Snow simulation method and system based on multi-spectral environment modeling and physical behavior feedback

Abstract

The application discloses a snowfield simulation method and system based on multi-spectrum environment modeling and physical behavior feedback and belongs to the technical field of computer virtual environment simulation. The application aims at the defects of the absence of snowfield simulation physical modeling, single spectrum, and the disconnection between the environment and the tactical decision, generates a multi-spectrum unified snowfield material library by loading target area geographic data and snow condition initial parameters, constructs a globally snowfield environment state layer which can be dynamically updated, combines the snow physical parameters, the snow quality parameters and the attributes such as the weight of the simulation entity and the ground contact area to calculate the entity maneuvering influence parameters, completes the whole life cycle management of the snowfield interaction traces, synchronously renders the multi-spectrum simulation picture, realizes the environment decision closed loop feedback and the tactical evaluation. The application can realize high-fidelity snowfield simulation and is suitable for military training, equipment performance evaluation and the like.

Description

Technical Field

[0001] This invention belongs to the field of computer virtual environment simulation and environment modeling technology, specifically involving a snow simulation method and system based on multi-spectral environment modeling and physical behavior feedback. Background Technology

[0002] In military training, mission rehearsals, and equipment evaluation, high-fidelity virtual environment simulation is crucial. Among related technologies, the simulation of battlefield snow scenes typically employs a combination of texture mapping and particle systems. Texture mapping often uses white textures to simulate the visual effect of snow accumulation, while particle systems are primarily used to simulate snowfall. However, this simulation method mainly focuses on the visible light visual level, and in practical applications under complex battlefield environments, it still suffers from the following core technological shortcomings: First, the environmental simulation lacks physical depth and interactive feedback. Existing technologies often treat snowfields as static visual backgrounds, failing to establish a quantitative mechanical relationship between the physical parameters of the snow layer and the mobility of simulated entities. These snow layer physical parameters include at least snow depth and shear strength, while simulated entities include vehicles, personnel, and other battlefield maneuvers. For example, the movement resistance, slip characteristics, and extreme physical states such as "getting stuck" when vehicles travel on snow are difficult to simulate realistically and quantitatively. This directly leads to a severe disconnect between the simulated environment and the physical behavior of real snowfield scenarios, making it impossible to support high-precision maneuver behavior simulation.

[0003] Second, the environmental simulation spectrum is limited, failing to support multi-method reconnaissance simulations. Most existing simulation systems only support visible light band rendering, completely ignoring the snow environment in core modern battlefield observation modes such as infrared thermal imaging and low-light night vision. In snow environments, due to the high albedo and specific thermal radiation characteristics of snow, existing technologies cannot simulate key elements such as the "thermal imprint" left by moving heat sources on the snow or the scene's photometric characteristics in snowy night environments. This prevents trainees from conducting target identification and reconnaissance training in multi-spectral backgrounds within the simulation environment, failing to meet the simulation requirements of multiple observation modes.

[0004] Third, environmental features lack dynamic persistence and tactical relevance. In traditional simulation schemes, traces left by the movement of simulated entities, such as tire tracks and footprints, are often merely temporary visual effects, or even disappear instantly, lacking a tactically significant lifecycle management and logical persistence mechanism for these traces. This makes it impossible for the simulation environment to simulate high-level tactical reasoning behaviors based on "trace freshness," such as pursuit, retreat, and concealment. An effective closed-loop feedback loop cannot be formed between environmental state variables and tactical decisions, making it difficult to support high-level tactical simulation training.

[0005] Fourth, there is insufficient coupling between sensor simulation and environmental factors. Existing technologies lack global quantitative modeling of snow conditions, making it impossible to accurately simulate the signal attenuation effects of environmental factors such as snowfall intensity and snow thickness on commonly used battlefield sensors such as sonar and laser rangefinders. It is also impossible to achieve dynamic linkage between sensor performance and snow environment conditions, making it difficult to meet the core requirements of modern military simulation for all-round, multi-spectral, and physically driven battlefield environment simulation.

[0006] In summary, existing snow environment simulation technologies suffer from core technical defects such as lack of physical modeling, limited simulation spectrum, broken interactive closed loops, and disconnect between environmental and tactical decision-making, which fail to meet the core requirements of modern high-precision military simulation. Summary of the Invention

[0007] The purpose of this invention is to address the core shortcomings of existing snow environment simulation technologies, such as lack of physical modeling, single simulation spectrum, broken interaction loops, and disconnect between environment and tactical decision-making. It provides a snow environment simulation method and system based on multi-spectral environment modeling and physical behavior feedback. This addresses the technical problems of existing technologies, such as snow environment simulation only remaining at the visual level, failing to establish a quantitative mechanical correlation between snow layer physical parameters and simulated entity mobility, and failing to reproduce real snow physical interaction behavior. It also solves the technical problems of existing technologies supporting only visible light single-spectral rendering, failing to achieve unified modeling and synchronous simulation of multi-spectral snow environments, and failing to support reconnaissance and simulation training under multiple observation modes. Furthermore, it addresses the technical problems of existing technologies failing to achieve full lifecycle persistent management of snow interaction traces, failing to establish a quantitative feedback loop between environmental state and tactical decision-making, and being unable to support high-level tactical simulations. Simultaneously, it addresses the technical problems of insufficient coupling between sensor simulation and snow environment factors, and failing to achieve dynamic linkage simulation of sensor performance and snow state. Ultimately, it achieves comprehensive, multi-spectral, and physically behavior-driven high-fidelity snow environment simulation, meeting the core simulation requirements of high-precision military training, mission rehearsals, and equipment evaluation scenarios.

