to different detection scenarios; a resource allocation algorithm is established to maximize the detection efficiency.

[0169] As an example, in step S3, area management of the engagement domain is realized.

[0170] In step S31, the entire scene space is divided into multiple calculation regions, and each region is responsible for by an independent engagement domain.

[0171] In step S32, a boundary processing mechanism between regions is designed to ensure the continuity and consistency of the engagement domain at the region boundary.

[0172] In step S33, a cooperation mechanism between engagement domains is designed to handle the electromagnetic interference effects across regions.

[0173] As an example, in step S4, a distributed interaction mechanism is designed to realize the interaction mechanism between agents and the engagement domain under the distributed interaction mechanism.

[0174] In step S41, the interaction event is initialized.

[0175] In step S42, the state of the engagement domain is updated. After the engagement domain related to the agent that emits the signal receives the signal, the affected area range is determined. The calculation error and resource occupancy are monitored in real time, and the grid density of each layer is dynamically adjusted to optimize the calculation efficiency while ensuring the calculation accuracy.

[0176] In step S43, the signal propagation is calculated.

[0177] In step S44, an engagement domain effect superposition model is constructed to realize the comprehensive simulation of multiple physical effects.

[0178] In step S45, the target impact is evaluated.

[0179] In step S46, the distributed event is fed back.

[0180] In step S47, the state consistency is maintained.

[0181] As an example, in step S5, according to the characteristics of the engagement domain calculation under the distributed interaction mechanism, a multi-level optimized distributed interaction mechanism is implemented. When the detected target enters a certain area or the physical quantity gradient changes violently, the grid density of this area is increased; when the detected target leaves or the physical quantity tends to be stable, the grid density is appropriately reduced to save computing resources.

[0182] In the third specific embodiment, step S5 further includes: designing a parameter optimization mechanism.

[0183] As an example, a system parameter optimization framework is established. The optimization of radar agent deployment adopts an improved Particle Swarm Optimization (PSO) algorithm. The radar deployment problem is modeled as a multi-objective optimization problem. The objective function includes three indicators: detection coverage rate, system redundancy, and deployment cost. By introducing an adaptive inertia weight and a mutation operator, the convergence performance of the algorithm is improved, and the detection coverage efficiency is enhanced. A beam allocation optimization model is constructed. The beam allocation optimization model is based on the deep reinforcement learning method. A Markov Decision Process (MDP) model including a state space (target position, speed, priority, etc.), an action space (beam pointing, power allocation, revisit interval, etc.), and a reward function (tracking accuracy, resource utilization, etc.) is constructed. The Double Deep Q-Network (Double DQN) algorithm is used to achieve the dynamic allocation of beam resources and improve the target tracking accuracy. A data fusion parameter optimization mechanism is established. The data fusion parameter optimization mechanism adopts an online learning method based on gradient descent. An optimization model including key parameters such as data association threshold, filtering gain, and feature weight is constructed. The adaptive adjustment of parameters is realized by minimizing the fusion error. The performance prediction model adopts a Long Short-Term Memory (LSTM) network. By inputting the historical performance data sequence, the predicted value of the performance index at the future moment is output. Key temporal features are extracted through the attention mechanism to accurately predict the performance of the radar agent, enhance the situation awareness ability, and support the trend analysis and decision-making of the detection efficiency.

[0184] As an example, in step S6, the distributed effect is evaluated and feedback is provided to implement the evaluation and feedback mechanism under the distributed interaction mechanism. The evaluation content includes a detection probability model, a measurement accuracy model, and a tracking stability model. A real-time evaluation engine is established to support the dynamic calculation and statistical analysis of performance indicators. The evaluation results are divided into detection performance evaluation results (including detection probability, false alarm probability, detection distance, etc.), measurement performance evaluation results (including distance measurement accuracy, speed measurement accuracy, angle measurement accuracy, etc.), and tracking performance evaluation results (including track establishment time, track loss rate, track intersection handling ability, etc.).

[0185] In the third specific embodiment, step S6 further includes: evaluating the collaborative detection efficiency.

[0186] As an example, an information gain model is established to quantify the performance improvement brought by collaborative detection; a coverage analysis model is constructed to evaluate the overlap and blind areas of the detection area; a threat assessment model is developed to analyze the anti-jamming ability of the system. The evaluation results of the collaborative detection efficiency are divided into information gain evaluation results (including information entropy gain, target recognition rate improvement, situation awareness ability improvement, etc.), coverage performance evaluation results (including spatial coverage rate, time coverage rate, detection blind area distribution, etc.), and anti-jamming ability evaluation results (including anti-electronic jamming ability, anti-deception ability, system survival ability, etc.).

