Intelligent Simulation

This paper proposes an agile model construction method based on meta-modeling theory to address issues such as low efficiency, poor cross-platform compatibility, and insufficient dynamic adaptability in military simulation model building in complex battlefield environments. The method utilizes a multi-domain integrated battlefield knowledge graph, a modular template library, and a dynamic verification mechanism to achieve rapid generation of heterogeneous simulation models, seamless integration across platforms, and enhanced system robustness. Case studies on typical scenarios, including counter-drone operations and joint air defense, validate the advantages of the method in cross-service coordination, environmental disturbance response, and tactical model adaptability. The study demonstrates that this method significantly improves the efficiency of building complex battlefield simulation systems and enhances real-time responsiveness, especially for modern joint operations and multi-platform collaborative simulations. By combining theory and practical applications, this paper provides an effective technical pathway and theoretical support for building adaptive, cross-platform compatible military simulation systems.

In recent years, the lattice Boltzmann method has achieved tremendous success as a novel numerical method in scientific computing. Compared with traditional computational fluid dynamics methods, the lattice Boltzmann method possesses unique advantages in parallel efficiency and program implementation. In this paper, the lattice Boltzmann method is employed to numerically simulate reactive flow in porous media. Specifically, the lattice Boltzmann method for solving the flow field is derived from incompressible macroscopic equations and formulated for porous media flow, while the lattice Boltzmann method for solving the species transport equation is based on a hybrid regularization approach and exhibits high numerical stability. In this study, the correctness of the program implementation is first verified by comparing the results with those obtained from other computational fluid dynamics methods reported in the literature and from commercial software. Subsequently, a porous media catalytic reactor model is simulated, and the computational results demonstrate that the lattice Boltzmann method is suitable for simulating such scenarios.

To address the problem of uniform experimental design in the constraint region, this paper proposes an improved ToPDE algorithm with higher performance based on the current more advanced two phase differential evolution (ToPDE) algorithm. Firstly, the proposed method optimizes the evolution of ToPDE algorithm, which improves the local optimization ability of the algorithm and accelerate convergence by introducing the parameter adaptive tuning strategy. In addition, the proposed method introduces the charge repulsion simulation evolution method in the second stage of the ToPDE algorithm, which enables the algorithm to have a finer global evolution capability. Experimental results show that the proposed method has better performance than the ToPDE algorithm.

With the continuous development of the times, hypersonic technology has become one of the focuses of current aviation research. The process of multi-body separation/load release is ubiquitous. In the process, the lumped uncertainties, including parameter uncertainty, unmodeled dynamics, and unknown disturbances, makes the design of the controller difficult. In order to maintain the stability of the lateral motion of the aircraft during the separation process, a backstepping control method based on command filter and neural network is proposed. A dual command filter to obtain the derivative of the differential signal between the virtual input variable and the virtual control law; Traditional neural network-based backstepping controller requires multiple neural network, and this kind of controller will lead to a more complex and fragile system. Our method requires only a single neural network to approximate matched uncertainties, effectively addressing the complexity and fragility issues inherent in traditional multi-neural-network-based backstepping controllers. Finally, the effectiveness of the designed separation control method is confirmed by simulation.

The Functional Mock-up Interface (FMI) has matured in co-simulation and data exchange for continuous-time systems; however, it still lacks sufficient support for event handling, especially for external events. Most existing master algorithms lack a unified modeling and processing mechanism for external events, which limits the application of the FMI standard in discrete event-driven systems. To address these issues, we first construct an external discrete event model based on the Discrete Event System Specification (DEVS) and propose a method for external event localization and handling within the FMI framework. On this basis, we extend the co-simulation master algorithm to support hybrid co-simulation involving external events. Experimental results demonstrate that the proposed method supports dynamic structural models and shows high accuracy and adaptability in external event simulation.

The accurate measurement and feedback of brake cylinder pressure is critical to the accuracy, reliability and availability of brake cylinder pressure control. To address the issues of unmeasurable brake cylinder pressure, as well as a faulty brake cylinder pressure sensor cannot provide accurate feedback value, a brake cylinder pressure observer was modeled. Firstly, the structure and principle of the brake system of railway vehicles were introduced and analyzed. Secondly, a mathematical model of the brake system was established, and the brake cylinder pressure observer based on LO (Luenberger Observer) was designed. Finally, case studies on the test bench were conducted, and the experimental results show that the proposed method enables precise estimation of the brake cylinder pressure via the EP (Electro-pneumatic) valve output chamber pressure and control signals, with a steady-state accuracy within ± 1 kPa. The research results provide a basis for closed-loop feedback control of brake cylinder pressure in the absence of sensing and sensor failure.