[0008] To achieve the above objectives, the first aspect of this invention provides a snow simulation method based on multi-spectral environment modeling and physical behavior feedback, comprising the following steps: S1 Snow Environment Initialization: Load the geographic data of the target simulation area, which includes at least elevation data and surface material classification map; set the initial snow condition parameters, which include at least snow depth, wind speed and sunshine conditions; generate a multi-spectral unified snow material library and construct a quantifiable and dynamically updated global snow environment state layer. S2 Physical Behavior Interaction and Environmental State Update: In response to the mobile interaction behavior of the simulated entity, the motion influence parameters corresponding to the simulated entity are calculated based on the snow physical parameters in the global snow environment state layer; at the same time, snow interaction traces corresponding to the simulated entity are generated, the snow interaction traces are managed persistently throughout their entire lifecycle, and the global snow environment state layer is updated synchronously based on the interaction behavior and trace update data. S3 Multi-band Synchronous Rendering: Based on the updated global snow environment state layer, multi-band synchronous rendering is performed to generate simulation images corresponding to the three spectral bands of visible light, infrared thermal imaging and low-light night vision, respectively. S4 Environmental Decision-Making Closed-Loop Feedback and Evaluation: Based on the quantitative parameters of the current global snow environment state layer, the sensor simulation performance and intelligent decision-making logic of the simulated entity are adjusted in conjunction with the simulation screen; and multi-spectral contrast analysis is performed in conjunction with the simulation screen to output tactical evaluation results for the simulated entity.

[0009] Furthermore, in step S1, the multi-spectral unified snow material library includes at least visible light material, infrared material, and terrain height offset parameters; the visible light material includes at least albedo, roughness, and normal map, and the infrared material includes at least thermal emissivity and basic thermal inertia parameters; the material data corresponding to all spectral bands share the same set of world coordinates, geometric model, and global time reference.

[0010] Furthermore, in step S2, based on the preset snow physics and mechanics model, combined with the snow physical parameters, snow quality parameters, and the physical property parameters of the simulated entity, the movement resistance coefficient and slip coefficient are calculated, and the maneuvering speed, steering effectiveness, fuel consumption rate, and stuck state determination results corresponding to the simulated entity are output as the maneuvering influence parameters; the snow quality parameters include at least snow surface hardness, the snow physical parameters include at least snow depth and shear strength, and the physical property parameters of the simulated entity include at least weight and ground contact area.

[0011] Furthermore, in step S2, the snow interaction traces include at least geometric traces and thermal traces; a global trace manager is established to persistently store the geometric shape, thermal signal intensity, visibility, and freshness of the snow interaction traces, and to perform attribute updates according to preset decay rules over time and environmental parameters until the cancellation conditions are met.

[0012] Furthermore, the infrared thermal imaging rendering in step S3 includes: calculating the radiance of the ground surface and the simulated entity based on a preset thermal radiation calculation model, generating thermal imaging images by combining dynamic range mapping, MTF modulation transfer function simulation and spatiotemporal noise simulation, and synchronously rendering the temporary temperature field shift corresponding to the thermal traces left by the moving heat source on the snow.

[0013] Furthermore, the low-light night vision rendering in step S3 includes: performing photoelectric conversion and phosphorescent screen shading simulation based on the scene HDR brightness map, and performing gain adjustment on the scene midtone brightness based on the high albedo characteristics of the snow surface, while simultaneously superimposing imaging defect simulations of halos, afterglow, and glow.

[0014] Furthermore, the intelligent decision-making logic of the simulated entity in step S4 includes: adjusting the decision weights or execution logic of the simulated entity in path planning, speed decision, formation selection, reconnaissance and pursuit, retreat and defense, and coordinated action based on the snow depth, snow surface hardness and trace freshness data in the global snow environment state layer.

[0015] Furthermore, snow depth weight, slope weight, and shadow area weight are introduced into the cost function of the path planning to achieve path optimization or real-time replanning of the simulated entity in a snow scene.

[0016] Furthermore, the method also includes: for large-scale cluster simulation entities, using a thermal proxy model to calculate the average thermal signal and random disturbance value of the area where the cluster is located, and using the thermal proxy model to render the cluster thermal signal.

[0017] Furthermore, the method also includes: adopting an asynchronous frame-by-frame update strategy for the global snow environment state layer, dividing different update frequencies according to the priority and relative distance of the simulation entities; and using parallel computing units to perform batch parallel calculations for the surface temperature conduction and environmental radiation calculation tasks.

[0018] Furthermore, in a distributed simulation scenario, the method also includes: using an environment state data bus based on a publish-subscribe model to synchronize regional incremental change events and regional update data of the global snow environment state layer across terminals, and updating local environment data on demand based on the regional incremental change events, so as to achieve environmental state consistency among multiple terminals.

[0019] A second aspect of this invention provides a snow simulation system based on multi-spectral environment modeling and physical behavior feedback, comprising: The snow environment initialization module is used to load the geographic data of the target simulation area and set the initial snow condition parameters, generate a unified multi-spectral snow material library, and construct a quantifiable and dynamically updated global snow environment state layer. The physical behavior interaction and environmental state update module is used to respond to the movement interaction behavior of the simulated entity, calculate the corresponding maneuvering influence parameters of the simulated entity based on the snow physical parameters in the global snow environment state layer, and manage snow interaction traces with a full life cycle to synchronously update the global snow environment state layer. The multi-band synchronous rendering module is used to synchronously render and generate simulation images corresponding to the three spectral bands of visible light, infrared thermal imaging, and low-light night vision based on the updated global snow environment state layer. The environmental decision-making closed-loop feedback and evaluation module is used to adjust the sensor simulation performance and intelligent decision-making logic of the simulated entity based on the quantitative parameters of the current global snow environment state layer, and to perform multi-spectral contrast analysis in conjunction with the simulation screen to output tactical evaluation results for the simulated entity.

[0020] Furthermore, it also includes a guidance and interaction module, which is used to realize the visualization adjustment of snow condition parameters, switching of multi-spectral views, and playback of simulation process and tactical data review and analysis.

