Distributed heterogeneous TT&C communication simulation system and hierarchical time synchronization method thereof

Abstract

The application provides a distributed heterogeneous TT&C communication simulation system and a hierarchical time synchronization method thereof, relates to the technical field of TT&C communication simulation, and solves the problem that the existing time synchronization scheme cannot effectively balance the global coordination accuracy and the implementation complexity. The system first configures multiple types of distributed heterogeneous simulation nodes and divides subsets of simulation tasks, then a system-level time synchronization module completes a first-level global coarse-grained time synchronization, a system-level time synchronization module completes a second-level task time synchronization, and an engineering-level time synchronization module finally completes a third-level device self hardware time synchronization in each node; after all nodes complete the three-level time synchronization in turn, the simulation task is executed and continuously monitored. The application sorts out the time synchronization requirements between different task domains and the time synchronization requirements between different devices in the domain, and assists with model order reduction, realizes global coordination from macro task to micro operation, and ensures efficient execution of complex simulation tasks.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of TT&C (Tracking, Telemetry and Command) communication simulation, and in particular to a distributed heterogeneous TT&C communication simulation system and a hierarchical time synchronization method thereof. BACKGROUND

[0002] A TT&C communication simulation system plays a vital role in modern training and testing. A traditional system architecture is usually based on unified structures, standards and algorithms, and connects various simulation applications distributed in different geographical locations through a network, aiming to support data exchange and interoperation between real, virtual and heterogeneous platform-level simulation applications distributed in different places. Such a system can meet the basic requirements of data transmission and collaborative simulation in a relatively single and determined task.

[0003] However, with the deepening and expansion of application scenarios, the TT&C communication simulation field is facing unprecedented challenges. On the one hand, the types of simulation projects have increased dramatically, from single simulation subjects to multiple and complex simulation subjects; on the other hand, the complexity and dynamics of tasks have significantly improved, and the task patterns have evolved from a pre-determined fixed mode to a highly diversified and uncertain direction. These changes have raised almost stringent requirements for the core performance indicators of the system, such as real-time performance, synchronization, TT&C accuracy and communication speed. In particular, when dealing with emergency and emergency events, the diversity and uncertainty of task requirements have shown an explosive trend.

[0004] The inherent rigid architecture of the traditional system based on unified structures and standards is not adaptable enough to meet the above-mentioned rapidly changing and highly dynamic task requirements. This architecture is difficult to adjust flexibly to adapt to the differentiated requirements of different task subsets for system resources, communication paths and data processing logic, resulting in difficulty in guaranteeing system efficiency and even failure when facing diversified and uncertain tasks. Therefore, it has become an urgent need for the current technology development to build a distributed heterogeneous TT&C communication simulation system with stronger task adaptability and dynamic reconstruction capability.

[0005] In a distributed heterogeneous TT&C communication simulation system, a core basic problem of ensuring the cooperative work of the whole system is time synchronization. The system is composed of multiple subsystems with different architectures, different standard protocols and located in different places, which must coordinate to complete the simulation task. The importance of time synchronization is reflected in ensuring data consistency. If there is a clock deviation between different node devices, it will directly lead to the misalignment of the timestamp of data update, thereby destroying the consistency of data state and affecting the correctness of system logic. The order of transaction execution also highly depends on accurate time reference, and timing disorder may lead to operation logic error, affecting the reliability of simulation results. In addition, high-precision TT&C, real-time interaction and strict synchronization of multi-node behavior all put forward very high requirements for the synchronization accuracy of the underlying clock. The existing conventional time synchronization method often has difficulty in achieving effective balance between global coordination accuracy and implementation complexity when facing highly heterogeneous, dynamic and extremely demanding complex task scenarios, becoming a key bottleneck restricting the performance improvement of the system. SUMMARY

[0006] The purpose of the present application is to solve the problem that the time synchronization method of the existing system cannot effectively balance the global coordination accuracy and the implementation complexity when facing diversified and high dynamic tasks, therefore a distributed heterogeneous TT&C communication simulation system and a hierarchical time synchronization method thereof are proposed. The present application adopts a three-level time synchronization scheme of system-level-system-level-engineering-level decomposition, according to different simulation task subsets, sorts out the time synchronization requirements between different task domains and the time synchronization requirements between different devices within the domain, and realizes the global coordination from macro task to micro operation with the aid of model order reduction, to ensure the efficient execution of complex simulation tasks.

