Joint simulation model integration method and tool, and medium

Abstract

The invention belongs to the technical field of system simulation, and relates to a joint simulation model integration method and tool and a medium. The method comprises the following steps: firstly, analyzing semantics of a model, including meta-model data analysis, conceptual model description analysis, static semantic analysis and dynamic semantic analysis; then performing interface function event association mapping to obtain a mapping relationship between a model interface and a standard interface; next, configuring a model integration environment, including a running environment, model assembly and scheduling parameters, a generation directory file, an agent model / model scheduling file and the like; and finally, integrating and packaging the packaged proxy model / model scheduling file and the like, and sending the packaged file to a model resource library. According to the method, automatic integration of most types of models (including components, entities, semi-physical models and systems) in the technical field of system simulation can be realized, the integrated polymorphic models can be expanded and reused, the simulation efficiency is greatly improved, and various existing model resources are fully utilized to shorten the integration period.

Description

Technical Field

[0001] The invention belongs to the technical field of joint simulation and relates to a joint simulation model integration method, tool and medium. Background Art

[0002] In the field of joint simulation technology, there are many elements involved, complex relationships, high modeling technology difficulty and large workload. It is very necessary to use existing model resources at low cost and high efficiency. However, in the actual simulation experiment process, there are a large number of mature model resources with different sources, different types, different resolutions, and different time and space benchmarks. With the continuous improvement of the scale, function and logical complexity of joint simulation systems, it is difficult to use existing models, the resource reuse rate is low, and the integration verification cycle is long, which cannot meet the needs of rapid multi-modal heterogeneous model (such as components, entities, semi-physical and systems) integration and application.

[0003] A semi-physical model is also called a semi-physical model. It refers to a model that introduces a part of the simulated object system into the simulation loop in the form of a physical object (or physical model) based on the simulation research content, and describes the rest of the simulated object system with a mathematical model, which is converted into a simulation calculation model, and a joint simulation model of real-time mathematical simulation and physical simulation is performed.

[0004] For example, during the integration process, there are problems such as type incompatibility, inconsistent coordinates, inconsistent dimensions, irregular syntax, unclear semantics, mismatched mappings, different granularity, time asynchrony, and non-sharing of resources among multi-modal heterogeneous models with different system frameworks, diverse sources, complex types, large resolution differences, and prominent differences in spatiotemporal benchmarks. These problems are very trivial and require familiarity with the detailed definitions of the model and simulation system interface at the same time, so that the integrators and modelers need to negotiate and resolve them one by one. Model integration in the existing technology must be carried out through customized conversion interfaces, lacking standardization and process-based development, making development difficult, integration efficiency low, and basically unable to achieve automation and scale. Therefore, there is an urgent need for the support of efficient auxiliary integration methods to help simulators achieve rapid reuse of large-scale multi-modal heterogeneous models. Summary of the invention

[0005] The purpose of the present invention is to provide a joint simulation model integration method and tool to solve the problems of high technical threshold, large development workload, low integration efficiency and high integration failure rate in the rapid integration and reuse of existing polymorphic heterogeneous models. The method can realize the standardized integration of four types of polymorphic heterogeneous models, namely, component, entity, system and semi-physical, provide support for simulation system calls and provide them for simulation personnel to use.

[0006] To solve this technical problem, the technical solution of the present invention is:

[0007] On the one hand, a joint simulation model integration method is provided, wherein the integration method is used to integrate and call four types of multi-modal heterogeneous models: component, entity, semi-physical and system; the integration steps are as follows:

[0008] Step 1: semantic analysis of model parameters: semantic analysis is performed on the interface parameters of the polymorphic heterogeneous model. The process of semantic analysis is divided into the following stages:

[0009] Metamodel data analysis, conceptual model description analysis, static semantic analysis, dynamic semantic analysis;

[0010] S1. Metamodel data analysis includes: extracting and analyzing basic model information such as model ID, model name, model version number, model description, model discipline, model creator, model creation time, model modifier, model modification time, model application scope, usage restrictions, model confidentiality level, etc.;

[0011] Metamodel data analysis analyzes and describes the model from a pragmatic level, providing some node knowledge for the semantic analysis knowledge base. It also provides key information for rapid retrieval of model applications in scenarios after the model is integrated into the database, improving the convenience of management and use of the model after it is integrated into the database.

