Non-real-time distributed joint simulation system supporting model dynamic connection
By employing technologies such as model integration and management modules, dynamic connection and scheduling engines, the dynamic connection and distributed collaborative operation of heterogeneous simulation models are realized, solving the problem of static model connection in existing technologies and improving the flexibility and efficiency of the simulation system.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- CHENGDU ANSHI ASIA PACIFIC TECH CO LTD
- Filing Date
- 2026-01-08
- Publication Date
- 2026-04-17
AI Technical Summary
Existing distributed simulation systems have static model connections during the design iteration process, lacking dynamic reconstruction capabilities, which limits the breadth and depth of design exploration.
By employing a model integration and management module, a dynamic connection and scheduling engine, a distributed simulation execution module, a data interaction and collaboration bus, and a process and task management platform, the system achieves standardized encapsulation, dynamic connection, and distributed collaborative operation of heterogeneous simulation models.
It supports dynamic model connections, improves the efficiency of simulation resource utilization and design agility, ensures data consistency and result reliability, and adapts to topology reorganization and dynamic allocation of computational tasks during the design iteration process.
Smart Images
Figure CN121881663A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of distributed simulation technology, specifically to a non-real-time distributed co-simulation system that supports dynamic model connection. Background Technology
[0002] As equipment systems become increasingly complex, their design process often involves deep collaboration across multiple disciplines, including mechanical, electrical, control, and software. To address this challenge, co-simulation systems are gradually evolving from centralized, tightly coupled architectures to distributed, loosely coupled architectures. Existing technologies typically employ frameworks based on HLA (High-Level Architecture) or DIS (Distributed Interactive Simulation) standards, aiming to achieve interconnectivity between heterogeneous simulation nodes. These systems provide basic communication, synchronization, and data distribution services through runtime infrastructure (RTI), enabling basic integration of simulation applications in a distributed environment.
[0003] At the model integration level, existing technologies primarily encapsulate different types of simulation models by defining standardized interfaces (such as FMI / FMU standards), which to some extent solves the problem of model format heterogeneity. Regarding distributed computing, existing systems can deploy simulation tasks across multiple computing nodes and provide preliminary time management mechanisms. However, in terms of data interaction, they generally rely on specific middleware or communication protocols to achieve data exchange between nodes.
[0004] Existing technologies have the following shortcomings: Most existing systems employ predefined static connection configurations. Data flows and calling relationships between simulation components must be fully determined and compiled into the executable file before simulation begins. If adjustments to the model topology or data dependencies are needed during design iterations, reconfiguration, compilation, and even restarting of the entire simulation system are often required. This one-off connection method cannot support frequent scheme modifications and optimizations in agile design processes, severely limiting the breadth and depth of design exploration. Summary of the Invention
[0005] To address the shortcomings of existing technologies, this invention provides a non-real-time distributed co-simulation system that supports dynamic model connections, solving the problem of static model connections and lack of dynamic reconstruction capabilities.
[0006] To achieve the above objectives, the present invention provides the following technical solution: a non-real-time distributed co-simulation system that supports dynamic model connection, including a model integration and management module for standardized encapsulation and unified management of heterogeneous simulation models;
[0007] The dynamic connection and scheduling engine communicates with the model integration and management module to dynamically construct logical connection relationships between simulation components and generate simulation workflows based on user configuration, as well as to schedule distributed computing resources.
[0008] The distributed simulation execution module is communicatively connected to the dynamic connection and scheduling engine, and is used to load and run simulation components assigned by the scheduling engine on multiple computing nodes.
[0009] The data interaction and collaboration bus connects the model integration and management module, the dynamic connection and scheduling engine, and the distributed simulation execution module, and is used to provide unified data communication services for each simulation component;
[0010] The process and task management platform communicates and connects with the aforementioned modules to provide configuration, monitoring, data management, and user interface for the simulation process.
[0011] In some embodiments, the model integration and management module includes:
[0012] The model abstraction and encapsulation unit is used to establish a unified interface for input, output, parameters and state variables for simulation models of different formats, and to transform heterogeneous models into standardized simulation components.
[0013] The model repository and version control unit are used for centralized storage, classification, and version management of the encapsulated simulation components.
[0014] In some embodiments, the dynamic connection and scheduling engine includes:
[0015] The connection configuration tool provides a graphical interface for defining data flow and execution logic between simulation components via drag-and-drop.
[0016] A simulation workflow orchestrator, used to compile user-defined connections into executable simulation workflows;
[0017] And a resource management and task scheduler, used to schedule each simulation component to the corresponding computing resources for execution according to the simulation workflow.