As the core of the equipment system, the command and control (C2) unit governs the operational processes of the equipment system. Accurately constructing command and control models has become a critical task in current equipment system simulation. To enhance the confidence level of model development, this study conducts an integrated analysis of the interaction between C2 units and other equipment, as well as the behavioral models of other equipment, from the perspective of model systems. This approach enables precise positioning of the C2 model at the system level and completes the functional and interface design of the C2 model. The modeling methods for functional modules of the C2 model, including network communication, situation awareness, command issuance, and control, are elaborated in detail. Application of the model in simulation tests demonstrates that the C2 model can simulate command and control processes and effectively support the requirements of system-level simulation.

Aiming at the problems of weak target search capability, poor coordination, and limitations in the field of view and attack capabilities of single-domain platforms in unknown suburban battlefield environments, this study proposes an air-ground coordinative dynamic target search and strike strategy based on IWPA-HDMPC and call mechanism. Models of air-ground agents and battlefield environments are proposed while constructing environmental map to enhance agents’ perception and search representation. Under the framework of distributed model predictive control (DMPC), objective functions are established, and an improved wolf pack algorithm with hierarchical mechanism is introduced to accelerate solution efficiency and improve search benefits. At the same time, the algorithm satisfies communication and safety constraints. Considering agent heterogeneity, a joint mission model integrating target search and coordinative strike is designed to provide agents with a closed-loop search-decision-action process. In simulation instantiated with real-world regions, compared with the latest method, the proposed method improves static area coverage efficiency by 25.9% and dynamic target discovery speed by 40%, demonstrating enhanced search efficiency and superior mission execution capability.

The land combat simulation environment primarily comprises urban and wilderness settings, where the path planning of military vehicles is mainly constrained and influenced by terrain and road. On the basis of designing and implementing a universal land combat environment service interface, in order to addresses the current scarcity of path planning algorithms in land battlefield environments and their poor real-time performance, an urban road path planning method for military vehicles was implemented based on OSRM(Open Source Routing Machine), and an incremental unit update algorithm for urban road networks was proposed to optimize the OSRM framework and enhance the efficiency of road networks update. Additionally, we implement a wilderness path planning method for military vehicles based on Theta*. Finally, the effectiveness and real-time performance of the proposed path planning methods in battlefield environments maneuvering simulation are validated through the construction of path planning simulation scenarios.

Advancements in space exploration technologies have garnered increasing global attention, with various nations initiating programs encompassing lunar missions, deep space exploration, and Mars expeditions. Nuclear thermal propulsion (NTP) rocket engines, serving as energy supply systems for space activities, offer significant advantages over traditional chemical fuel and solar electric propulsion by achieving high specific impulse and thrust with minimal nuclear fuel consumption. Among various NTP system cycles, the closed expander cycle demonstrates superior engine performance, making it particularly suitable for deep space missions. This study employs the Modelica language and the OpenModelica platform to design and simulate a closed expander cycle-based NTP system. Mathematical models for the propellant and key components of the propulsion system are established, followed by the construction of a simulation model. Comparative analyses of the simulation results are conducted to validate the accuracy of the model, and dynamic simulations of orbital maneuvers are performed. Addressing the current challenges in China's NTP field, such as its late inception and significant gaps in simulation technologies, this research provides a reference for the modeling and simulation design of space nuclear thermal propulsion systems by integrating system modeling of the propulsion module using Modelica.

Unmanned aerial vehicle (UAV) clusters offer enhanced resilience, coverage and mission flexibility but demand robust state estimation to counter disturbances and uncertainties. This paper proposes a distributed algorithm that fuses a Quaternion‑based Unscented Kalman Filter (QUKF) with a Generalized Proportional‑Integral Observer (GPIO) for improved state estimation. The decentralized design ensures scalability and fault tolerance across large formations, while the GPIO compensates for linearization errors, sharpening quaternion accuracy. Accurate state estimates are then integrated into cooperative control laws to maintain stable formation and ensure trajectory convergence. Simulations confirm that this approach achieves precise trajectory tracking and formation synchronization under sensor noise and model uncertainties.