[0187] As an example, the performance of the electromagnetic engagement domain in this embodiment under the multi-dimensional engagement domain interactive digital simulation engine is shown in Table 3.

[0188] Table 3 Performance Table of the Multi-Dimensional Engagement Domain Interactive Digital Simulation Engine in the Third Specific Embodiment In this embodiment, the collaborative detection simulation of land-sea-air integration is realized, supporting the collaboration of multi-radar agents in complex environments; an accurate radar wave propagation model is established to realize the propagation characteristic simulation of different spatial regions; an efficient data fusion mechanism is provided, significantly improving the collaborative detection efficiency; a complete performance evaluation system is realized to support system optimization and decision-making support; through the distributed computing framework, the real-time performance in large-scale scenarios is guaranteed.

[0189] In the fourth specific embodiment of this application, multi-type combat platforms such as satellites, airborne aircraft, ground armored vehicles, naval ships, and underwater submarines are used as agents. Taking the multi-domain collaborative combat of land-sea-air-space as an example, the application of the construction method of the multi-dimensional engagement domain interactive digital simulation engine in a complex battlefield environment is demonstrated. The scenario is set to include multi-type combat platforms such as space satellites, airborne aircraft, ground armored vehicles, naval ships, and underwater submarines, which conduct collaborative combat in a unified battlefield environment. The multi-dimensional engagement domain interactive digital simulation process needs to simulate the propagation characteristics and their interactions of multi-physical fields such as electromagnetic, acoustic, and optical, while considering the influence of environmental factors such as the atmosphere, ocean, and terrain.

[0190] To achieve such complex multi-domain collaborative simulation, the system adopts a hierarchical distributed architecture. The simulation tasks of different combat domains are assigned to dedicated computing server clusters. For example, the space surveillance and early warning tasks are responsible for by the space simulation server cluster, and the underwater confrontation is processed by the ocean simulation server cluster. Each physical field engagement domain is also deployed on independent high-performance computing nodes to achieve the collaborative calculation of multi-physical fields through the field coupling mechanism.

[0191] As an example, in step S1, a distributed engine is constructed.

[0192] In step S11, the standardized interfaces of agents such as satellite agents, airborne aircraft agents, ground armored vehicle agents, naval ship agents, and underwater submarine agents are defined. The standardized interfaces include: basic attributes such as position, attitude, speed, etc.; general behaviors such as maneuvering, detecting, and striking.

[0193] Then, specific attributes and capabilities are extended for different types of agents, such as the orbital parameters of satellites, the aerodynamic characteristics of aircraft, and the hydrodynamic characteristics of ships.

[0194] In step S12, a basic framework of the multi-physical field engagement domain is established to achieve the spatial and computational grid division of the engagement domain. First, a three-dimensional battlefield environment including land, sea, air, and space is constructed. In the spatial layer, the orbital motion and space surveillance of satellite agents are realized. In the atmosphere layer, the maneuvers of aerial agents such as aircraft agents and missile agents are simulated. In the surface layer, static elements such as terrain and buildings and dynamic targets such as ground armored vehicle agents are established. In the ocean layer, the activities of surface ship agents and submarine agents are realized. Information exchange is achieved between each layer through a unified coordinate system and data interface.

[0195] As an example, the multi-physical field engagement domain includes multiple physical field engagement domains such as space, electromagnetic, acoustic, and optical, and realizes the coupled calculation between the multi-physical field engagement domains.

[0196] Among them, the physical model for constructing the optical engagement domain includes: realizing the propagation characteristics of electromagnetic waves such as visible light and infrared, which are used for optoelectronic reconnaissance, infrared guidance, etc. Based on the scene organization of the space engagement domain, considering factors such as atmospheric attenuation, cloud cover, and day-night changes, the propagation effect of optical signals is calculated.

[0197] In the fourth specific embodiment, after step S12, it further includes: designing a field domain coupling calculator and establishing an interaction model between the physical field engagement domains. The space engagement domain serves as the basic support, providing a unified space reference and target state information for other physical field engagement domains. For example, the reflection and scattering effects generated when electromagnetic waves encounter the sea surface will affect the radar detection effect, which requires the space engagement domain to provide accurate sea surface state and target position information. Sound waves may be affected by electromagnetic interference during propagation, which requires the space engagement domain to coordinate the spatial relationships between different physical field engagement domains. These complex physical interactions are realized through the field domain coupling calculator.