[0021] Furthermore, it also includes an environment status data bus based on a publish-subscribe model, used to realize the publication and subscription response of incremental environment status events between modules, as well as the synchronization of incremental change events across terminals and regions in distributed scenarios and the on-demand update of local environment data.

[0022] This invention addresses the core shortcomings of existing snow environment simulation technologies, such as lack of physical modeling, single simulation spectrum, fragmented interactive closed loops, and disconnect between environmental and tactical decision-making. Through a comprehensive technical solution encompassing multi-spectral snow environment unified modeling, quantitative interaction of snow physical behavior, full lifecycle management of interaction traces, and closed-loop feedback between environmental and tactical decision-making, it achieves all-round, multi-spectral, and physically behavior-driven high-fidelity snow environment simulation. The core beneficial effects are as follows: This invention achieves deep physical interaction between the snow environment and the simulated entity's maneuvering behavior through quantitative physical and mechanical modeling of the snow environment, overcoming the limitation of existing technologies that can only perform visual representations of snow. Based on a pre-defined snow physical and mechanical model, this invention combines snow physical parameters, snow quality parameters, and the physical property parameters of the simulated entity to accurately calculate the corresponding drag coefficient and slip coefficient, and outputs the simulated entity's maneuvering speed, steering effectiveness, fuel consumption rate, and stuck-in-the-ground status determination results, establishing a quantifiable mechanical relationship between the physical properties of the snow layer and the maneuvering behavior of the simulated entity. The snow physical parameters include at least snow depth and shear strength, the snow quality parameters include at least snow surface hardness, and the physical property parameters of the simulated entity include at least weight and ground contact area. All parameters are completely matched with the physical characteristics of the real snow environment, effectively solving the technical problem of a serious disconnect between simulation results and real snow physical behavior in existing technologies. This significantly improves the physical fidelity of snow simulation and can directly support core simulation scenarios such as high-precision equipment maneuvering performance evaluation and personnel snow maneuvering training.

[0023] This invention achieves high-fidelity full-spectrum simulation of snow environments under multiple observation modes by constructing a unified multi-spectral snow material library and a multi-spectral synchronous rendering mechanism, filling the technical gap of existing technologies that can only simulate single-spectral bands of visible light. The unified multi-spectral snow material library constructed in this invention covers material properties for three core spectral bands: visible light, infrared thermal imaging, and low-light night vision. All material data for all spectral bands share the same set of world coordinates, geometric models, and global time reference, enabling the synchronous generation of spatially consistent and temporally synchronized multi-spectral simulation images. Simultaneously, this invention utilizes techniques such as thermal radiation physics calculations, synchronous rendering of thermal imprints from moving heat sources, and adaptation to the high albedo characteristics of snow surfaces to realistically reproduce the core performance of snow environments under mainstream observation modes in modern battlefields. This effectively solves the technical problem that existing technologies cannot support multi-spectral reconnaissance and simulation training, fully meeting the core requirements of modern military simulation for multi-observation modes and full-spectrum environment simulation.

[0024] This invention achieves a systematic closed-loop interaction between the snow environment and simulated entities by persistently managing snow interaction traces throughout their entire lifecycle, significantly enhancing the tactical depth of the simulation environment. The invention persistently stores the geometric and thermal traces generated by the movement of simulated entities. A global trace manager controls the geometric shape, thermal signal intensity, visibility, and freshness of the snow interaction traces throughout their entire lifecycle, updating attributes according to preset decay rules over simulation time and environmental parameters until the traces meet the cancellation conditions. This solution completely changes the limitations of existing technologies where snow is merely a static visual background and traces are only temporary visual effects. It transforms the snow environment into an interactive, storable, and traceable dynamic tactical element, realistically reproducing the systematic impact of snow traces on entity concealment, tracking, and reconnaissance behaviors, providing a complete environmental interaction foundation for high-level tactical simulation.

[0025] This invention solves the core problem of the disconnect between environmental state and tactical decision-making in existing technologies by constructing a quantitative feedback closed loop between environmental state and tactical decision-making, thus achieving seamless integration from environmental simulation to tactical evaluation. Based on quantifiable parameters such as snow depth, snow surface hardness, and trace freshness in the global snow environment state layer, this invention can directly link and adjust the decision weights and execution logic of simulated entities in path planning, speed decisions, formation selection, reconnaissance and pursuit, retreat and defense, and coordinated actions. Simultaneously, it combines multi-spectral simulation images to perform multi-spectral contrast analysis, outputting tactical evaluation results for simulated entities. This establishes a complete closed loop of environmental state changes, decision logic adjustments, tactical result output, and debriefing evaluation, enabling the snow environment to directly drive the dynamic optimization of tactical decisions. It supports high-level tactical reasoning training based on snow environment characteristics and can quantitatively analyze the impact of environmental factors on tactical mission results, achieving full-process closed-loop control and capability enhancement in simulation training.

[0026] This invention significantly improves the engineering feasibility and scenario adaptability of its technical solutions while ensuring high-fidelity simulation effects through multi-dimensional performance optimization and distributed scene adaptation schemes. By employing asynchronous frame-by-frame update strategies, batch processing by parallel computing units, and clustered hot proxy rendering, this invention drastically reduces the computational load of large-scale scene simulations. Simultaneously, through a publish-subscribe model-based environment state data bus, it synchronizes regional incremental change events and regional update data of the global snow environment state layer across terminals in distributed simulation scenarios. Based on regional incremental change events, it updates local environment data on demand, significantly reducing network bandwidth consumption in distributed simulations. This enables real-time snow simulation of large-scale battlefield scenarios and multiple cluster entities, while also meeting the complex scenario requirements of distributed joint training and multi-terminal collaborative simulation, demonstrating strong engineering practicality and scenario expansion capabilities. Attached Figure Description

[0027] Figure 1 This is a flowchart of the snow simulation method of the present invention; Figure 2 This is a system schematic diagram of the snow simulation method of the present invention. Detailed Implementation