[0007] The present application adopts the following technical solutions to achieve the purpose:

[0008] A distributed heterogeneous TT&C communication simulation system, comprising:

[0009] A node configuration management module for configuring and managing multiple types of distributed heterogeneous simulation nodes in the simulation system, determining the nodes and their physical locations required for executing simulation tasks in the simulation system, and dividing the nodes required for the simulation tasks into corresponding multiple simulation task subsets;

[0010] A system-level time synchronization module for serving as a first-level time synchronization standard to complete the global coarse-grained time synchronization of all nodes required for the simulation task through NTP protocol;

[0011] A system-level time synchronization module for serving as a second-level time synchronization standard to complete the task time synchronization of the nodes in each simulation task subset through NTP+PTP hybrid protocol based on the global coarse-grained time synchronization for each simulation task subset;

[0012] an engineering-level time synchronization module, configured to complete hardware time synchronization of various specific devices in each node by a hybrid strategy of NTP+PTP+model order reduction as a third-level time synchronization standard based on task time synchronization;

[0013] a simulation task execution module, configured to execute the simulation task and continuously monitor time synchronization states of all nodes in the execution process when all nodes required by the simulation task sequentially complete the third-level time synchronization.

[0014] Specifically, in the node configuration management module, the physical positions of the multiple types of distributed heterogeneous simulation nodes are determined by their deployment platforms; the deployment platforms include a space-based platform, an air-based platform and a ground-based platform, and multiple simulation nodes of the same type or different types are deployed in each platform; the simulation nodes deployed in the space-based platform include high-orbit satellite static simulation nodes and medium and low-orbit satellite dynamic simulation nodes; the simulation nodes deployed in the air-based platform include air static simulation nodes and air dynamic simulation nodes; and the simulation nodes deployed in the ground-based platform include ground static simulation nodes and ground dynamic simulation nodes.

[0015] Specifically, the system-level time synchronization module specifically includes:

[0016] a time reference unification unit, configured to establish a globally unified time reference;

[0017] a coarse synchronization execution unit, configured to make the multiple types of distributed heterogeneous simulation nodes advance the simulation process under the globally unified time reference by using a distributed NTP protocol.

[0018] Preferably, the system-level time synchronization module specifically includes:

[0019] a server configuration unit, configured to configure a time source server of a corresponding level for each node in a simulation task subset, and preset different time synchronization precision requirements for time source servers of different levels; and take a corresponding index of global coarse-grained time synchronization as a reference time signal of each time source server;

[0020] a dynamic request response unit, configured to obtain a time request initiated by a low-level time source server to a high-level time source server, and generate response data carrying high-level clock information and synchronization parameters based on the high-level time source server;

[0021] a cross-level / cross-domain synchronization unit, configured to sequentially deliver the response data of the node corresponding to the high-level time source server to the node corresponding to the low-level time source server, until the nodes in the simulation task subset all obtain corresponding response data, and complete task time synchronization.

[0022] Specifically, the engineering-level time synchronization module specifically includes:

[0023] The precision detection unit is used to form the engineering simulation model of the corresponding index of the task time synchronization of each node, and further detect whether the hardware resources of each type of specific device in the node meet the resource requirements of the engineering simulation model thereof;

[0024] The dynamic reduction decision unit is used to, when the hardware resources of each type of specific device in the node meet the resource requirements of the engineering simulation model thereof, not use the model reduction scheme, and only use the NTP+PTP hybrid protocol as the hybrid strategy; otherwise, when the resource requirements of the engineering simulation model are not met, select the reduction scheme from the preset engineering-level model reduction scheme library to form the NTP+PTP+model reduction hybrid strategy;

[0025] The hybrid synchronization execution unit is used to execute the hybrid strategy corresponding to each node respectively, and complete the hardware time synchronization of each type of specific device in each node.

[0026] Preferably, the dynamic reduction decision unit is further used to select the reduction scheme from the preset engineering-level model reduction scheme library through an artificial intelligence mode according to the weight value selection result of the optimal function in the built-in preferred strategy thereof; and the weight value of the optimal function is determined by the multi-dimensional features corresponding to the extraction of a plurality of reduction schemes.

[0027] The reduction schemes in the preset engineering-level model reduction scheme library include a neural network scheme, an eigen-orthogonal decomposition scheme, an equivalent circuit scheme, a Krein subspace scheme, a dynamic modal decomposition scheme, a singular value decomposition scheme and a balanced truncation scheme.

[0028] Preferably, the engineering-level time synchronization module further comprises a preferred strategy updating unit; the preferred strategy updating unit is used to perform reverse time synchronization efficiency evaluation on the nodes that have completed the hardware time synchronization, trace back to the corresponding index of the global coarse-grained time synchronization, and iteratively update the weight value selection process of the optimal function in the built-in preferred strategy of the dynamic reduction decision unit according to the evaluation result.