Ensuring passenger safety in metro stations is paramount. This research investigates emergency evacuation events and evaluate station evacuation efficiency under various passenger behavioral decisions. We establish an improved social force model incorporating these behavioral decisions. By employing advanced AnyLogic simulation techniques, this research comprehensively deconstructs metro station emergency evacuation dynamics, revealing that passenger behavioral decisions, emotional contagion, and spatial configurations significantly influence evacuation efficiency. It includes various scenarios such as normal conditions, phototaxis behavior, following guidance and exit failures. Compared with the normal conditions, the growth rate of evacuation time under the influence of phototaxis, taking into account the panic psychology, has reached 19.71%. When different exits fail, the evacuation time growth rates of Exit 3 and Exit 4 are the most significant. Considering the panic psychology, their evacuation time growth rates are 53.93% and 28.59% respectively. The study's innovative social force model integrates psychological profiles and spatial factors, providing transportation planners with a nuanced framework for understanding crowd behavior during crisis scenarios. By quantitatively mapping individual psychological characteristics to evacuation performance, the research offers actionable insights into designing more effective emergency response strategies, ultimately enhancing passenger safety in high-density urban transit environments.

This paper proposes a high-fidelity hardware-in-the-loop (HIL) simulation method for radar seeker verification in complex electromagnetic environments, achieving equivalent reconstruction of electromagnetic environments between microwave anechoic chambers and real-world outfield scenarios through collaborative modeling of outfield environmental signal computation and indoor radio frequency (RF) simulation. The model integrates an out-field environment signal reception power high-fidelity calculation module with an in-field radio frequency simulation module. Innovatively employing target scenario signal simulation and triad equivalent simulation techniques, it generates target echo signal parameter sets through multi-physics joint simulation. Combined with array coarse/fine control units and a high-dynamic-range power allocation algorithm, this approach thus overcomes the trade-off between large dynamic range and high spatial resolution inherent in traditional RF anechoic chamber near-field synthesis techniques. Experimental results demonstrate that this method significantly improves power field reconstruction accuracy under rain/snow/fog conditions, enables equivalent simulation of multi-equipment scenarios through loading outfield test data, and confines the deviation between seeker-received signal power and outfield measurements within required specifications, providing effective HIL simulation support for anti-jamming verification of new guided weapon systems.

The phenomenon of water skipping, in addition to being commonly observed in daily life, is widely applied in simulations such as spacecraft recovery and vehicle-road interaction mechanics. This paper conducts a theoretical mechanical analysis of the water-skipping body’s motion process, deriving the contact forces generated when the body interacts with the water surface under different entry conditions. Subsequently, a numerical simulation analysis of this theory is performed based on the MWORKS platform. The results demonstrate that the theory can effectively describe the displacement and attitude changes of the water-skipping body during its motion.

Modeling of the maritime environment lays a foundation for path planning at sea. For this reason, a nautical chart binarization method based on a pixel matrix is put forward with a grid model and discretization. Red-green-blue images were obtained by adopting the imread function in Matlab software, while the rgb2gray function in Matlab was employed to convert these red-green-blue images into grayscale images, so as to gather the exclusive pixel information. By extracting the pixel information of navigable sea areas, grayscale images were directionally transformed into a binary nautical chart. Based on actual needs, grayscale images could be transformed into red or black to achieve the binarization of the nautical chart.

Obtaining crowdedness level and crowd movement trends on campus can help in understanding campus dynamics, and therefore, improve campus management efficiency and increase faculty and student satisfaction. Given the limitations of traditional data collection methods that heavily rely on fixed IoT (Internet of Things) devices or mobile sensors, this research turns to crowdsourcing as an alternative approach, capitalizing on the widespread use of mobile devices and web technology. This work introduces a mobile crowdsourcing approach for capturing campus dynamics within universities. We included three HCI (Human-Computer Interaction) modes for mobile crowdsourcing – passive mode, questionnaire mode, and dialogue mode. The passive mode collects the user's movement trajectory upon user’s approval. The questionnaire mode asks for campus dynamics in-formation such as the user's current activity location, activity time, and activity type through a pre-designed questionnaire. The dialogue mode collects the same information as the questionnaire mode, however, through dialogues between chatbots and users. We tested the feasibility of the proposed crowdsourcing approach on a university campus. We found that in the passive mode, the participants perceived the lowest cognitive workload. The questionnaire mode and the dialogue mode can not only produce valuable subjective perception and evaluation of campus dynamics, but also provide acceptable output quality of objective data compared to the passive mode. The dialogue mode outperforms the questionnaire mode in terms of output quality. This work provides valuable implications in designing crowdsourcing apps and improving university management.