[0018] In some embodiments, the distributed simulation execution module includes:
[0019] Simulation node agents deployed on each computing node are used to receive scheduling instructions, load local simulation components, and perform simulation calculations.
[0020] And a simulation clock synchronizer, used to provide a logical time synchronization mechanism for all distributed simulation components in a non-real-time simulation environment.
[0021] In some embodiments, the data interaction and coordination bus includes:
[0022] A unified data bus, built on a high-performance message middleware, is used to transmit input and output data between various simulation components;
[0023] Data format standardization and converters are used to define standard formats for data exchange within the system and provide format conversion services for non-standard data.
[0024] In some embodiments, the process and task management platform includes:
[0025] The simulation workflow monitoring panel is used to visualize the execution status, component operation, and resource load of the simulation workflow in real time.
[0026] The data management and analysis unit is used to record, store, query, and analyze data throughout the simulation process, and generate analysis reports; and the user and permission management center is used to manage system users, roles, and their operating permissions.
[0027] In some embodiments, the connection configurator is further configured to provide users with automatic connection recommendations based on data dependencies of the simulation component interface.
[0028] In some embodiments, the resource management and task scheduler operates by dynamically decomposing tasks and allocating resources based on the logical structure of the simulation workflow and the real-time load status of each computing node.
[0029] Compared with existing technologies, this invention provides a non-real-time distributed co-simulation system that supports dynamic model connection, and has the following beneficial effects:
[0030] A non-real-time distributed co-simulation system supporting dynamic model linking offers significant advantages over existing technologies. The system completely breaks down the barriers of model heterogeneity, achieving "plug-and-play" and efficient reuse of cross-disciplinary simulation models through unified standardized encapsulation and a model repository. By introducing a graphical dynamic linking and intelligent scheduling engine, the system solves the problems of static connection relationships and rigid resource scheduling in traditional solutions. This allows for flexible and efficient reorganization of the simulation topology during design iterations and dynamic allocation of computational tasks based on real-time load, greatly improving the utilization efficiency of simulation resources and the agility of collaborative design. Simultaneously, the system's high-performance data bus and advanced simulation clock synchronization mechanism effectively overcome the shortcomings of low data interaction efficiency and timing errors in non-real-time simulation scenarios, ensuring data consistency and result reliability in the distributed simulation process. With an integrated process and task management platform, the system achieves refined monitoring and data-driven decision-making throughout the entire lifecycle of complex simulation tasks. Attached Figure Description
[0031] Figure 1 This is a schematic diagram of the overall process of the non-real-time distributed co-simulation system of the present invention; Detailed Implementation
[0032] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0033] In the description of this invention, it should be noted that the terms "upper," "lower," "inner," "outer," "front end," "rear end," "both ends," "one end," and "the other end," etc., indicate the orientation or positional relationship based on the orientation or positional relationship shown in the accompanying drawings. They are used only for the convenience of describing this invention and for simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this invention. Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance.
[0034] In the description of this invention, it should be noted that, unless otherwise explicitly specified and limited, the terms "installed," "equipped with," "connected," etc., should be interpreted broadly. For example, "connection" can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium; it can be a connection within two components. Those skilled in the art can understand the specific meaning of the above terms in this invention based on the specific circumstances.
[0035] Please see Figure 1 This implementation plan relates to a non-real-time distributed co-simulation system that supports dynamic model connection. The system achieves standardized integration, dynamic connection and distributed collaborative operation of heterogeneous simulation models through a modular architecture.
[0036] Model Integration and Management Module
[0037] Functional positioning: This unit is the system's "model standardization factory", responsible for converting various heterogeneous models into callable components in a unified format.
[0038] Specific components:
[0039] Model abstraction and encapsulation unit:
[0040] A standardized model interface specification is defined using an Interface Description Language (IDL), which includes input ports, output ports, parameter configuration areas, and state variable areas.
[0041] Unified encapsulation of simulation models in different formats such as FMU (Functional Mock Unit), Simulink models, and Modelica models.
[0042] Establish a model metadata standard, including core information such as model type, version number, creation time, and interface definition.
[0043] Model repository and version control system:
[0044] The model repository is built using a distributed file system, which supports distributed storage and highly available access to model files.
[0045] Implement a Git-based version control mechanism to record model change history and support version rollback and branch management.
[0046] Establish a model classification system, classifying models in multiple dimensions according to professional fields (mechanical, electrical, control, etc.) and functional types (dynamics, thermodynamics, fluid, etc.).
[0047] Technical details:
[0048] The model encapsulation adopts containerization technology, packaging each simulation model and its dependent environment into an independent container image.
[0049] The model interface uses the Protobuf protocol for serialization to ensure data format consistency.