[0012] S2. Conceptual model description and analysis includes: extracting and analyzing model interaction patterns, describing the model's internal interaction process in the form of event sequences between entities; the elements involved in this process include model state transitions, entity types, and event types.

[0013] Through the analysis of the conceptual model description, the model interaction mode, interaction process, state transition, interaction-related entity types and event types are analyzed and described from the semantic level. Each model interaction mode description completes a specific function through one or more events, one event corresponds to an action, and each event is associated with an event inside the model or an external interaction model entity.

[0014] The conceptual model description analysis can describe the interaction process inside or outside the model, improve the model interaction, combination ability and interoperability, and help enhance the ability to understand data exchange.

[0015] S3, static semantic analysis includes: extracting the model interface function list and analyzing the input and output parameter ID, input and output parameter name, input and output parameter data type, input and output parameter unit, input and output parameter value upper limit, input and output parameter value lower limit and input and output parameter row number and column number in the input and output data file;

[0016] S4. Dynamic semantic analysis includes: obtaining the pointer of the model external interface function, then calling each model external interface function and checking the model output data, and analyzing and checking the output data. When calling the model external interface function, it is necessary to obtain the data type, unit and value range of the corresponding external interface input parameter based on the results of static semantic analysis, and then use random data generation technology or manually fill in data randomly to generate the specific value of the input parameter of the model external interface function.

[0017] Through the static semantic analysis and dynamic semantic analysis, the interfaces and parameters of the multi-modal heterogeneous models are analyzed, described and verified from the semantic and grammatical levels, so as to check whether the model has defects, whether the model can be called normally and accurately reflect the pragmatic and semantic information described in the metamodel data analysis and conceptual model description analysis. Models with problems will not be integrated further, and the preliminary screening of the integrated model will improve the efficiency of model integration.

[0018] Step 2: Model interface association mapping: Perform event association mapping on the interface functions of the multi-modal heterogeneous model.

[0019] For the model interface function list in the semantic analysis result described in step 1, event association is mapped to the final required model specification framework;

[0020] The association mapping process refers to: mapping the interface function list in the original model to the model interface function list after integration, wherein the model interface function list after integration refers to the standard interface required by the final model specification framework;

[0021] The model interface association mapping can map the model interface function list in a graphical and visual way, automatically generate mapping code, and improve the combing of interface integration logic and model integration efficiency.

[0022] Step 3: Model integration configuration: Integrate and configure the multi-modal heterogeneous models

[0023] Configure the imported multi-modal heterogeneous model integration environment, including: configuring the model running environment, setting model integration parameters, generating model catalog files, generating model assembly files, generating model scheduling files and other integration configuration files.

[0024] The integration configuration file refers to all the procedural files generated as needed during the integration configuration of multi-modal heterogeneous models; the integration configuration process generates the integrated target model catalog file, model assembly file and model scheduling file based on the semantic analysis results of the model and the model integration parameters. These three types of model description files are provided to the target simulation system for integration call reference, solving the problem of model configuration in multi-modal heterogeneous integration.

[0025] Step 4: Model integration and storage: The multi-modal heterogeneous models are integrated and stored in the database; the model integration and storage means that the models are packaged and sent to the model resource library for unified management after integration;

[0026] According to the integration configuration file described in step three, all relevant files of the multi-modal heterogeneous model are packaged and integrated as needed, and finally sent to the model resource library for storage, or directly provided to the simulation system for link calling.

[0027] Furthermore, when the method integrates a semi-physical model or a system model, a digital proxy model needs to be created before step 1, and the digital proxy model is used to replace the semi-physical model or the system model. Specifically, in step 4, when the multi-modal heterogeneous model is a component or a physical model, the original model file and the integration configuration file are packaged and integrated; when the multi-modal heterogeneous model is a system or a semi-physical model, the digital proxy model and the integration configuration file are packaged and integrated.

[0028] This method can be directly integrated for component or physical models. For semi-physical / system models, it is necessary to create a digital proxy model in advance and use the digital proxy model to replace the original semi-physical / system model, which can greatly improve the optimization design efficiency and reduce the optimization difficulty.