[0028] Example 1

[0029] like Figure 1 As shown, this embodiment provides a snow simulation method based on multi-spectral environment modeling and physical behavior feedback, including the following steps: S1 Snow Environment Initialization: Load the geographic data of the target simulation area, which includes at least elevation data and surface material classification map; set the initial snow condition parameters, which include at least snow depth, wind speed and sunshine conditions; generate a multi-spectral unified snow material library and construct a quantifiable and dynamically updated global snow environment state layer. S2 Physical Behavior Interaction and Environmental State Update: In response to the mobile interaction behavior of the simulated entity, the motion influence parameters corresponding to the simulated entity are calculated based on the snow physical parameters in the global snow environment state layer; at the same time, snow interaction traces corresponding to the simulated entity are generated, the snow interaction traces are managed persistently throughout their entire lifecycle, and the global snow environment state layer is updated synchronously based on the interaction behavior and trace update data. S3 Multi-band Synchronous Rendering: Based on the updated global snow environment state layer, multi-band synchronous rendering is performed to generate simulation images corresponding to the three spectral bands of visible light, infrared thermal imaging and low-light night vision, respectively. S4 Environmental Decision-Making Closed-Loop Feedback and Evaluation: Based on the quantitative parameters of the current global snow environment state layer, the sensor simulation performance and intelligent decision-making logic of the simulated entity are adjusted in conjunction with the simulation screen; and multi-spectral contrast analysis is performed in conjunction with the simulation screen to output tactical evaluation results for the simulated entity.

[0030] As one implementation method, in step S1 of this embodiment, the multi-spectral unified snow material library includes at least visible light material, infrared material and terrain height offset parameters; the visible light material includes at least albedo, roughness and normal map, and the infrared material includes at least thermal emissivity and basic thermal inertia parameters; the material data corresponding to all spectral bands share the same set of world coordinates, geometric model and global time reference.

[0031] As one implementation method, in step S2 of this embodiment, based on a preset snow physics and mechanics model, combined with snow physical parameters, snow quality parameters, and the physical property parameters of the simulated entity, the movement resistance coefficient and slip coefficient are calculated, and the maneuvering speed, steering effectiveness, fuel consumption rate, and vehicle stuck state determination results corresponding to the simulated entity are output as the maneuvering influence parameters; the snow quality parameters include at least snow surface hardness, the snow physical parameters include at least snow depth and shear strength, and the physical property parameters of the simulated entity include at least weight and ground contact area.

[0032] As one implementation method, in step S2 of this embodiment, the snow interaction traces include at least geometric traces and thermal traces; a global trace manager is established to persistently store the geometric shape, thermal signal intensity, visibility and freshness of the snow interaction traces, and to perform attribute updates with time and environmental parameters according to a preset decay rule until the cancellation conditions are met.

[0033] As one implementation method, the infrared thermal imaging rendering in step S3 of this embodiment includes: calculating the radiance of the ground surface and the simulated entity based on a preset thermal radiation calculation model, generating a thermal imaging image by combining dynamic range mapping, MTF modulation transfer function simulation and spatiotemporal noise simulation, and synchronously rendering the temporary temperature field shift corresponding to the thermal traces left by the moving heat source on the snow.

[0034] As one implementation method, the low-light night vision rendering in step S3 of this embodiment includes: performing photoelectric conversion and phosphorescent screen coloring simulation based on the scene HDR brightness map, and performing gain adjustment on the mid-tone brightness of the scene based on the high albedo characteristics of the snow surface, while superimposing imaging defect simulations of halos, afterglow and halo.

[0035] As one implementation method, the intelligent decision-making logic of the simulation entity in step S4 of this embodiment includes: adjusting the decision weight or execution logic of the simulation entity in path planning, speed decision, formation selection, reconnaissance and pursuit, retreat and defense, and coordinated action based on the snow depth, snow surface hardness and trace freshness data in the global snow environment state layer.

[0036] As one implementation method, this embodiment introduces snow depth weight, slope weight, and shadow area weight into the cost function of the path planning to achieve path optimization or real-time replanning of the simulated entity in a snow scene.

[0037] As one implementation method, the method described in this embodiment further includes: for a large-scale cluster simulation entity, using a thermal proxy model to calculate the average thermal signal and random disturbance value of the area where the cluster is located, and using the thermal proxy model to complete the rendering of the cluster thermal signal.

[0038] As one implementation method, the method described in this embodiment further includes: adopting an asynchronous frame-by-frame update strategy for the global snow environment state layer, dividing different update frequencies according to the priority and relative distance of the simulation entities; and using parallel computing units to perform batch parallel calculations for the surface temperature conduction and environmental radiation calculation tasks.

[0039] As one implementation method, in a distributed simulation scenario, the method further includes: synchronizing regional incremental change events and regional update data of the global snow environment state layer across terminals based on a publish-subscribe environmental state data bus, and updating local environmental data on demand based on the regional incremental change events, so as to achieve environmental state consistency among multiple terminals.

[0040] like Figure 2 As shown, this embodiment is a system for implementing the snow simulation method described in Embodiment 1, and its overall process architecture is as follows. Figure 2 As shown. The overall execution flow of this system is divided into three core stages: simulation initialization, simulation main loop, and multi-spectral band comparative analysis and tactical evaluation. Each stage achieves data synchronization through a unified global snow environment state layer, and realizes command linkage and event response between modules through an environment state data bus based on a publish-subscribe model.

[0041] The simulation process begins at the simulation start node. After the task starts, it first enters the simulation initialization phase, completing the initialization of the battlefield environment and loading of geographic data and snow condition parameters. This phase is executed by the snow environment initialization module. This module first loads the geographic data of the target simulation area and the initial snow condition parameters corresponding to the task scenario. The geographic data includes at least the elevation data and surface material classification map of the target area, and the initial snow condition parameters include at least the global snow depth baseline value, ambient wind speed, and sunshine conditions. Subsequently, based on the loaded geographic data and snow condition parameters, a corresponding multi-spectral unified snow material library is generated for various surface materials, and a quantifiable and dynamically updated global snow environment state layer is constructed to provide basic environmental data support for the subsequent main simulation loop.