[0029] The application also provides a hierarchical time synchronization method for a distributed heterogeneous TT&C communication simulation system, which comprises the following steps:

[0030] S1, configuring the distributed heterogeneous simulation nodes required by the simulation task, determining the physical positions thereof, and dividing the nodes into a plurality of simulation task subsets;

[0031] S2, performing system-level time synchronization: completing the global coarse-grained time synchronization of all nodes through the NTP protocol;

[0032] S3, performing system-level time synchronization: for each simulation task subset, on the basis of the global coarse-grained time synchronization, completing the task time synchronization of the nodes in the subset through the NTP+PTP hybrid protocol;

[0033] S4, performing engineering-level time synchronization: for each node, on the basis of task time synchronization, hardware time synchronization of various devices in the node is completed through a mixed strategy of NTP+PTP+model order reduction;

[0034] S5, when all nodes complete the three-level time synchronization, the simulation task is executed and the node time synchronization state is continuously monitored.

[0035] Further, in step S3, the system-level time synchronization is executed, specifically including:

[0036] S31, a hierarchical time source server is configured for each simulation task subset, and different levels of time source servers are preset with different time synchronization precision requirements; and the corresponding index of the global coarse-grained time synchronization is taken as the reference time signal of each time source server;

[0037] S32, a time request initiated by a low-level time source server to a high-level time source server is obtained, and response data carrying high-level clock information and synchronization parameters is generated based on the high-level time source server;

[0038] S33, the response data of the corresponding node of the high-level time source server is transmitted to the corresponding node of the low-level time source server step by step, until the nodes in the simulation task subset obtain the corresponding response data, and the task time synchronization is completed.

[0039] Further, in step S4, the engineering-level time synchronization is executed, specifically including:

[0040] S41, generating an engineering simulation model of the node according to the task time synchronization index, and detecting whether the hardware resources of various specific devices in the node meet the resource requirements of the engineering simulation model;

[0041] S42, if the hardware resources meet the requirements, a mixed protocol of NTP+PTP is used as a mixed strategy; if the hardware resources do not meet the requirements, an order reduction scheme is selected from a preset engineering-level model order reduction scheme library based on an artificial intelligence optimization strategy, and a mixed strategy of NTP+PTP+model order reduction is formed;

[0042] S43, the mixed strategy corresponding to each node is executed, and the hardware time synchronization of various specific devices in the node is completed.

[0043] As described above, since the technical solution is adopted, the application has the following advantages:

[0044] The three-level linkage time synchronization scheme is formed by three-level mapping decomposition of system-level time synchronization requirements-system-level time synchronization requirements-engineering-level time synchronization requirements.

[0045] The engineering-level time synchronization requirements are obtained after layer-by-layer decomposition, and then a suitable engineering-level model reduction scheme is selected.

[0046] The application first determines a unified time reference for the system-level time synchronization requirements, and only uses the NTP protocol to meet the coarse synchronization requirements of each element in the system. BRIEF DESCRIPTION OF DRAWINGS

[0047] The embodiments and technical solutions of the application are further illustrated by the following drawings, specifically including three drawings as follows:

[0048] Figure 1 The logical principle diagram of the three-level time synchronization process in the application is shown in the figure;

[0049] Figure 2 The deployment configuration diagram of the various distributed heterogeneous simulation nodes in the system of the application is shown in the figure;

[0050] Figure 3 The detailed flowchart of the hierarchical time synchronization method of the application is shown in the figure. DETAILED DESCRIPTION

[0051] To make the objects, technical solutions and advantages of the embodiments of the present application clearer, the following will clearly and completely describe the technical solutions in the embodiments of the present application with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only some but not all of the embodiments of the present application. The components of the embodiments of the present application described and shown in the drawings can be arranged and designed in various different configurations.

[0052] Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the claimed present application, but only represents selected embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative work are within the scope of protection of the present application.

[0053] Embodiment 1

[0054] A distributed heterogeneous TT&C communication simulation system, comprising the following functional modules:

[0055] A node configuration management module, configured to configure and manage multiple types of distributed heterogeneous simulation nodes in the simulation system, determine the nodes and their physical locations required for executing a simulation task in the simulation system, and divide the nodes required for the simulation task into corresponding multiple simulation task subsets;

[0056] A system-level time synchronization module, configured to serve as a first-level time synchronization standard, and complete global coarse-grained time synchronization of all nodes required for the simulation task through an NTP protocol;

[0057] A system-level time synchronization module, configured to serve as a second-level time synchronization standard, and complete task time synchronization of the nodes in each simulation task subset through an NTP+PTP hybrid protocol on the basis of the global coarse-grained time synchronization, for each simulation task subset;

[0058] An engineering-level time synchronization module, configured to serve as a third-level time synchronization standard, and complete hardware time synchronization of each type of specific device itself in each node through an NTP+PTP+model order reduction hybrid strategy, for each distributed heterogeneous simulation node, on the basis of the task time synchronization;

[0059] A simulation task execution module, configured to execute the simulation task and continuously monitor the time synchronization state of all nodes in the execution process, when all nodes required for the simulation task have completed the three-level time synchronization in turn.