[0198] As an example, the field domain coupling calculator adopts a hierarchical coupling architecture, including a physical effect layer, a data exchange layer, and a control and coordination layer. The physical effect layer realizes the interaction calculation between each physical field engagement domain, such as the reflection and scattering model of electromagnetic waves and the sea surface, and the mutual influence model of sound waves and electromagnetic interference. The data exchange layer is responsible for data transmission between different engagement domains, adopting a combination of shared memory and message queue to ensure the efficient exchange of large-scale data. The control and coordination layer realizes the timing synchronization and calculation coordination between the engagement domains, and ensures the causal consistency of multi-physical field calculations through a token passing mechanism.

[0199] Compared with the engagement domain coupling mechanism of the third specific embodiment, the field domain coupling mechanism of this embodiment focuses more on the accurate calculation of complex physical effects, adopts a finer-grained coupling model and a more rigorous synchronization mechanism to support complex physical interactions in multi-domain operations on land, sea, air, and space. For example, the reflection and scattering effects generated when electromagnetic waves encounter the sea surface will affect the radar detection effect, which requires the spatial engagement domain to provide accurate sea surface state and target position information. Sound waves may be affected by electromagnetic interference during propagation, which requires the spatial engagement domain to coordinate the spatial relationships between different physical fields. Through this multi-level coupling mechanism, accurate interaction calculations between different physical fields are achieved.

[0200] In step S13, a distributed scheduling method is designed to coordinate the operation timing of components on the nodes of each server.

[0201] As an example, in step S2, an intelligent agent distributed dynamic deployment mechanism is designed.

[0202] In step S21, a server performance evaluation model is established. Calculate the hierarchical levels according to the engagement domains of each physical field, specifically divided into the following levels: the space layer, responsible for satellite intelligent agent orbit calculation, space target intelligent agent tracking, etc.; the atmosphere layer, responsible for simulating air intelligent agents such as aircraft intelligent agents and missile intelligent agents; the surface layer, responsible for calculating the movement of ground target intelligent agents; the ocean layer, responsible for simulating surface ship intelligent agents and submarine intelligent agents; the physical field layer, responsible for calculating field effects such as electromagnetic engagement domains, acoustic engagement domains, and optical engagement domains; dedicated computing server resources are configured for each level. For example, space simulation tasks are assigned to a high-precision orbit calculation server cluster, and underwater simulation tasks are processed by an acoustic calculation server cluster.

[0203] In step S22, the server performance evaluation model is used to achieve intelligent agent load balancing. Monitor the load status of each computing node in real time and dynamically adjust the computing resource allocation according to the evolution of the combat situation. When the confrontation intensity in a certain combat domain increases, the computing resources in this domain can be dynamically increased.

[0204] In the fourth specific embodiment, after step S22, it further includes: communication optimization, designing an efficient inter-domain communication mechanism, and adopting a hierarchical message routing strategy, which specifically includes the following steps: First, establish an intra-warfare domain routing layer responsible for direct communication between nodes within the same warfare domain; second, establish an inter-warfare domain routing layer to realize message forwarding between different domains through boundary routing nodes; finally, establish a global routing layer to maintain the inter-domain routing table and optimize the routing path. Dynamically adjust the routing strategy according to the communication frequency, establish direct point-to-point communication links for frequently interacting nodes, and reduce the routing overhead. Adopt a distributed routing table synchronization mechanism to ensure the consistency of routing information. Through this three-layer routing architecture, the cross-domain communication delay is significantly reduced, the real-time performance of the system is improved, and the cross-warfare domain communication overhead is reduced. For frequently interacting nodes, establish direct communication links.

[0205] As an example, in step S3, implement the partition management of the warfare domain.

[0206] In step S31, divide the entire scene space into multiple computing regions, and each region is responsible for by an independent warfare domain.

[0207] In step S32, design a boundary processing mechanism between regions to ensure the continuity and consistency of the warfare domain at the region boundary.

[0208] In step S33, design a cooperation mechanism between warfare domains to handle the cross-region impact.

[0209] As an example, in step S4, design a distributed interaction mechanism to realize the interaction mechanism between the agent and the warfare domain under the distributed interaction mechanism.