[0042] After initialization, the system enters the main simulation loop phase, which is the core operation phase of the system. Figure 2 As shown, it includes two core functional modules that run in parallel: a rendering module and a physics module. The two modules share the same set of global snow environment state layer data, ensuring the spatial consistency and temporal synchronization of multi-dimensional simulation data.

[0043] like Figure 2 As shown, the rendering module employs a serially executed rendering pipeline, sequentially comprising a visible light rendering unit, an infrared thermal imaging rendering unit, and a night vision low-light rendering unit. The visible light rendering unit receives terrain and material data from the global snow environment state layer and outputs a visible light rendering image of the snow-covered scene. The infrared thermal imaging rendering unit, based on real-time updated dynamic temperature field data from the global snow environment state layer, calculates the thermal radiation of the scene's surface and simulated entities, combines dynamic range mapping and noise simulation to generate a thermal imaging image, and simultaneously recreates the thermal imprints left by moving heat sources on the snow. The night vision low-light rendering unit, based on the scene HDR brightness map output by the visible light rendering pipeline, performs photoelectric conversion and night vision scene simulation, adjusts the scene brightness based on the high albedo characteristics of the snow surface, and outputs a low-light night vision image after superimposing imaging defect simulation. The multi-spectral simulation images output by the rendering module are synchronously transmitted to the multi-spectral contrast analysis unit, providing a data foundation for tactical evaluation.

[0044] like Figure 2As shown, the physics module employs a serially executed physics simulation pipeline, sequentially comprising a mobility impact calculation unit, a trace generation and persistence unit, and a sensor performance modulation unit. The mobility impact calculation unit receives movement control commands from the simulated entity, retrieves snow physical parameters and snow quality parameters at the corresponding location in the global snow environment state layer, and, combined with the simulated entity's weight, ground contact area, and other physical properties, calculates the simulated entity's drag coefficient and slip coefficient based on a preset snow physics and mechanics model. Simultaneously, it determines the vehicle's stuck state and outputs mobility impact parameters such as the simulated entity's speed, steering effectiveness, and fuel consumption rate, directly influencing the simulated entity's movement control logic. The trace generation and persistence unit simultaneously generates geometric traces, logical traces, and thermal traces based on the simulated entity's movement trajectory. The geometric traces are generated using deformable terrain. The system generates tire tracks and footprints that match the movement trajectory, and generates temporary temperature field offsets on the movement trajectory based on thermal traces. Logical traces create independent logical data objects for each trace instance. All traces are included in the global trace manager for full lifecycle persistence management, dynamically decaying and deregistering trace attributes according to preset rules. Simultaneously, the global snow environment state layer is updated synchronously based on the trace data. The sensor performance modulation unit adjusts the maximum effective distance and measurement accuracy of sensors such as sonar and laser rangefinders in the scene according to preset ratios based on snow thickness and snowfall intensity parameters in the global snow environment state layer, achieving dynamic linkage between sensor performance and snow environment state. All calculation results from the physics module are synchronously updated to the global snow environment state layer and fed back to the simulation main loop, realizing dynamic updating and closed-loop interaction of the environmental state.

[0045] The output data from the main simulation loop is synchronously fed into the multispectral contrast analysis and tactical evaluation stage, which is executed by the environmental decision-making closed-loop feedback and evaluation module. This module receives the multispectral simulation images output by the rendering module, the simulation entity state data output by the physics module, and the full-cycle environmental data of the global snow environment state layer. It performs multispectral contrast analysis and, combined with the target requirements of the mission, outputs the corresponding tactical evaluation results. Simultaneously, this module feeds back the environmental state parameters and tactical evaluation results to the intelligent decision-making unit of the simulation entity, adjusting the subsequent decision logic of the simulation entity, such as path planning, velocity decision, and tactical execution, to achieve a complete closed-loop feedback between environmental state and tactical decision-making.

[0046] In an optional embodiment of this example, the system is further provided with a guidance and control interaction module. The guidance and control interaction module provides a visual interactive interface for training guidance and control personnel, supports real-time adjustment of snow condition parameters, seamless switching of multi-spectral views, playback control of the simulation process, and review and analysis of tactical data, so as to realize the control and intervention of the entire simulation training process.

[0047] In an optional implementation of this embodiment, the system also includes an environment state data bus based on a publish-subscribe model. This environment state data bus serves as the central data hub of the system, receiving incremental environment state change events published by each module and pushing updated data to modules that subscribe to the corresponding events. It also supports cross-terminal data synchronization in distributed simulation scenarios, synchronizing only regional incremental change events and updating local environment data as needed, reducing network bandwidth usage in distributed simulation, and ensuring consistency of environment state among multiple terminals.

[0048] Example 2

[0049] This embodiment presents a snow simulation system that can be standardized, deployed, and modularly expanded. It can be directly integrated into various application scenarios such as military training simulation platforms, equipment performance testing systems, and digital twin battlefield environments. The system's functional modules adopt a decoupled design, allowing for flexible tailoring and expansion according to simulation needs, adapting to simulation requirements across all scenarios, from individual soldier tactical training to large-scale joint campaign exercises. The core functional modules and their specific implementations are as follows: This system is set up with four core functional modules. Each module achieves data communication through a unified global snow environment state layer and completes command linkage through standardized interfaces, forming a complete simulation closed loop from environment initialization, physical interaction, multi-spectral rendering to decision evaluation.