Recently, 3D Gaussian Splatting has gained significant attention in the novel view synthesis domain due to its high-quality rendering and computational efficiency. However, when camera views (e.g., focal length, scene depth) vary, the rendered images exhibit aliasing artifacts, including erosion during zoom-in and dilation during zoom-out. The fundamental cause of this phenomenon lies in the fixed-scale 2D dilation filter, which can not adapt to dynamic changes in sampling frequency. To address this issue, a mechanism for adaptive frequency constraint on Gaussians has been proposed, which dynamically adjusts frequency characteristics of Gaussians to match varying sampling requirements. Additionally, based on spectral discrepancies in images, a frequency regularization term is designed to enhance consistency in frequency domain representation. These approaches effectively suppress aliasing artifacts while maintaining rendering efficiency.

Today, skill transfer from humans to agents via demonstrations is a widely adopted approach. However, prior research primarily focused on recording data generated during demonstrations, leading to challenges such as poor interpretability of the skill transfer process, limited transferability and generalization capabilities of demonstrations, and insufficient visualization. To address these issues, this paper proposes a method for agent skill learning and knowledge graph generation based on demonstrations. The contributions include the following. First, to enhance the interpretability of the skill transfer process, we construct demonstration skills. Second, to improve the transferability and generalization of demonstrations, we propose a DeepSeek-based method for decomposing demonstration skills and generating behavior trees. Finally, to strengthen visualization, we introduce a hierarchical skill generation method for agents using knowledge graphs. Experiments conducted in 3C assembly and satellite assembly scenarios demonstrate that our method substantially enhances the interpretability of the demonstration process.

3D point cloud segmentation is a key task in urban scene reconstruction, especially for extracting building structures, which are diverse in scale and geometry. Existing segmentation methods mainly rely on supervised deep learning, which suffers from limited generalization across different scenes and requires large amounts of annotated 3D data and computational resources. In contrast, 2D image segmentation has achieved significant progress. This work proposes a generalized 3D building segmentation framework based on 2D–3D fusion. By leveraging state-of-the-art 2D segmentation models such as Mask2Former and SAM, and combining them with 3D point clouds reconstructed by COLMAP, we establish correspondences between 2D masks and 3D points. This approach enables effective segmentation of 3D buildings without 3D supervision, and lays a foundation for downstream tasks such as urban scene reconstruction, measurement, and mapping.

The verification of the adaptability and robustness of infrared target recognition algorithms in complex and variable environments requires a large amount of target and interference test image data. Currently, various experimental verification methods generally suffer from insufficient scenario coverage and inadequate realism. To address this, this paper proposes a method for generating complex infrared interference scenarios based on few-shot data. The method first constructs a deep learning algorithm based on a semantic generation model to extract three-dimensional modeling features of targets and interference. Then, it combines knowledge transfer and deep diffusion generative models to achieve controllable simulation generation of typical infrared interference scenarios based on few-shot data. Finally, it uses image fusion algorithms to fuse infrared target and interference scenario images, achieving a close fit to the sample data. Experimental results have shown that the method in this paper has an effective ability to generate complex infrared interference scenarios, and the generated results have high credibility.

Multi-UAV (Unmanned Aerial Vehicle) cooperative task planning has demonstrated great potential in dealing with complex and dynamic environments. However, most existing multi-UAV task planning methods rely on static modeling under the assumption of homogeneous UAVs, along with manually designed heuristic algorithms or solvers. These methods struggle to adapt to real-time changes in task demands and environmental conditions. Heterogeneous UAV—with differences in capabilities, endurance—introduce further challenges in coordination and task allocation. To address these challenges, this paper proposes a dynamic heterogeneous multi-UAV task planning method based on reinforcement learning, aiming to overcome the limitations of traditional methods. Specifically, a sequential Markov decision process is formulated to model the dynamic heterogeneous multi-UAV task planning problem, and a dual-encoder lightweight single-decoder policy network is designed, incorporating self-attention and cross-attention mechanisms. Experiment results demonstrate that the proposed method outperforms widely-used solvers in both solution quality and computational speed, verifying its effectiveness and superiority.

The MARL (Multi-Agent Reinforcement Learning) training framework for CGF (Computer Generative Force) behavior decision-making provides RL interaction and iteration for engagement-level models in military field. To ensure compatibility with various RL algorithms and meet the requirements of high-fidelity military simulations, We present a RL training framework which supports training engagement-level models in military field, and enables faster-than-real-time iteration. We also propose the Graph Normalized MAPPO(GNMAPPO) to enhance the adaptability of agents with local observations by integrating relationship information into their learning processes and successfully apply it to an unmanned aerial vehicle (UAV) reconnaissance mission. The experimental results demonstrate that our framework supports simultaneous interaction among multiple heterogeneous entities and GNMAPPO significantly outperforms baseline algorithm in UAV reconnaissance mission.