[0210] In step S41, the agent in the scene emits a signal, and the information obtained by the agents in each physical field warfare domain is first preliminarily processed within the same warfare domain. The information includes target detection data, situation awareness results, threat assessment, etc.

[0211] In step S42, after the warfare domain related to the agent that emits the signal receives the signal, determine the affected area range.

[0212] In step S43, after determining the affected area range, within the determined affected range, the warfare domain calculates the propagation path and intensity attenuation of the signal, and at the same time combines the influence of obstacles such as terrain and buildings in the scene on the signal.

[0213] In step S44, after calculating the propagation path and intensity attenuation of a single signal, considering the influence of other agents or the environment existing in the scene, calculate the combined field strength of each other agent.

[0214] In the fourth specific embodiment, after step S44, it further includes: receiving data from each physical field engagement area, and generating a unified signal through a multi-source data fusion algorithm. The fusion process takes into account the timeliness, credibility, and accuracy of the data to form a coherent situation awareness.

[0215] In step S45, based on the fused signal, the engagement area evaluates the impact of the signal on the target agent, allocates tasks and resources to the agents in each physical field engagement area, and at the same time considers the performance of the target agent to obtain the final evaluation result. For example, after a space satellite agent discovers a target, it can direct an aerial agent to conduct tracking, and at the same time notify the ground air defense system to get ready.

[0216] In step S46, the engagement area sends the evaluation result to the relevant agents through a distributed event bus. After receiving these distributed events, the target agent triggers corresponding response actions. The evaluation results include: task completion rate (>95%), target detection probability (>90%), strike accuracy (<5m), collaborative response time (<1s), system survival rate (>85%), etc. The evaluation results show that the agents in each physical field engagement area demonstrate excellent task execution capabilities and collaborative combat effectiveness in complex confrontation environments, and all indicators meet or exceed the design requirements.

[0217] In step S47, to ensure the state consistency of each engagement area under the distributed interaction mechanism, a distributed state synchronization mechanism is adopted. Through regular state verification and incremental updates between each engagement area, the consistency of the engagement area values is maintained.

[0218] As an example, in step S5, the distributed interaction mechanism is optimized. In view of the characteristics of engagement area calculation under the distributed interaction mechanism, a multi-level optimized distributed interaction mechanism is implemented.

[0219] As an example, in step S6, the distributed effect is evaluated and feedback is provided to implement the evaluation and feedback mechanism under the distributed interaction mechanism. After all engagement areas in the scenario calculate the evaluation results, they first evaluate the impact on the agents locally, then notify the relevant agents of the evaluation results through the distributed event bus to trigger corresponding response actions, and finally, through the distributed state synchronization mechanism, integrate the evaluation results of all engagement areas in the scenario and achieve full-scenario-level response and state synchronization through the distributed event bus to ensure that the scenario states in all areas are consistent.

[0220] In the fourth specific embodiment, after step S6, it further includes: step S7: Design multi-engagement area visualization and interaction to achieve three-dimensional display and interactive control of a complex battlefield environment, including: steps S71 to S73.

[0221] Step S71: Render the multi-dimensional scene. Adopt the hierarchical rendering technology to achieve continuous scene display from space to the seabed. Support multi-view switching and free zooming from the global situation to local details.

[0222] Step S72: Display the physical field effects. Use specialized visualization algorithms to show the distribution and changes of various physical fields. For example, display the electromagnetic field strength in pseudo-color and show the sound wave propagation path by ray tracing.

[0223] Step S73: Design an interactive control interface. Provide a multi-level interactive control interface. The operator can conduct macro command on the global battlefield situation map or conduct micro control on specific agents.

[0224] As an example, the performance of the electromagnetic engagement domain in the multi-dimensional engagement domain interactive digital simulation engine of this embodiment is shown in Table 4.

[0225] Table 4 Performance table of the multi-dimensional engagement domain interactive digital simulation engine in the fourth specific embodiment In this embodiment, an integrated simulation of multi-domain coordinated operations of land, sea, air, and space is realized, supporting joint operation deduction in complex battlefield environments; a coupled calculation framework for the multi-physical field engagement domain is established to accurately simulate the interaction of physical effects in complex environments; an efficient distributed computing and data fusion mechanism is realized to ensure real-time performance in large-scale scenarios; an intuitive 3D visualization interface is provided to support multi-level situation display and interactive control; through a hierarchical calculation framework and optimization mechanism, a smooth simulation experience in complex scenarios is realized.