[0050] The snow environment initialization module is the basic input unit for the system's simulation. It performs four core operations: loading basic environmental data for the target simulation area, quantifying and configuring snow condition parameters, generating a multi-spectral material library, and constructing a global snow environment state layer. This provides a unified environmental data benchmark for the entire system's simulation operation. Specifically, this module first loads the geographic data of the target simulation area. This geographic data includes at least the digital elevation model data of the target area, a surface material classification raster map, and regional meteorological basic data. The surface material classification raster map at least labels five basic surface types: grassland, paved roads, woodland, bare rock, and water bodies, providing a foundation for subsequent differentiated snow cover modeling. Subsequently, based on the simulation task scenario, two modes are supported: globally unified configuration and regionally customized configuration. Initial snow condition parameters are set, which include at least the basic snow depth value, ambient wind speed, sunshine conditions, and ambient reference temperature. At the same time, differentiated snow quality parameters are configured for different surface types and different simulation areas. These snow quality parameters include at least the snow surface hardness, snow shear strength, pressure-subsidence characteristic parameters, and thermal inertia parameters. Corresponding parameter thresholds are preset for three basic snow quality types: new snow, old snow, and ice shell, providing a quantitative basis for subsequent physical behavior simulation. Based on the loaded geographic data and configured snow condition parameters, this module generates a corresponding multi-spectral unified snow material library for each combination of surface material and snow quality type. This material library includes at least visible light materials, infrared materials, and terrain height offset parameters. The visible light materials include at least albedo, roughness, and normal maps matching snow depth and snow quality, while the infrared materials include at least thermal emissivity and basic thermal inertia parameters matching snow quality and temperature. All spectral material data share the same set of world coordinates, geometric models, and global time references, ensuring spatial consistency and temporal synchronization in multi-spectral rendering. Finally, this module constructs a rasterized global snow environment state layer. This layer uses fixed-resolution raster textures as its carrier and stores six quantifiable and dynamically updated environmental parameters for each raster unit of the entire map: snow depth, snow surface hardness, shear strength, surface temperature, snow quality type, and trace status, providing a unified environmental data foundation for all modules in the system.

[0051] The Physical Behavior Interaction and Environmental State Update Module is the core unit of the system's physical simulation. It enables deep physical interaction between the snow environment and simulated entities, performing three core operations: quantitative simulation of the simulated entity's maneuvering behavior, full lifecycle management of snow interaction traces, and dynamic updating of the global environmental state. It serves as the central hub connecting environmental simulation and entity behavior. This module comprises two parallel subunits: a maneuvering behavior simulation subunit and a trace lifecycle management subunit. These two subunits share real-time data from the global snow environment state layer. The maneuvering behavior simulation subunit responds to the simulated entity's mobile interaction behavior, performing quantitative calculations and outputting the maneuvering impact parameters of the simulated entity. In practice, the system first retrieves the snow physical parameters and snow quality parameters of the corresponding grid cell in the global snow environment state layer based on the real-time position of the simulated entity. The snow physical parameters include at least snow depth and shear strength, and the snow quality parameters include at least snow surface hardness and pressure-sinking characteristics. Simultaneously, the system acquires the physical attribute parameters of the simulated entity. For personnel entities, these parameters include at least total soldier weight, load, and shoe sole contact area; for vehicle entities, these parameters include at least vehicle curb weight, track / tire contact area, wheelbase, and engine output power. Then, based on a preset snow physics and mechanics model, the system calculates the current movement resistance coefficient and slip coefficient of the simulated entity using the above parameters. Based on the calculation results, the system outputs the simulated entity's real-time maneuvering speed, steering angle effectiveness, and fuel / energy consumption rate. When the calculated movement resistance exceeds a preset vehicle-trapping threshold, a vehicle-trapping state determination is triggered. These parameters collectively serve as the maneuvering influence parameters of the simulated entity and are directly sent to the entity's motion control logic, achieving realistic physical feedback from the snow environment to the entity's maneuvering behavior. The trace lifecycle management subunit is used to synchronously generate corresponding snow interaction traces during the movement of the simulated entity, completing the persistent management of the traces throughout their entire lifecycle. Specifically, based on the simulated entity's movement trajectory, total weight, ground contact area, and current snow quality parameters, three types of snow interaction traces are generated simultaneously: geometric traces, thermal traces, and logical traces. Geometric traces utilize deformable terrain meshes or decal technology to generate tire tracks and footprint mesh depressions matching the movement trajectory in the area traversed by the entity. The depression depth is correlated with the entity's weight and snow hardness according to a preset relationship. Thermal traces correspond to temporary temperature field shifts on the entity's movement trajectory, forming thermal imprints left by the moving heat source. The initial temperature of the thermal imprint is correlated with the entity's heat source power and the ambient reference temperature according to a preset relationship. Logical traces create an independent logical data object for each trace instance, storing the trace's location information, generation time, geometric parameters, thermal signal intensity, visibility, freshness, and the entity information it belongs to.All logical data objects of all traces are incorporated into the global trace manager for unified persistent management. The global trace manager dynamically updates the thermal signal intensity, visibility, and geometry of each trace according to preset attenuation rules, combined with simulation time, ambient wind speed, sunlight conditions, and snow quality parameters. When the thermal signal intensity or visibility of a trace falls below a preset cancellation threshold, the trace is cancelled from the global trace manager, completing the full lifecycle management of the traces. Simultaneously, this unit sends the trace update data and changes in snow compaction state caused by entity interaction to the global snow environment state layer, completing the dynamic update of the environmental parameters of the corresponding raster unit.