[0050] Version control uses semantic version numbers (e.g., v1.2.3) and supports model dependency management.
[0051] Dynamic connection and scheduling engine
[0052] Functional positioning: This unit is the "nerve center" of the system, responsible for the dynamic configuration and resource scheduling of simulation components.
[0053] Core components:
[0054] Connection relationship configurator:
[0055] A web-based graphical configuration interface that supports drag-and-drop component connections.
[0056] It provides an automatic connection recommendation function, which intelligently matches based on interface data type and semantic similarity.
[0057] Supports visual configuration of complex logical relationships such as conditional branches and loops.
[0058] Simulation workflow orchestrator:
[0059] Transform graphical configuration into an execution workflow based on a directed acyclic graph (DAG).
[0060] Achieve persistent storage and templated management of workflows.
[0061] Supports version control and reuse of workflows
[0062] Resource Management and Task Scheduler:
[0063] Real-time monitoring of CPU, memory, network, and other resource status of each computing node.
[0064] Task scheduling optimization based on genetic algorithm to minimize overall simulation time.
[0065] Supports dynamic load balancing, adjusting task allocation based on runtime status.
[0066] Implementation details:
[0067] The connection configuration is stored in JSON format, containing complete information such as component ID, connection relationship, and data mapping.
[0068] The workflow execution engine supports fault tolerance mechanisms such as breakpoint resumption and error recovery.
[0069] The scheduler adopts a microservice architecture, supporting horizontal scaling and high-concurrency scheduling.
[0070] Distributed simulation execution module
[0071] Functional positioning: This unit is the "execution terminal" of the system, responsible for loading and running simulation components on each computing node.
[0072] Key components:
[0073] Simulation node agent:
[0074] Lightweight service program deployed on each compute node
[0075] Receive scheduling instructions and pull the specified version of the model container image.
[0076] Manage the lifecycle of the local simulation process and monitor its running status.
[0077] Simulated clock synchronizer:
[0078] Employing a Parallel Discrete Event Simulation (PDES) Mechanism Based on Optimistic Synchronization
[0079] Implement event timestamp ordering and causal consistency guarantees between logical processes (LPs).
[0080] Supports adaptive time advancement, dynamically adjusting the time step based on component computation latency.
[0081] Operating mechanism:
[0082] When a node agent starts up, it registers its own capabilities (supported model types, computing resources, etc.) with the scheduler.
[0083] The clock synchronizer ensures consistent time progression in distributed simulations through a conservative barrier mechanism.
[0084] It supports the dynamic addition and removal of compute nodes without affecting the overall simulation operation.
[0085] Data Interaction and Collaboration Bus
[0086] Functional Architecture: This unit is the system's "information superhighway," undertaking the data communication tasks between all modules.
[0087] Core elements:
[0088] Unified data bus:
[0089] Build a publish-subscribe communication model based on DDS (Data Distribution Service).
[0090] Implement Quality of Service (QoS) policies, including reliability, persistence, and deadlines.
[0091] Supports adaptive selection of multiple transport protocols (TCP, UDP, shared memory, etc.)
[0092] Data format standardization and converter:
[0093] Define the system's internal standard data format (SIDL), which includes metadata such as data type, unit, and timestamp.
[0094] Provides a plug-in data transformation framework that supports custom transformation logic.
[0095] Achieve zero-copy optimization for data serialization and improve transmission efficiency.
[0096] Process and Task Management Platform
[0097] System Functions: This unit is the "command and control center" of the system, providing simulation management capabilities throughout the entire lifecycle.
[0098] Functional modules:
[0099] Simulation process monitoring panel:
[0100] Real-time display of workflow execution progress and component running status
[0101] Visualize resource usage and support comparison of real-time and historical data.
[0102] Provides abnormal alarm and fault location functions
[0103] Data Management and Analysis Unit:
[0104] Simulation data is stored using a time-series database, supporting efficient querying and analysis.
[0105] Built-in data mining algorithms automatically identify patterns and anomalies during the simulation process.
[0106] Supports custom report templates and allows for one-click generation of simulation analysis reports.
[0107] User and Permission Management Center:
[0108] Implement a role-based access control (RBAC) mechanism
[0109] Supports operation auditing, recording all users' key operations.
[0110] Provides project-level access control to ensure data security across different projects. Specific Implementation
[0111] Taking the joint simulation of the attitude control system of a certain type of satellite as an example:
[0112] System deployment environment:
[0113] A distributed computing cluster is built using 5 high-performance servers.
[0114] Each server is configured with: 32-core CPU / 128GB RAM / 10Gbps network.
[0115] The operating system used is CentOS 7.6, and the container runtime is Docker 20.10.