[0029] In particular, in aircraft co-simulation, the input interface parameters of the aircraft physical model (or the digital proxy model created by the semi-physical / system model) include waypoints, cruise routes, cruise airspace, maneuvering control instructions, input performance parameters and simulation time; the output interface parameters include carrier motion status information, fuel consumption rate, longitude, latitude, altitude, pitch angle, roll angle, yaw angle and speed scalar. In order to integrate the input parameters required by the aircraft physical model and combine the aircraft characteristics, the custom data types in the aircraft physical model are as follows:

[0030] Define the waypoint data structure that the aircraft model needs to input, including the three-dimensional position coordinates composed of longitude, latitude and altitude on the map. Define the cruise route data structure that the aircraft model needs to input, including a list of waypoints. Define the cruise area data structure that the aircraft model needs to input, including the airspace surrounded by the cruise route. Define the maneuvers that the aircraft model needs to input through maneuvering control instructions, in addition to basic acceleration, deceleration, circling, turning, etc., there are also special somersaults, separation S, rolls, J-turns, leaf drifts, barrel rolls, etc. Define the additional parameters of the above maneuvers, such as acceleration, turning angle, circling altitude, etc. Define the performance parameters that the aircraft model needs to input, including speed, altitude, range, flight time, load, etc.

[0031] The standard interface functions of the aircraft entity model (or digital proxy model) include creating an instance (for registering the model), initializing, subscribing to data, stepping (for passing in model data before single-step simulation), publishing data, and deregistering an instance (for model destruction).

[0032] Optionally, the model parameter semantic analysis process further includes: intelligent semantic analysis and manual semantic analysis.

[0033] The intelligent semantic analysis refers to the automatic analysis of extracted semantic parameters by searching the semantic analysis knowledge base or other intelligent methods, thereby achieving the function of automatically completing the semantic parameters and automatically correcting them according to the recommended values, thereby improving the automation level and integration efficiency of the model integration.

[0034] The manual semantic analysis means that in the process of semantic parameter extraction and analysis of output result data, manual participation is required to select or set configuration parameters (the configuration parameters shown include model model, corresponding first-level classification, second-level classification, main purpose, functional components, internal and external interface functions, model library dependencies, parameter dimensions and coordinate systems, etc.).

[0035] Specifically, during the semantic analysis of model parameters, parameters can be selected or set based on empirical knowledge, or a semantic analysis knowledge base can be referenced or queried. The combination of intelligent semantic analysis and manual semantic analysis can not only improve the efficiency of model integration, but also achieve the accuracy of semantic analysis.

[0036] The model interface association mapping in step 2 includes automatic association mapping or manual association mapping.

[0037] The automatic association mapping refers to mapping the model interface function list to the standard interface of the required model specification framework one by one according to the automatic analysis results of the semantic analysis knowledge base.

[0038] The manual association mapping refers to manually mapping the model interface function list to the standard interface of the required model specification framework one by one through the experience and knowledge of the integrator, and also includes manually correcting the results of the automatic association mapping.

[0039] The combination of automatic association mapping and manual association mapping can not only improve the efficiency of model integration, but also achieve the accuracy of model interface association mapping.

[0040] The model specification framework in step 2 includes FMI specification, BOM specification, SMP specification, etc., or other custom specifications, wherein the custom specification refers to the standard interface specification of the integration engine defined by the user.

[0041] The event association mapping strategies in step 2 include "one-to-one", "one-to-many", "many-to-many", "many-to-one", "one-to-zero", and "zero-to-one". The association mapping logics used based on the strategies include adding functions, deleting functions, renaming functions, merging functions, decomposing functions, and rearranging APIs.

[0042] In step 3, when the system environment for generating the model is not compatible with the system environment for running the model, it is necessary to configure the model running environment, such as deploying a virtualized environment in advance. The virtualized environment is deployed by encapsulating the model to be integrated in the virtual model running environment, and then communicating with the target simulation system in the heterogeneous model running system environment through the model service bus. The virtual model running environment includes multiple implementation methods such as virtual machines and containers.

[0043] The use of virtualized environment deployment can expand the integration method of multi-modal heterogeneous models, so that cross-platform black box multi-modal models can also be integrated and called, improving the reusability of existing multi-modal heterogeneous models.

[0044] In step three, during the integration configuration of multi-modal heterogeneous models, it is necessary to set the model integration parameters, including the integration target type, target interface format, target language, target bit number, communication interface type, target operating environment, dependencies, integration model name, integration description, host environment, etc.

[0045] The simulation system refers to a set of software and hardware systems that implements the model of the simulated system in accordance with the requirements of the simulation task to conduct simulation experiments, including simulation equipment, operators and models.