[0052] The multi-spectral synchronous rendering module is the system's visualization output unit. Based on a real-time updated global snow environment state layer, it performs synchronous rendering of multi-spectral scenes, outputting simulated images in multiple observation modes consistent with the characteristics of a real snow environment. Specifically, this module works in real-time with the global snow environment state layer, sharing the same geometric channels and global time reference. It synchronously executes rendering processes for the visible light, infrared thermal imaging, and low-light night vision spectrums, ensuring spatial consistency and temporal synchronization of the three spectrum images. The visible light rendering uses a physically based rendering pipeline. Based on the snow depth and snow quality parameters in the global snow environment state layer, it calls the corresponding visible light materials from the unified multi-spectral snow material library to render the terrain, simulated entities, and scene elements modified by snow cover, restoring the realistic visual effects of snow-covered scenes under different snow depths and qualities. Infrared thermal imaging rendering is based on dynamic temperature field data and thermal emissivity corresponding to snow quality parameters in the global snow environment state layer. It calls a preset thermal radiation calculation model to calculate the radiance of the scene's surface and various simulated entities. Combining dynamic range mapping, MTF modulation transfer function simulation, and spatiotemporal noise simulation, it generates corresponding thermal images. Simultaneously, it renders the temporary temperature field offset corresponding to the thermal traces left by the movement of simulated entities, restoring the thermal imprints left by the moving heat source in the thermal image. Low-light night vision rendering is based on the scene HDR brightness map output from the visible light rendering pipeline. It completes photoelectric conversion and phosphorescent screen coloring simulation, performs gain adjustment on the scene's midtone brightness based on the high albedo characteristics of the snow surface, and simultaneously overlays imaging defect simulations such as halos, afterglow, and glow to generate low-light night vision images consistent with real snow-covered night vision scenes. The simulation images of the three spectral bands output by this module are simultaneously sent to the environmental decision-making closed-loop feedback and evaluation module, providing a data foundation for multi-spectral contrast analysis and tactical assessment.

[0053] The Environmental Decision-Making Closed-Loop Feedback and Evaluation Module is the core unit for decision-making and evaluation in the system. It achieves closed-loop feedback between the snow environment state and the decision-making behavior of simulated entities, completing three core operations: dynamic modulation of sensor performance, linkage adjustment of intelligent decision logic, and multi-spectral contrast analysis and tactical evaluation output. This enables seamless integration from environmental simulation to tactical evaluation. Specifically, based on the snow thickness, snowfall intensity, and snow quality parameters in the current global snow environment state layer, this module dynamically attenuates the maximum effective distance and measurement accuracy of various sensors such as sonar, laser rangefinders, and optical observation equipment within the scene according to a preset proportional function. This achieves real-time dynamic linkage between sensor simulation performance and the snow environment state, reproducing the impact of the real snow environment on battlefield reconnaissance equipment. Simultaneously, based on quantifiable parameters such as snow depth, snow surface hardness, and trace freshness in the global snow environment state layer, the intelligent decision-making logic of simulated entities within the scene is adjusted in a coordinated manner. Specifically, this includes adjusting the decision weights and execution logic in simulated entity path planning, speed decision, formation selection, reconnaissance and pursuit, retreat and defense, and coordinated actions. Among them, the cost function of path planning introduces snow depth weight, slope weight, shadow area weight, and trace freshness weight, enabling simulated entities to automatically select the sunward travel route with shallower snow depth and harder snow quality. At the same time, it can adjust the reconnaissance and pursuit strategy according to the freshness of enemy traces, realizing path optimization and real-time replanning in snow scenes. This module synchronously receives multi-spectral simulation images output by the multi-spectral synchronous rendering module. Combining the full-cycle environmental data of the global snow environment state layer, the maneuver data and decision-making behavior data of the simulated entities, it performs multi-spectral contrast analysis. The analysis focuses on the contrast difference between the target and the snow background under different spectral bands, and the identifiability of target traces under multiple spectral bands. Combined with the tactical objectives of the simulation task, it outputs quantitative tactical evaluation results on the tactical execution effect of the simulated entities and the completion of the training task. At the same time, it generates a quantitative analysis report on the impact of environmental factors on tactical results, completing the closed-loop management of the entire simulation process.

[0054] As an optional implementation of this embodiment, the system is also provided with a guidance and control interaction module, which provides a visual human-computer interaction interface for training guidance and control personnel, and realizes control and intervention over the entire simulation process. Specifically, it includes four core functions: First, a snow condition parameter visualization and adjustment function, providing a global parameter adjustment slider and a regional brush drawing tool, allowing operators to adjust ambient temperature, wind speed, snowfall intensity, snow depth and snow quality type in specific areas in real time, and quickly modify simulation scenarios; Second, a multi-spectral view switching function, allowing operators to seamlessly switch between visible light view, infrared thermal imaging view, low-light night vision view, and multi-spectral overlay view, while independently enabling or disabling snow depth layer, surface temperature layer, and trace distribution layer, achieving full-dimensional battlefield situational awareness; Third, a simulation process replay function, providing timeline control for the entire simulation cycle, allowing operators to pause, fast forward, rewind, and slow down the simulation process, restoring the battlefield environment and entity behavior at any time point; Fourth, a tactical data review and analysis function, which can draw movement efficiency curves and speed-snow depth correlation curves for any simulated entity, generate heat maps of firefight hotspots, dense footprint areas, and high-incidence areas of vehicle entrapment, quantitatively analyze the impact of environmental factors on tactical results, and support training review and capability assessment.

[0055] As an optional implementation of this embodiment, the system also includes an environment state data bus based on a publish-subscribe model. This environment state data bus serves as the central data hub of the entire system, employing a publish-subscribe message communication model to achieve decoupled data interaction between modules and cross-terminal synchronization in distributed scenarios. Specifically, each functional module within the system can act as a publisher, publishing incremental environment state change events with regional boundaries to the data bus, such as snow compaction status update events published by the physics module and snowfall intensity change events published by the weather module. Each functional module can also act as a subscriber, subscribing to environment change events required for its own business. When the data bus receives a corresponding event, it immediately pushes it to the module that subscribed to the event, triggering the module's event-driven response, eliminating the need for direct coupling communication between modules. In distributed simulation scenarios, the environment state data bus only synchronizes regional incremental environment state change events across terminals, updating local environmental data of each terminal as needed based on incremental events, without synchronizing full-scale environmental map data. This significantly reduces network bandwidth consumption in distributed simulation, ensures environmental state consistency among multiple terminals, and supports the stable operation of multi-node joint simulation training scenarios.