[0060] The system of the embodiment is crucial in realizing, for example, Figure 1The three-level time synchronization process shown gradually refines from global to local, ensuring that all nodes of the distributed heterogeneous TT&C communication simulation system remain highly coordinated in the simulation task. The system-level time synchronization module initiates global coarse-grained time synchronization, establishes a unified time reference through the standard Network Time Protocol (NTP), and covers all simulation nodes, providing a foundation guarantee for system-level planning. This layer of synchronization adapts to scenarios with a large time span, such as the coordination of nodes in space-based platforms and ground platforms, ensuring that different levels of elements advance in unison in high-level planning.

[0061] Subsequently, the system-level time synchronization module expands on the basis of global synchronization and conducts task time synchronization for each simulation task subset. It uses a hybrid protocol that combines the Network Time Protocol (NTP) and the Precision Time Protocol (PTP) to achieve hierarchical time information propagation, transmitting from high-level time sources to low-level servers, improving transmission efficiency and accuracy. This layer adapts to complex cross-domain environments and corresponds to each task subset in the simulation task, such as communication interactions between certain nodes of air platforms and certain satellites in space-based platforms, ensuring real-time and reliable execution of task instructions among multiple levels of nodes.

[0062] The function of the engineering-level time synchronization module is further into each distributed heterogeneous simulation node, on the premise of task time synchronization, for specific equipment hardware time synchronization. When the time synchronization accuracy is decomposed from system level to engineering level, if the engineering-level time synchronization accuracy requirement is within the scope of the resources of each node, there is no need for model reduction; but when the accuracy requirement is very high, the engineering simulation model of the node introduces a model reduction scheme through a hybrid strategy, simplifying its multi-dimensional system description, preserving key characteristics and optimizing computational efficiency. This layer of synchronization targets specific devices such as circuits and radio frequencies within the device, ensuring the accuracy of engineering-level information transmission, while dynamically selecting a reduction scheme through an artificial intelligence strategy to respond to dynamically changing requirements.

[0063] The entire three-level synchronization process progresses from coarse to fine, ultimately achieving full-node synchronization monitoring in the simulation task execution module. Through this hierarchical mechanism, the system balances time accuracy and computational complexity, ensuring efficient operation of the simulation task.

[0064] Embodiment 2

[0065] Based on Embodiment 1, this embodiment provides a detailed or preferred introduction to the details of each functional module in the system.

[0066] In the node configuration management module, the physical locations of multiple types of distributed heterogeneous simulation nodes are determined by their deployment platforms; the deployment platforms include space-based platforms, air platforms, and ground platforms, and each platform has multiple same or different simulation nodes deployed. Figure 2An exemplary system architecture configuration is shown, in which there are distributed heterogeneous simulation nodes and their interconnection relationship.

[0067] The simulation nodes deployed in the space-based platform include high-orbit satellite static simulation nodes, which have relatively stable corresponding time reference; and medium and low-orbit satellite dynamic simulation nodes, which have greater dynamic change.

[0068] The simulation nodes deployed in the air-based platform include air static simulation nodes, such as air floating boats, measurement and control balloons, etc.; and air dynamic simulation nodes, such as unmanned aerial vehicles, cruise missiles, etc.

[0069] The simulation nodes deployed in the ground-based platform include ground static simulation nodes, such as fixed ground measurement and control stations, etc.; and ground dynamic simulation nodes, such as mobile ground stations, manned vehicles, unmanned vehicles and unmanned boats, etc.

[0070] Among the above types of nodes, the time reference corresponding to the static simulation nodes is relatively stable, and the dynamic simulation nodes have greater time variability. Using the above different types of nodes, different measurement and control communication scene architectures can be simulated and built according to different simulation task requirements, and simulation tasks can be performed.

[0071] In this embodiment, the system-level time synchronization module specifically includes:

[0072] The time reference unification unit is configured to establish a globally unified time reference.

[0073] The coarse synchronization execution unit uses the distributed NTP protocol to make the multiple types of distributed heterogeneous simulation nodes advance the simulation process under the globally unified time reference.

[0074] The system-level time synchronization module has a large time span and low time synchronization requirements, so it can meet the requirements only by using the NTP protocol, so that the entire simulation system architecture is globally time-synchronized, and different elements in the simulation task scene can advance the simulation process under the overall time reference controlled by the system.

[0075] In this embodiment, the system-level time synchronization module specifically includes:

[0076] The server configuration unit is configured to configure a corresponding level of time source server for each node in a simulation task subset, and different levels of time source servers are preconfigured with different time synchronization accuracy requirements; and the corresponding index of the global coarse-grained time synchronization is used as the reference time signal of each time source server.