[0046] On the other hand, a joint simulation model integration tool is provided, which includes a model parameter semantic analysis function module, a model interface association mapping function module, a model integration configuration function module and a model integration warehousing function module.

[0047] The model parameter semantic analysis function module includes: a meta-model data analysis unit, a concept model description analysis unit, a static semantic analysis unit and a dynamic semantic analysis unit:

[0048] The metamodel data analysis unit analyzes and describes the model from the pragmatic level. The unit converts the basic information of the model to be integrated into an analysis file in XML format and outputs it, providing partial node knowledge for the subsequent semantic analysis knowledge base.

[0049] Specifically in the integration tool system interface: users can enter the basic information of the model to be integrated through the UI interface, including model ID, model name, model version number, model description, model discipline, model creator, model creation time, model modifier, model modification time, model application scope, usage restrictions, model confidentiality and other basic model information, and output an analysis file in XML format to provide partial node knowledge for the subsequent semantic analysis knowledge base.

[0050] The conceptual model description and analysis unit analyzes and describes the model interaction mode, interaction process, state transition, interaction-related entity types and event types from a semantic level. The unit allows the user to input the model interaction mode through the UI interface, define the event type and state transition conditions of each model interaction mode, define the model entity type associated with each event, and output an analysis file in XML format, which can describe the interaction process inside or outside the model.

[0051] The static semantic analysis unit enables the user to input relevant information of the model to be integrated through the UI interface, including the input and output parameter ID, input and output parameter name, input and output parameter data type, input and output parameter unit, input and output parameter value upper limit, input and output parameter value lower limit, and input and output parameter row number and column number in the input and output data file, and outputs an analysis file in XML format to provide basic data for the implementation of subsequent model encapsulation of various functional units.

[0052] The dynamic semantic analysis unit supports extracting and dynamically calling the model's external interface functions, inputs initial parameters in a certain format for the model input interface data format, then runs the model file for trial operation, outputs the model's calling interface return data, and implements the model interface verification function.

[0053] The model interface association mapping function module maps the interface functions in the model to be integrated with the standard interfaces required by the final model specification framework, and automatically generates mapping codes; the module inputs the model interface function list and the mapping connection relationship input on the interface, and outputs a relationship mapping code file; specifically, in the integration tool system interface: the interface function list in the model to be integrated and the standard interface list required by the final model specification framework are listed separately, and the interface functions are mapped one by one by connecting lines.

[0054] The model integration configuration function module is responsible for configuring the imported multi-modal heterogeneous model integration environment; the model integration configuration function module includes a virtualization environment deployment unit and a parameter configuration unit:

[0055] The parameter configuration unit enables the user to input parameters to configure the model operating environment and set model integration parameters through the UI interface, and outputs integrated configuration files such as model catalog files, model assembly files, and model scheduling files in XML format.

[0056] The virtualized environment deployment unit encapsulates and deploys the model to be integrated in the virtual model running environment, and then communicates with the target simulation system in the heterogeneous model running system environment through the model service bus. The virtual model running environment includes multiple implementation methods such as virtual machines and containers. The input of this unit is the model library to be encapsulated and the integration configuration file, and the output is the deployment directory and startup method of the virtualized environment.

[0057] The model integration and warehousing function module will package and integrate all relevant files of the multi-modal heterogeneous model as needed, and finally send them to the model resource library for storage, or directly provide them to the target simulation system for link call. This module allows users to select the required file name on the UI interface, select the model resource library address to be stored, fill in the package name, and finally output a compressed package and send it to the specified location.

[0058] The present invention also provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and the computer program implements the functions of the integrated tool when executed by a processor.

[0059] The beneficial effects of the present invention are:

[0060] The present invention can realize the integration of most types of models in the field of joint simulation technology, and can expand and reuse the integrated subject models, which is convenient and flexible to use. It greatly improves the efficiency of simulation integration, expands the scope of model application, and effectively and fully utilizes various existing model resources, as follows:

[0061] (1) The model parameter semantic analysis described in this method describes and analyzes the interface parameters of multi-modal heterogeneous models, including metamodel data analysis, conceptual model description analysis, static semantic analysis, and dynamic semantic analysis. It clarifies the model description items and model verification methods required before model integration, standardizes the interpretability of model behavior, effectively determines the availability of the model to be integrated, and realizes model inspection and understanding before the implementation of model integration.