[0056] The snow simulation system described in this embodiment adopts a decoupled modular design, with clear boundaries and responsibilities for each functional module. It can be flexibly tailored and expanded according to simulation needs, adapting to various simulation scenarios from individual soldier tactical training to large-scale joint campaign simulations. At the same time, the system performs quantitative modeling of snow quality parameters and physical properties of simulated entities throughout the entire process, realizing deep linkage between the snow environment, entity behavior, and tactical decision-making. It completely solves the core defects of existing snow simulation technologies, such as superficiality, single spectrum, broken interactive closed loop, and disconnect between environment and tactics, and has extremely strong engineering feasibility and scenario adaptability.

[0057] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. It should be noted that any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A snow simulation method based on multi-spectral environment modeling and physical behavior feedback, characterized in that, Includes the following steps: S1 Snow Environment Initialization: Load the geographic data of the target simulation area, which includes at least elevation data and surface material classification map; set the initial snow condition parameters, which include at least snow depth, wind speed and sunshine conditions; generate a multi-spectral unified snow material library and construct a quantifiable and dynamically updated global snow environment state layer. S2 Physical Behavior Interaction and Environmental State Update: In response to the mobile interaction behavior of the simulated entity, the motion influence parameters corresponding to the simulated entity are calculated based on the snow physical parameters in the global snow environment state layer; at the same time, snow interaction traces corresponding to the simulated entity are generated, the snow interaction traces are managed persistently throughout their entire lifecycle, and the global snow environment state layer is updated synchronously based on the interaction behavior and trace update data. S3 Multi-band Synchronous Rendering: Based on the updated global snow environment state layer, multi-band synchronous rendering is performed to generate simulation images corresponding to the three spectral bands of visible light, infrared thermal imaging and low-light night vision, respectively. S4 Environmental Decision Closed-Loop Feedback and Evaluation: Based on the quantitative parameters of the current global snow environment state layer, the sensor simulation performance and intelligent decision-making logic of the simulation entity are adjusted in conjunction. The simulation image is then combined with multi-band contrast analysis to output tactical evaluation results for the simulated entity.

2. The method according to claim 1, characterized in that, In step S2, the snow interaction traces include at least geometric traces and thermal traces; a global trace manager is established to persistently store the geometric shape, thermal signal intensity, visibility and freshness of the snow interaction traces, and to perform attribute updates over time and according to preset decay rules, until the cancellation conditions are met.

3. The method according to claim 1, characterized in that, Infrared thermal imaging rendering in step S3 includes: calculating the radiance of the ground surface and the simulated entity based on a preset thermal radiation calculation model; generating thermal images by combining dynamic range mapping, MTF modulation transfer function simulation, and spatiotemporal noise simulation; and synchronously rendering the temporary temperature field shift corresponding to the thermal traces left by the moving heat source on the snow.

4. The method according to claim 1, characterized in that, The intelligent decision-making logic of the simulation entity in step S4 includes: adjusting the decision weights or execution logic of the simulation entity in path planning, speed decision, formation selection, reconnaissance and pursuit, retreat and defense, and coordinated action based on the snow depth, snow surface hardness and trace freshness data in the global snow environment state layer.

5. The method according to claim 4, characterized in that, Snow depth weight, slope weight, and shadow area weight are introduced into the cost function of the path planning to achieve path optimization or real-time replanning of the simulated entity in a snow scene.

6. The method according to claim 1, characterized in that, The method further includes: adopting an asynchronous frame update strategy for the global snow environment state layer, dividing different update frequencies according to the priority and relative distance of the simulation entities; and using parallel computing units to perform batch parallel calculations for the surface temperature conduction and environmental radiation calculation tasks.

7. The method according to claim 1, characterized in that, In a distributed simulation scenario, the method further includes: an environment state data bus based on a publish-subscribe model, synchronizing regional incremental change events and regional update data of the global snow environment state layer across terminals, and updating local environment data on demand based on the regional incremental change events, so as to achieve environmental state consistency among multiple terminals.

8. A snow simulation system based on multi-spectral environment modeling and physical behavior feedback, characterized in that, include: The snow environment initialization module is used to load the geographic data of the target simulation area and set the initial snow condition parameters, generate a unified multi-spectral snow material library, and construct a quantifiable and dynamically updated global snow environment state layer. The physical behavior interaction and environmental state update module is used to respond to the movement interaction behavior of the simulated entity, calculate the corresponding maneuvering influence parameters of the simulated entity based on the snow physical parameters in the global snow environment state layer, and manage snow interaction traces with a full life cycle to synchronously update the global snow environment state layer. The multi-band synchronous rendering module is used to synchronously render and generate simulation images corresponding to the three spectral bands of visible light, infrared thermal imaging, and low-light night vision based on the updated global snow environment state layer. The environmental decision-making closed-loop feedback and evaluation module is used to adjust the sensor simulation performance and intelligent decision-making logic of the simulated entity based on the quantitative parameters of the current global snow environment state layer, and to perform multi-spectral contrast analysis in conjunction with the simulation screen to output tactical evaluation results for the simulated entity.

9. The system according to claim 8, characterized in that, It also includes a guidance and interaction module, which is used to realize the visualization adjustment of snow condition parameters, the switching of multi-spectral views, and the playback of the simulation process and the analysis of tactical data review.

10. The system according to claim 8, characterized in that, It also includes an environment status data bus based on a publish-subscribe model, which is used to realize the publication and subscription response of incremental environment status events between modules, as well as the synchronization of incremental change events across terminals and regions and the on-demand update of local environment data in distributed scenarios.

Images