[0077] The dynamic request response unit is configured to obtain a time request initiated by a low-level time source server to a high-level time source server, and generate response data carrying high-level clock information and synchronization parameters based on the high-level time source server.

[0078] A cross-layer / cross-domain synchronization unit is configured to pass the response data of the high-level time source server corresponding node to the low-level time source server corresponding node level by level until the nodes in the entire simulation task subset obtain the corresponding response data, and the task time synchronization is completed.

[0079] The system-level time synchronization module is based on the NTP+PTP hybrid protocol to realize the task time synchronization of each simulation task subset. It corresponds to the time synchronization requirement for completing a sub-task, and is executed in a cross-layer and cross-domain multi-node simulation environment. Cross-layer means that a simulation task subset contains nodes in a space-based platform and nodes in an air-based platform. Cross-domain means that the nodes in a simulation task subset include nodes belonging to a control domain for transmitting instructions and nodes belonging to a state communication domain for transmitting text interactive information.

[0080] The time source servers of nodes at different levels have different accuracy presets, and the highest level time source usually has the highest accuracy, such as the high-orbit satellite static simulation node of the space-based platform. The lower level servers regularly send time requests to the higher level servers, and the higher level servers respond to these requests using their own clocks or synchronization information with higher level time sources. In this way, time information is propagated from higher level time sources to lower level servers to achieve time synchronization between task domains in the entire TT&C communication network.

[0081] In this embodiment, the engineering-level time synchronization module specifically includes:

[0082] The precision detection unit is configured to form an engineering simulation model of each node according to the corresponding index of the task time synchronization, and then detect whether the hardware resources of various specific devices in the node meet the resource requirements of the engineering simulation model.

[0083] The dynamic reduction decision unit is configured to, when the hardware resources of various specific devices in the node meet the resource requirements of the engineering simulation model, not use the model reduction scheme, and only use the NTP+PTP hybrid protocol as the hybrid strategy; otherwise, when the resource requirements of the engineering simulation model are not met, select a reduction scheme from the preset engineering-level model reduction scheme library to form an NTP+PTP+model reduction hybrid strategy.

[0084] The hybrid synchronization execution unit is configured to execute the hybrid strategy corresponding to each node respectively to complete the hardware time synchronization of various specific devices in each node.

[0085] As a preferred embodiment, the dynamic reduction decision unit is further configured to select a reduction scheme from the preset engineering-level model reduction scheme library according to the weight value selection result of the optimal function in the built-in preferred strategy by using artificial intelligence; and the weight value of the optimal function is determined by the multi-dimensional features extracted from the multiple reduction schemes.

[0086] The time synchronization of the engineering level is extremely high in the resolution requirement index due to the internal synchronization requirements of specific devices in specific nodes, such as circuits, antennas, radio frequency components, etc. Therefore, the embodiment first converts the task synchronization index by the precision detection unit. The unit maps the upper time synchronization requirement into the engineering simulation model of the specific device, and evaluates the support ability of the hardware resources in the node to the model requirement in real time. When the detection confirms that the resources are sufficient, the dynamic order reduction decision unit maintains the basic protocol combination and only uses the hybrid strategy of the network time protocol and the precise time protocol to complete the synchronization. At this time, unnecessary model simplification can be effectively avoided, the original high precision of the device within the capability range is guaranteed, and even nanosecond engineering application precision can be achieved.

[0087] However, the hardware resources of the devices in a single node cannot bear the extremely high precision calculation requirement of the engineering simulation model, so the dynamic order reduction decision unit activates its intelligent decision-making process. The unit dynamically selects the optimal solution in the engineering level model order reduction scheme library according to the preset multi-dimensional feature evaluation system.

[0088] In addition, as a preferred embodiment of the present embodiment, the engineering level time synchronization module further comprises a preferred strategy updating unit. The preferred strategy updating unit is used for reverse time synchronization performance evaluation of the node that completes the hardware time synchronization, until the corresponding index of the global coarse-grained time synchronization is traced back, and the weight selection process of the optimal function in the preferred strategy built in the dynamic order reduction decision unit is iteratively updated according to the evaluation result.

[0089] The iterative update of this part enables the feature weight of the optimal function in the preferred strategy to be continuously optimized through historical performance data, so that the artificial intelligence strategy can accurately adapt to the real-time working conditions of different devices. After selecting the scheme, the system automatically generates a three-layer hybrid strategy that integrates the protocol and the order reduction technology.