[0062] (2) The semantic analysis of model parameters and the association mapping of the model interface described in this method demonstrate the key steps in the model integration process, among which the intelligent semantic analysis and automatic association mapping functions provide an efficient model integration method, realize automatic model packaging and integration, and effectively improve the model integration efficiency.

[0063] (3) This method provides a rapid integration solution for models of different forms (including components, entities, semi-physical objects, and systems), enabling efficient integration of models of different forms on the same platform and truly realizing the interconnection of complex elements in multidisciplinary joint simulation.

[0064] (4) In particular, when the physical model is an aircraft entity model, the aircraft entity model of the present invention customizes multiple data types, which can not only meet the tight coupling of aircraft entity attributes, but also meet the cross-platform operation and call of the model. It is suitable for the association mapping of the model interface and the semantic analysis of the model parameters, and is convenient for the integrated configuration and integrated storage of the model.

[0065] (5) The virtualized environment deployment described in this method provides a feasible integration method for cross-platform black box models, realizes the integration mode of model cross-platform operation and call, and expands the application scope of the model. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] In order to more clearly illustrate the technical solutions implemented in the present invention, the following is a brief explanation of the drawings required to be used in the examples of the present invention. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0067] Figure 1 A flow chart of the method for integrating the co-simulation model;

[0068] Figure 2 Integrate tool logic block diagram for joint simulation model;

[0069] Figure 3 Provide a flow chart for static semantic analysis of aircraft platform models;

[0070] Figure 4 It is the result diagram of aircraft platform model interface association mapping;

[0071] Figure 5 Deploy a cell structure diagram for a virtual environment. DETAILED DESCRIPTION

[0072] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0073] The features of various aspects of the embodiments of the present invention will be described in detail below. In the detailed description below, many specific details are proposed to fully understand the present invention. However, it is obvious to those skilled in the art that the present invention can also be implemented without these specific details. The following description of the embodiments is only for a better understanding of the present invention by illustrating examples of the present invention. The present invention is not limited to any specific settings and methods provided below, but covers all product structures, any improvements, replacements, etc. of methods covered without departing from the spirit of the present invention.

[0074] In the various drawings and the following description, well-known structures and technologies are not shown to avoid unnecessary ambiguity of the present invention. Figure 1 In the embodiment of the present invention, a typical heterogeneous aircraft physical model integration and a semi-physical aircraft model integration are taken as examples to further specifically illustrate the integration method of the present invention:

[0075] The existing entity model is an airplane entity model, and the model files include a .dll format dynamic library file (plane.dll), a .h format header file (plane.h), and a .csv format interface description file (plane.csv). The plane.dll file defines the simulation program of the airplane entity model, the plane.h file defines the airplane entity model external interface function declaration, and the plane.csv file defines the airplane entity model external interface function calling method description and parameter semantics interpretation, as well as the use of the airplane entity model.

[0076] The target simulation system for model integration is based on a custom model specification framework, which provides a set of inherent model call standard interfaces. The simulation system operating environment is based on the Linux operating system. According to the traditional model integration method, the aircraft physical model source code can only be recompiled into a .so format dynamic library file in the Linux environment, and the external interface functions exposed by the model in the source code are re-customized and developed to be completely consistent with the model standard interface of the model specification framework. This integration method requires the model to be operated at the source code level, which is not conducive to the decoupling of the model and the simulation system's behavior, and is not conducive to the maintenance of the model's intellectual property rights. In addition, when the model provider cannot provide the source code, it is basically impossible to achieve integrated application of this cross-platform heterogeneous model.

[0077] For external heterogeneous aircraft entity models, since there is no model description file that conforms to the XML format specification, it is necessary to subject the plane.csv interface description file to model parameter semantic analysis (intelligent semantic analysis + manual semantic analysis) and generate the aircraft entity model description file (plane.xml) through the simulation model integration tool. Secondly, after the model interface association mapping and model integration configuration steps, the aircraft entity model is deployed in the Windows virtualization environment for operation. The virtualization environment deployment unit adopts the model service bus technology, which can realize the command and data communication between the encapsulated integrated aircraft entity model and the simulation system. Finally, the virtualization environment of the encapsulated integrated aircraft entity model dll dynamic library file (planeCAE.dll) is run on the Linux system and provided to the simulation system for calling.

[0078] For the integration of semi-physical aircraft models, it is first necessary to build a digital proxy model of the semi-physical aircraft and then adopt the method mentioned in the present invention.