[0226] The method for constructing the multi-dimensional engagement domain interactive digital simulation engine provided by this application realizes the efficient coupled calculation of multi-physical fields such as electromagnetic, acoustic, and optical through the engagement domain concept and the sub-region management mechanism, supporting the interaction of multiple effects in complex scenarios. Adopt a distributed engine core architecture and an intelligent agent dynamic deployment mechanism to solve the coordination problem of tens of thousands of intelligent agents and the calculation of the complex physical field engagement domain, taking into account both calculation accuracy and real-time performance. Through the distributed interaction mechanism and state consistency maintenance, achieve millisecond-level human-computer interaction response, supporting real-time intervention by the operator and adjustment of simulation parameters. Support the flexible extension of new physical field models and interaction mechanisms through standardized interfaces. Utilize distributed computing optimization and a multi-level parallel computing framework to realize intelligent scheduling and dynamic load balancing of heterogeneous computing resources, improving resource utilization efficiency. Through the engagement domain partition management and boundary data exchange mechanism, realize the efficient management and coordination of multiple engagement domains in a large-scale space, ensuring the accuracy and continuity of the simulation.

[0227] The method for constructing a multi-dimensional engagement domain interactive digital simulation engine provided by this application can be applied to the fields of military exercises and drills: supporting joint military exercises of multiple departments to enhance combat command capabilities; war gaming and decision-making fields: assisting in the formulation of military strategies and tactical analysis to provide decision-making support; teaching and training fields: used for teaching practice in military academies to improve the actual combat capabilities of students; emergency management fields: simulating response plans for emergencies to improve processing efficiency; virtual reality fields: combining VR / AR technologies to provide immersive simulation experiences.

[0228] Although this application has been disclosed above by way of examples, it is not intended to limit this application. Any person with ordinary knowledge in the relevant technical field may make some modifications and refinements without departing from the spirit and scope of this application. Therefore, the protection scope of this application shall be subject to that defined by the appended patent application scope.

Claims

1. A method for constructing a multi-dimensional interactive digital simulation engine, characterized in that: include, Step S1: Build a distributed engine; Step S2: Design a distributed dynamic deployment mechanism for intelligent agents; Step S3: Design engagement domain partition management; Step S4: design a distributed interaction mechanism; Among them, the engagement domain is the medium for information exchange and status update between different intelligent agents.

2. The method for constructing a multi-dimensional interactive digital simulation engine according to claim 1, characterized in that: The step S1 includes: Step S11: define the standardized interface of the agent, including the basic attributes and behaviors of the agent; Step S12: Establishing the basic framework of the engagement domain and realizing the spatial and computational grid division of the engagement domain; Step S13: Design a distributed scheduling method to coordinate the running sequence of components on each server node.

3. The method for constructing a multi-dimensional interactive digital simulation engine according to claim 2, characterized in that: The step S2 includes: Step S21: Establishing a server performance evaluation model; Step S22: Using the server performance evaluation model to achieve agent load balancing.

4. The method for constructing a multi-dimensional combat domain interactive digital simulation engine according to claim 3, characterized in that: The step S3 includes: Step S31: In view of the characteristics of large-scale scenes, the entire scene space is divided into multiple calculation areas, each area is managed by an independent combat domain; Step S32: Designing a boundary processing mechanism between regions to ensure the continuity and consistency of the engagement domain at the region boundary; Step S33: Design a coordination mechanism between combat domains to handle cross-regional impacts.

5. The method for constructing a multi-dimensional interactive digital simulation engine according to claim 4, characterized in that: The step S4 includes: Step S41: Initialize the interaction event, the agent sends a signal, and the signal contains detailed information of the agent; Step S42: Update the state of the engagement domain. After receiving the signal, the engagement domain associated with the agent that sent the signal first determines the affected area range and calculates the computational grid that needs to be updated; Step S43: Calculate signal propagation, including: signal propagation path and intensity attenuation; Step S44: superimposing the combat domain effect, including: superimposing the influence of other intelligent agents or environments in the scene to obtain an updated distribution state of the combat domain; Step S45: Evaluate the impact of the target agent and obtain an evaluation result based on the updated distribution state of the engagement domain; Step S46: Feedback distributed events, send evaluation results to relevant agents, and trigger corresponding response actions; Step S47: state consistency maintenance, using a distributed state synchronization mechanism, and maintaining the consistency of the values ​​of the engagement domains through regular state verification and incremental updates between the engagement domains.