[0116] Model integration phase:
[0117] The onboard computer control model (Simulink format) is standardized through model abstraction and encapsulation units, defining 6 input ports (sensor data) and 4 output ports (actuator commands).
[0118] The attitude dynamics model (Modelica format) was packaged and registered in the model repository, version v2.1.0.
[0119] The environmental perturbation model (custom Python) is packaged into a container image and stored in the model repository.
[0120] Simulation configuration phase:
[0121] Users can drag and drop three model components in the connection configuration tool to establish connections.
[0122] The configurator automatically recommends the optimal connection scheme based on the interface data type (such as angular velocity, torque, etc.).
[0123] The workflow orchestrator generates a simulated workflow containing 10 execution steps.
[0124] The task scheduler decomposes the task into: Node 1 operation control model, Node 2 operation dynamics model, and Node 3 operation disturbance model.
[0125] Simulation execution process:
[0126] Each node agent receives scheduling instructions and pulls the corresponding version of the container image from the model repository.
[0127] The simulation clock synchronizer initializes the logic time, and the simulation step size is set to 10 milliseconds.
[0128] The control model calculates actuator instructions and sends them to the dynamic model via the data bus.
[0129] The dynamic model calculates the satellite's motion state, and the attitude data is updated by combining the output of the perturbation model.
[0130] The monitoring panel displays real-time curves showing changes in key parameters such as satellite pitch and roll angles.
[0131] Finally, it should be noted that the above descriptions are merely preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.
Claims
1. A non-real-time distributed co-simulation system supporting dynamic connection of models, characterized in that, include: The model integration and management module is used for standardized encapsulation and unified management of heterogeneous simulation models; The dynamic connection and scheduling engine communicates with the model integration and management module to dynamically construct logical connection relationships between simulation components and generate simulation workflows based on user configuration, as well as to schedule distributed computing resources. The distributed simulation execution module is communicatively connected to the dynamic connection and scheduling engine, and is used to load and run simulation components assigned by the scheduling engine on multiple computing nodes. The data interaction and collaboration bus connects the model integration and management module, the dynamic connection and scheduling engine, and the distributed simulation execution module, and is used to provide unified data communication services for each simulation component; The process and task management platform communicates and connects with the aforementioned modules to provide configuration, monitoring, data management, and user interface for the simulation process.
2. The non-real-time distributed co-simulation system supporting dynamic model connection according to claim 1, characterized in that: The model integration and management module includes: The model abstraction and encapsulation unit is used to establish a unified interface for input, output, parameters and state variables for simulation models of different formats, and to transform heterogeneous models into standardized simulation components. The model repository and version control unit are used for centralized storage, classification, and version management of the encapsulated simulation components.
3. A non-real-time distributed co-simulation system supporting dynamic model connection according to claim 1, characterized in that: The dynamic connection and scheduling engine includes: The connection configuration tool provides a graphical interface for defining data flow and execution logic between simulation components via drag-and-drop. A simulation workflow orchestrator, used to compile user-defined connections into executable simulation workflows; And a resource management and task scheduler, used to schedule each simulation component to the corresponding computing resources for execution according to the simulation workflow.
4. A non-real-time distributed co-simulation system supporting dynamic model connection according to claim 1, characterized in that: The distributed simulation execution module includes: Simulation node agents deployed on each computing node are used to receive scheduling instructions, load local simulation components, and perform simulation calculations. And a simulation clock synchronizer, used to provide a logical time synchronization mechanism for all distributed simulation components in a non-real-time simulation environment.
5. A non-real-time distributed co-simulation system supporting dynamic model connection according to claim 1, characterized in that: The data interaction and collaboration bus includes: A unified data bus, built on a high-performance message middleware, is used to transmit input and output data between various simulation components; Data format standardization and converters are used to define standard formats for data exchange within the system and provide format conversion services for non-standard data.
6. A non-real-time distributed co-simulation system supporting dynamic model connection according to claim 1, characterized in that: The process and task management platform includes: The simulation workflow monitoring panel is used to visualize the execution status, component operation, and resource load of the simulation workflow in real time. The data management and analysis unit is used to record, store, query, and analyze data throughout the simulation process and generate analysis reports; and the user and permission management center is used to manage system users, roles, and their operating permissions.
7. A non-real-time distributed co-simulation system supporting dynamic model connection according to claim 3, characterized in that: The connection relationship configurator is further configured to provide users with automatic connection recommendations based on the data dependencies of the simulation component interface.
8. A non-real-time distributed co-simulation system supporting dynamic model connection according to claim 3, characterized in that: The resource management and task scheduler works by dynamically decomposing tasks and allocating resources based on the logical structure of the simulation workflow and the real-time load status of each computing node.