[0090] The hybrid synchronization execution unit is responsible for the final implementation of the strategy. The unit customizes the execution of the corresponding hybrid strategy for each node to achieve hardware time synchronization at the device level. Especially when the order reduction scheme is used, the system significantly reduces the computational load of devices such as radio frequency circuits by retaining key characteristics and simplifying multi-dimensional descriptions, so that high-precision synchronization can still be achieved under limited resource conditions. The whole process forms a closed loop of resource awareness and intelligent decision-making, which not only meets the synchronization requirements of the engineering level in the order of microseconds, but also ensures the overall operation efficiency of the system.

[0091] In the embodiment, the order reduction schemes in the preset engineering level model order reduction scheme library include: a neural network scheme, an eigen-orthogonal decomposition scheme, an equivalent circuit scheme, a Krein subspace scheme, a dynamic modal decomposition scheme, a singular value decomposition scheme, and a balanced truncation scheme.

[0092] The adaptation objects of each reduction scheme are inconsistent, but according to the preset multi-dimensional feature evaluation system, such as circuit, signal, image, antenna, radio frequency; linear, nonlinear; large data volume, small data volume and other indicators, multi-dimensional classification can be carried out, the characteristics of each reduction scheme are extracted, thereby serving as the parameters and weights of the optimal function in the optimization strategy, to select the finally determined reduction scheme to be used.

[0093] The pros and cons of each reduction scheme under its scene application are different, which can be selected according to actual conditions, as follows:

[0094] The neural network scheme is to learn the low-dimensional representation of the high-dimensional system by using the deep learning model, which is mainly applied to complex electromagnetic environment real-time simulation and interference mode fast prediction in electronic warfare system scenes; the advantage of this scheme is to adapt to strong nonlinearity and end-to-end training, but a large amount of data is needed to train the model.

[0095] The proper orthogonal decomposition scheme is to extract the dominant mode (orthogonal basis) by collecting the snapshot of the system response, and to construct a low-dimensional subspace, which is mainly applied to electromagnetic field simulation, circuit transient analysis and other scenes; the advantage of this scheme is to adopt a data-driven way and adapt to nonlinear systems, but it depends on training data and has limited extrapolation ability.

[0096] The equivalent circuit scheme is to simplify the distributed parameter system into a lumped parameter circuit, which is mainly applied to the interconnection model simplification of high-speed digital circuits and the fast simulation of power electronic devices; the advantage of this scheme is that it is physically intuitive and can be compatible with Spice simulator, but the precision will decrease at high frequency.

[0097] The Krein subspace scheme is to match the transfer function matrix of the original system at a certain frequency point, and to construct a low-dimensional approximation, which is mainly applied to large-scale linear circuit model reduction and impedance characteristic simplification of chip power distribution network; the advantage of this scheme is to retain the frequency domain characteristics and efficient calculation, but it is only suitable for linear time-invariant systems.

[0098] The dynamic modal decomposition scheme is to extract dynamic modes from time series data and predict system evolution, which is mainly applied to target dynamic modeling in radar signal processing and communication network traffic prediction; the advantage of this scheme is that it does not need explicit system equations and is suitable for experimental data, but it is sensitive to noise.

[0099] The singular value decomposition scheme is to perform singular value decomposition on the system matrix, and to retain the subspace corresponding to the significant singular values, which is mainly applied to compressed sensing in image / signal processing and multi-input multi-output system channel reduction; this scheme is numerically stable and suitable for high-dimensional data, but it needs to be combined with other methods for nonlinear systems.

[0100] Balanced truncation method is to separate controllability and observability of system state by balanced transformation, and to truncate weakly coupled states. It is mainly applied to filter design, communication channel modeling, and reduced order equivalent of RF devices. Its main advantages are clear global error bound and stability preservation, but it has high computational complexity.

[0101] Embodiment 3

[0102] Based on the above-mentioned embodiments, the embodiment provides a hierarchical time synchronization method of a distributed heterogeneous TT&C communication simulation system, Figure 3 A flow diagram of the hierarchical time synchronization method is shown, which can be synchronized. The method comprises the following key steps:

[0103] S1, configure the distributed heterogeneous simulation nodes required by the simulation task, determine their physical positions, and divide the nodes into multiple simulation task subsets;

[0104] S2, perform system-level time synchronization: complete global coarse-grained time synchronization of all nodes through NTP protocol;

[0105] S3, perform system-level time synchronization: for each simulation task subset, on the basis of global coarse-grained time synchronization, complete task time synchronization of nodes in the subset through NTP+PTP hybrid protocol;

[0106] S4, perform engineering-level time synchronization: for each node, on the basis of task time synchronization, complete hardware time synchronization of various devices in the node through NTP+PTP+model reduction hybrid strategy;

[0107] S5, when all nodes complete the three-level time synchronization, execute the simulation task and continuously monitor the node time synchronization state.