In order to improve the problem of low efficiency of tactical rules relying on manual summarization in traditional behavior modeling, an intelligent modeling method of air combat behavior based on LLM is proposed. Under our existing air combat simulation framework, the reasoning ability of DeepSeek-R1-Distill-Qwen-7B model is used to output structured coding to support the reading and operation of behavior model. RAG retrieval enhancement generation and thinking chain are used to improve the output effect of the model. In addition, the rerank secondary re-ranking method is introduced in the similarity retrieval stage, and an indicator for evaluating the generation effect of structured html coding text is proposed. The results show that this method has good accuracy when processing a certain amount of situation information, improves the construction efficiency, and the simulation results have verified the effectiveness of the method.

To address the low efficiency and long debugging cycle in modeling air combat tactics, this paper proposes an intelligent modeling method for air force tactical confrontation behaviors based on YOLO and CLIP. By integrating object detection and cross-modal semantic understanding techniques, the method achieves high-precision detection of key elements in hand-drawn tactical sketches, such as trajectory and phase number elements. It further extracts semantic relationships among these elements to automatically generate structured models of tactical behaviors. Experimental results demonstrate that the proposed method enables end-to-end tactical modeling from hand-drawn sketches. In simulation verification, the execution trajectories of both offensive and defensive sides align well with the intended tactics in the sketches. This approach provides an innovative solution for efficient transformation of expert knowledge and tactical validation.

Neural operators have emerged as a powerful tool for solving partial differential equations (PDEs) and other complex scientific computing tasks. However, the performance of single operator block is often limited, thus often requiring composition of basic operator blocks to achieve better performance. The traditional way of composition is staking those blocks like feedforward neural networks, which may not be very economic considering parameter-efficiency tradeoff. In this paper, we propose a novel dual path architecture that significantly enhances the capabilities of basic neural operators. The basic operator block is organized in parallel two paths which are similar with ResNet and DenseNet. By introducing this parallel processing mechanism, our architecture shows a more powerful feature extraction and solution approximation ability compared with the original model. We demonstrate the effectiveness of our approach through extensive numerical experiments on a variety of PDE problems, including the Burgers’ equation, Darcy Flow Equation and the 2d Navier-Stokes equation. The experimental results indicate that on certain standard test cases, our model achieves a relative improvement of over 30% compared to the basic model. We also apply this structure on two standard neural operators (DeepONet and FNO) selected from different paradigms, which suggests that the proposed architecture has excellent versatility and offering a promising direction for neural operator structure design.

This paper explores the feasibility of using physical information neural network to solve the time-dimensional infinite plate temperature field problem, and compares the method with the traditional numerical method. Physical information neural networks integrate physical principles into learning algorithms and provide a new method to handle complex partial differential equation problems. In this study, the temperature field of a one-dimensional time-varying infinite plate with an internal and no internal heat source is studied respectively. The results show that the physical information neural network can achieve the same solution accuracy and speed by relying only on random sampling points. Nevertheless, the effectiveness of physical information neural networks in solving more complex heat conduction problems still needs further research.

To address the need for high-precision modeling in digital twin battlefields, this study proposes a six-degree-of-freedom digital twin model of a fixed-wing aircraft. The model adopts a modular and component-based architecture, enabling separate modeling of key components such as the fuselage and wings to enhance modeling granularity and flexibility. An improved Euler method is applied to increase the accuracy of dynamic simulation. Environmental response mechanisms—including standard atmosphere, icing, and rainfall models—are incorporated to dynamically adjust the aircraft’s characteristics under complex conditions. A genetic algorithm is employed to optimize the PID controller, enabling adaptive tuning with limited parameters. The proposed approach is innovative in its integration of fine-grained modeling with complex environmental responses, effectively addressing the limitations of traditional models in terms of low granularity and insufficient environmental representation. Simulation results demonstrate that the model exhibits strong stability and responsiveness, offering reliable support for digital twin modeling of battlefield equipment.