6. The method for constructing a multi-dimensional combat domain interactive digital simulation engine according to claim 5, characterized in that: After step S4, the method further includes: Step S5: Optimize the distributed interaction mechanism, and implement a multi-level optimized distributed interaction mechanism according to the characteristics of the engagement domain calculation under the distributed interaction mechanism; adopt adaptive grid technology, use finer grid division in key areas, and use coarser grids in distant areas to improve computing efficiency; decompose large-scale engagement domain numerical computing tasks into multiple subtasks, and assign them to different computing nodes in the engagement domain for parallel processing to achieve dynamic allocation of computing tasks; establish a distributed caching mechanism to cache static or slowly changing scene features, reduce cross-node data transmission and repeated calculations, and improve the efficiency of engagement domain numerical computing.

7. The method for constructing a multi-dimensional interactive digital simulation engine according to claim 6, characterized in that: After step S5, the method further includes: Step S6: Evaluate and feedback distributed effects, implement the evaluation and feedback mechanism under the distributed interaction mechanism, and after all engagement domains in the scene calculate the evaluation results, evaluate the impact on the intelligent agent locally; notify the relevant intelligent agents of the evaluation results and trigger corresponding response actions; integrate the evaluation results of all engagement domains in the scene through the distributed state synchronization mechanism, and realize the response and state synchronization of the entire scene through the distributed event bus to ensure that the scene state in all areas remains consistent.

8. The method for constructing a multi-dimensional interactive digital simulation engine according to claim 3, characterized in that: The operator is set to intervene and control the intelligent agent in the scene in real time. After step S13, it also includes: designing an interactive control method to process the operator's real-time instructions; after step S22, it also includes: the operator adds or removes the intelligent agent in real time through the control interface to dynamically adjust the scale of the interactive digital simulation of the multi-dimensional acoustic combat domain.

9. The method for constructing a multi-dimensional combat domain interactive digital simulation engine according to claim 2, characterized in that: After step S12, the method further includes: adopting a combat domain coupling mechanism to establish an interactive interface of the combat domain, wherein the interactive interface adopts an asynchronous communication mechanism based on a message queue, realizes data synchronization through timestamps and version numbers, and realizes bidirectional coupling of the combat domain; when the combat domain changes, the spatial relationship data is updated through atomic operations.

10. The method for constructing a multi-dimensional interactive digital simulation engine according to claim 6, characterized in that: After step S5, the method further includes: designing a parameter optimization mechanism, constructing a beam allocation optimization model, and realizing dynamic allocation of beam resources; establishing a data fusion parameter optimization mechanism, adopting an online learning method based on gradient descent, constructing an optimization model including key parameters such as data association threshold, filter gain, and feature weight, and realizing adaptive adjustment of parameters by minimizing fusion error; and adopting a long short-term memory network to realize the prediction of the performance of the intelligent agent.

11. The method for constructing a multi-dimensional interactive digital simulation engine according to claim 2, characterized in that: After step S12, it also includes: designing a field coupling calculator, establishing an interaction model between physical field engagement domains, taking the space engagement domain as the basic support, and providing a unified spatial reference and target state information for other physical field engagement domains; the field coupling calculator adopts a layered coupling architecture, including a physical effect layer, a data exchange layer and a control coordination layer, the physical effect layer realizes the interaction calculation between the physical field engagement domains; the data exchange layer is responsible for data transmission between different engagement domains, and adopts a combination of shared memory and message queues to ensure efficient exchange of large-scale data; the control coordination layer realizes timing synchronization and calculation coordination between engagement domains, and ensures the causal consistency of multi-physical field calculations through a token passing mechanism.

12. The method for constructing a multi-dimensional interactive digital simulation engine according to claim 3, characterized in that: After step S22, the method further includes: designing an inter-domain communication mechanism, adopting a hierarchical message routing strategy, establishing an intra-combat domain routing layer, and being responsible for direct communication between nodes in the combat domain; establishing an inter-combat domain routing layer, and realizing message forwarding between different domains through boundary routing nodes; establishing a global routing layer, maintaining an inter-domain routing table and optimizing routing paths; dynamically adjusting the routing strategy according to the communication frequency, and establishing direct point-to-point communication links for nodes that interact frequently, thereby reducing routing overhead.

Images