[0108] In step S3 of the embodiment, system-level time synchronization is performed, which specifically comprises:

[0109] S31, configure a hierarchical time source server for each simulation task subset, and different levels of time source servers are preset with different time synchronization precision requirements; and take the corresponding index of global coarse-grained time synchronization as the reference time signal of each time source server;

[0110] S32, obtain the time request initiated by the low-level time source server to the high-level time source server, and generate response data carrying high-level clock information and synchronization parameters based on the high-level time source server;

[0111] S33, the response data of the corresponding node of the high-level time source server is transmitted to the corresponding node of the low-level time source server level by level, until the nodes in the entire simulation task subset obtain the corresponding response data, and the task time synchronization is completed.

[0112] In step S4 of the embodiment, engineering-level time synchronization is performed, specifically including:

[0113] S41, generating an engineering simulation model of the task time synchronization index according to the node, detecting whether the hardware resources of various specific devices in the node meet the resource requirements of the engineering simulation model;

[0114] S42, if the hardware resources meet the requirements, using NTP+PTP hybrid protocol as the hybrid strategy; if the hardware resources do not meet the requirements, selecting a reduction scheme from the pre-set engineering-level model reduction scheme library based on the artificial intelligence optimization strategy to form a hybrid strategy of NTP+PTP+model reduction;

[0115] S43, executing the hybrid strategy corresponding to each node to complete the hardware time synchronization of various specific devices in the node.

[0116] In the embodiment, after all nodes complete the three-level time synchronization, the time synchronization efficiency can be immediately evaluated in the reverse order of engineering-level-system-level-system-level to verify whether it meets the engineering-level requirements of the simulation task. If it can meet the requirements, the time synchronization is completed, and the simulation task can be executed and continuously monitored. If it cannot meet the requirements, the execution process of the engineering-level time synchronization is re-entered, and the selection of the reduction scheme is adjusted in the case of applying model reduction to achieve time synchronization optimization.

[0117] After each time synchronization process of the method is completed, the efficiency evaluation result of the three-level time synchronization can be used to further optimize the optimal function of the optimization strategy used to select the reduction scheme during the engineering-level time synchronization, so that the optimization strategy itself can be dynamically adjusted, thereby improving the accuracy of the model reduction scheme, and when the scheme parameters change, the scheme is supplemented to the engineering-level model reduction scheme library, which better supports the high-precision time synchronization requirements of the subsequent distributed heterogeneous measurement and control communication simulation system.

Claims

1. A distributed heterogeneous TT&C communication simulation system, characterized in that, The simulation system comprises: a node configuration management module, configured to configure and manage multiple types of distributed heterogeneous simulation nodes in the simulation system, determine nodes and their physical locations required for executing a simulation task in the simulation system, and divide the nodes required for the simulation task into multiple simulation task subsets; a system-level time synchronization module, configured to serve as a first-level time synchronization standard, and complete global coarse-grained time synchronization of all nodes required for the simulation task through an NTP protocol; a system-level time synchronization module, configured to serve as a second-level time synchronization standard, and complete task time synchronization of nodes in each simulation task subset through an NTP+PTP hybrid protocol on the basis of the global coarse-grained time synchronization; an engineering-level time synchronization module, configured to serve as a third-level time synchronization standard, and complete hardware time synchronization of each type of specific device in each node through an NTP+PTP+model reduction hybrid strategy on the basis of the task time synchronization; a simulation task execution module, configured to execute the simulation task and continuously monitor the time synchronization state of all nodes during the execution when all nodes required for the simulation task have completed the three-level time synchronization in turn. The system-level time synchronization module comprises: a server configuration unit, configured to configure a corresponding level of time source server for nodes in each simulation task subset, and preset different time synchronization accuracy requirements for different levels of time source server; and use the corresponding index of the global coarse-grained time synchronization as the reference time signal of each time source server; a dynamic request response unit, configured to obtain a time request initiated by a low-level time source server to a high-level time source server, and generate response data carrying high-level clock information and synchronization parameters based on the high-level time source server; a cross-level / cross-domain synchronization unit, configured to pass the response data of the corresponding node of the high-level time source server to the corresponding node of the low-level time source server level by level, until the nodes in the entire simulation task subset obtain the corresponding response data and complete the task time synchronization. The engineering-level time synchronization module comprises: an accuracy detection unit, configured to form an engineering simulation model of each node by using the corresponding index of the task time synchronization, and then detect whether the hardware resources of each type of specific device in the node meet the resource requirements of the engineering simulation model; a dynamic reduction decision unit, configured to, when the hardware resources of each type of specific device in the node meet the resource requirements of the engineering simulation model, not use a model reduction scheme, and only use an NTP+PTP hybrid protocol as the hybrid strategy; otherwise, select a reduction scheme from a preset engineering-level model reduction scheme library, and form an NTP+PTP+model reduction hybrid strategy; a hybrid synchronization execution unit, configured to execute the corresponding hybrid strategy of each node, and complete the hardware time synchronization of each type of specific device in each node.