The wave of digital transformation is on the rise, and building a digital world corresponding to the physical world is likely to become the trend of future development and the foundation to support a great variety of intelligent applications. In the process of digital modeling, digital twin modeling language plays an important role. In this paper, we explain the related concepts, sort out the basic information and description form of commonly used digital twin modeling languages, such as asset administration shell and digital twin definition language, analyze the related key technologies, including the semantic enhancement of digital twin languages, and the construction and transformation of digital twin models. This paper can provide reference for the research in related fields.

As distributed simulation systems grow in scale and complexity, ensuring real-time data dissemination among heterogeneous subsystems becomes increasingly challenging. This paper proposes a data distribution optimisation framework for simulation agents, combining multi-stage pipeline parallelism and resource reservation for critical task handling. By parallelising data reception, parsing, processing, forwarding, and delivery, the framework reduces latency for time-sensitive tasks. Additionally, resource reservation ensures stable performance under high load by allocating dedicated CPU and memory to critical tasks. Experiments in multi-agent scenarios demonstrate significant reductions in end-to-end latency and improvements in real-time responsiveness. The framework is lightweight, platform-independent, and scalable, offering an effective solution for large-scale distributed simulation.

With the increasing complexity of simulation systems and the emergence of cloud computing, heterogeneous cloud-based simulation has become a new trend in the field of distributed simulation. However, the geographical distribution of simulation nodes pose significant challenges to data transmission performance. This paper proposes a lightweight, high-performance communication library tailored for heterogeneous simulation environments. The library incorporates adaptive protocol selection, high-performance data serialization with Protocol Buffers, and a hybrid transmission mechanism utilizing both HTTP/3 and TCP for reliability. Through a series of experiments simulating LAN and WAN conditions, the proposed library is shown to significantly outperform traditional TCP/XML-based solutions in terms of latency and stability. The results demonstrate the library’s capability to improve real-time performance and enhance the robustness of distributed simulations in complex network environments.

The definition for the helicopters general quality characteristics simulation model based on the requirements of helicopters general quality characteristics simulation is proposed in this paper. It suggests that the helicopters general quality characteristics simulation model needs VV&A (Verification, Validation, and Accreditation) based on the function and characteristics of the simulation model. According to the criteria of VV&A, a new VV&A scheme for the simulation model in combination with the characteristics and development process of the simulation model is proposed in this study. The VV&A activities should be through the whole process of it is emphasized. The foundation for further research on the implementation of the helicopters general quality characteristics simulation project is laid in this research.

On radio frequency (RF) simulation fidelity evaluation, The achievements of FISG are adopted to describe and define the fidelity of RF simulation, A fidelity evaluation index system in the time, space, frequency, energy, and polarization domains is established; a fidelity evaluation method based on error models and a feature consistency measurement method based on evidence frameworks are proposed. This approach enables the quantification of RF simulation fidelity, providing technical support for effectively utilizing simulation results in decision-making.

High-fidelity models can better meet the increasing requirements of contemporary flight dynamics simulation, making the assurance of their credibility a crucial issue. The credibility assessment of a simulation system must be conducted through Verification, Validation and Accreditation (VV&A). Main validation methods are systematically analyzed. The specific implementation process of a high-fidelity flight dynamics model is investigated from three dimensions: validation items determination, simulation experiment design, and results validation. Model refinement is guided by simulation results analyses involving basic performance parameters, aerodynamic coefficients, and dynamic performance parameters, which provides reference for the validation and optimization of other simulation systems. It also supports the comprehensive application of validation methods and the acquisition of output data from simulation models.

The multi-entity collaborative simulation system is an important experimental system for improving cluster collaborative verification capabilities. The multi-entity collaborative simulation system can integrate multiple entity models and various collaborative algorithms to perform a full-process simulation of cluster collaboration, supporting the effective verification of collaborative system architectures, collaborative task processes, key algorithms, and hardware and software systems under typical collaborative task scenarios. This article focuses on introducing the overall design, visual generation system, collaborative simulation and testing system, and network interface design of the multi-entity collaborative simulation system. The system has been successfully applied in multi-entity collaborative simulation experiments and has passed various collaborative simulation validation assessments.

In current research, there is an urgent need to quickly reconstruct 3D scenes from detected scene data for scenario design, rapid verification, and other applications across various fields. This paper proposes a method for rapid 3D scene reconstruction based on the U-Net network model. By dividing real-world scenes with complex and multi-scale elements into terrain elements, ground elements, and surface elements based on their size, and using image recognition and segmentation techniques to extract these multi-scale elements, the method leverages the UE4 engine to match and reconstruct the data. This approach enables the rapid establishment of 3D scenes.