2. The distributed heterogeneous TT&C communication simulation system according to claim 1, characterized in that: In the node configuration management module, the physical positions of the multiple types of distributed heterogeneous simulation nodes are determined by their deployment platforms; the deployment platforms include space-based platforms, air-based platforms, and ground-based platforms, and multiple simulation nodes of the same type or different types are deployed in each platform; The simulation nodes deployed in the space-based platforms include high-orbit satellite static simulation nodes and medium-orbit and low-orbit satellite dynamic simulation nodes; the simulation nodes deployed in the air-based platforms include air static simulation nodes and air dynamic simulation nodes; The simulation nodes deployed in the ground-based platforms include ground static simulation nodes and ground dynamic simulation nodes.

3. The distributed heterogeneous TT&C communication simulation system according to claim 1, wherein, The system-level time synchronization module specifically includes: a time reference unification unit configured to establish a globally unified time reference; a coarse synchronization execution unit configured to make the multiple types of distributed heterogeneous simulation nodes advance the simulation process under the globally unified time reference by using a distributed NTP protocol.

4. The distributed heterogeneous TT&C communication simulation system according to claim 1, characterized in that: The dynamic order reduction decision unit is further configured to select an order reduction scheme from a preset engineering-level model order reduction scheme library by using an artificial intelligence method according to a weight value selection result of an optimal function in the built-in optimal strategy; and the weight value of the optimal function is determined by the multiple dimensions of the extracted features in the multiple order reduction schemes. The order reduction schemes in the preset engineering-level model order reduction scheme library include a neural network scheme, an eigen-orthogonal decomposition scheme, an equivalent circuit scheme, a Krein subspace scheme, a dynamic modal decomposition scheme, a singular value decomposition scheme, and a balanced truncation scheme.

5. The distributed heterogeneous TT&C communication simulation system according to claim 4, characterized in that: The engineering-level time synchronization module further includes an optimal strategy updating unit; the optimal strategy updating unit is configured to perform a reverse time synchronization efficiency evaluation on the nodes that have completed the hardware time synchronization, trace back to the corresponding indicators of the global coarse-grained time synchronization, and iteratively update the weight value selection process of the optimal function in the built-in optimal strategy of the dynamic order reduction decision unit according to the evaluation result.

6. A hierarchical time synchronization method for the distributed heterogeneous TT&C communication simulation system according to any one of claims 1-5, characterized in that, The method includes the following steps: S1, configuring the distributed heterogeneous simulation nodes required for the simulation task, determining the physical positions of the nodes, and dividing the nodes into multiple simulation task subsets; S2, performing system-level time synchronization: completing the global coarse-grained time synchronization of all nodes by using an NTP protocol; S3, performing system-level time synchronization: for each simulation task subset, completing the task time synchronization of the nodes in the subset by using an NTP+PTP hybrid protocol on the basis of the global coarse-grained time synchronization; S4, performing engineering-level time synchronization: for each node, completing the hardware time synchronization of the devices in the node by using an NTP+PTP+model order reduction hybrid strategy on the basis of the task time synchronization; S5, after all the nodes complete the three-level time synchronization, performing the simulation task and continuously monitoring the node time synchronization state.

7. The hierarchical time synchronization method of claim 6, wherein, In step S3, the system-level time synchronization is specifically performed as follows: S31, configuring a hierarchical time source server for each simulation task subset; different levels of the time source servers are preset with different time synchronization accuracy requirements; and the corresponding indicators of the global coarse-grained time synchronization are used as the reference time signals of the time source servers; S32, obtaining a time request initiated by a low-level time source server to a high-level time source server, and generating response data carrying high-level clock information and synchronization parameters based on the high-level time source server; S33, the response data of the high-level time source server corresponding node is transmitted to the low-level time source server corresponding node level by level until the nodes in the entire simulation task subset obtain the corresponding response data, and the task time synchronization is completed.

8. The hierarchical time synchronization method of claim 6, wherein, In step S4, the engineering level time synchronization is performed, specifically including: S41, generating an engineering simulation model of the node according to the task time synchronization index, and detecting whether the hardware resources of various specific devices in the node meet the resource requirements of the engineering simulation model; S42, if the hardware resources meet the requirements, using NTP+PTP hybrid protocol as a hybrid strategy; if the hardware resources do not meet the requirements, selecting a reduction scheme from a pre-set engineering level model reduction scheme library based on an artificial intelligence optimization strategy to form a hybrid strategy of NTP+PTP+model reduction; S43, executing the hybrid strategy corresponding to each node to complete the hardware time synchronization of various specific devices in the node itself.

Images