[0079] Building a digital proxy model is actually to realize the interaction between the semi-physical aircraft model and the simulation system, which can be realized by middleware. The digital proxy model needs to receive the data packets sent by the semi-physical model through the TCP / IP network protocol, and convert the received data packets into model object class attributes or interaction class instances defined in the simulation system through the middleware and forward them to the simulation system. In this way, the interconnection and intercommunication between the semi-physical aircraft model and the simulation system can be realized.

[0080] The digital agent model can receive management messages, behavioral task instructions, environmental situation information, etc. from the simulation system, and can publish the state information, target and interference information of the semi-physical aircraft model to the simulation system. The digital agent model has the functions of mission planning analysis, heterogeneous system data interaction and protocol conversion.

[0081] The mission planning analysis function analyzes the mission planning sent by the simulation system and converts it into an action plan file data format that can be used in semi-physical simulation.

[0082] The heterogeneous system data interaction function converts data between simulation systems and semi-physical models, and interacts with data, such as scenario data, situation data, detection data, etc.

[0083] The protocol conversion function converts the protocols of various control instructions of the simulation system to realize the effective management and control of the semi-physical model by the simulation system, such as simulation advancement, simulation service, model service, time management, event management, scheduling management, clock synchronization, etc.

[0084] Except for building a digital proxy model of a semi-physical aircraft in advance, the integration process of the semi-physical aircraft model and the component / physical model is the same. Therefore, the following takes the aircraft physical model integration as an example and describes in detail the joint simulation model integration tool of the present invention in combination with the accompanying drawings. Figure 2 As shown in the figure, the specific method flow for realizing the integration of aircraft entity model is as follows:

[0085] Step 1: Build a digital agency model

[0086] The aircraft physical model is not a semi-physical / system model, so there is no need to build a digital proxy model.

[0087] Import model files, including the aircraft entity model dll dynamic library file (plane.dll) and the csv format interface table file (plane.csv). The dll dynamic library file is mainly a binary program file, and the csv interface table describes the interface functions and parameters exposed by the aircraft entity model dll dynamic library file. In this step, there will be interface files of different formats depending on the modeling differences. It may be a user manual or operating instructions with a txt or docx file format suffix, an interface table with an xlsx or csv format suffix, a program header file with an h format suffix, or a language description file with an xml or json format suffix. For these types of imported model description files, users are first required to conduct preliminary manual analysis and understanding before they can be automatically analyzed and integrated through tools.

[0088] Step 2: Metamodel Data Analysis

[0089] Based on the basic understanding of the imported model in the above steps, the user fills in the basic information of the input aircraft entity model through the UI interface of the joint simulation model integration tool, including model ID, model name, model version number, model description, model discipline, model creator, model creation time, model modifier, model modification time, model application scope, usage restrictions, model confidentiality and other basic model information, and outputs it as a plane.xml file, fills in the node values ​​related to the metamodel data in the file, and also provides some node knowledge for the subsequent semantic analysis knowledge base.

[0090] Step 3: Conceptual model description and analysis

[0091] Based on the analysis and description of the aircraft entity model in the above steps, continue to analyze and describe the aircraft entity model interaction mode, interaction process, state transition, interaction-related entity types and event types from the semantic level. The user inputs the aircraft entity model interaction mode through the UI interface of the joint simulation model integration tool, defines the event type and state transition conditions of each model interaction mode, and defines the external model entity type associated with each event. Finally, the output is the plane.xml file, fill in the node values ​​related to the conceptual model description analysis in the file, and describe the interaction process inside or outside the model.

[0092] Step 4: Static semantic analysis

[0093] According to the above steps, the user inputs the aircraft entity model interface function and the input and output parameter ID, input and output parameter name, input and output parameter data type, input and output parameter unit, input and output parameter value upper limit, input and output parameter value lower limit, and input and output parameter row number and column number in the input and output data file through the UI interface of the joint simulation model integration tool. Finally, the output is the plane.xml file, and the node values ​​related to the static semantic analysis in the file are filled in.

[0094] This step can be associated with the semantic analysis knowledge base. The static semantic analysis algorithm flow chart is as follows: Figure 3 As shown. By constructing a semantic knowledge graph in the aviation field, the default values ​​of the information or the recommended values ​​of the intelligent algorithm are filled in, or reset according to these values, and finally an airplane entity model description xml file (plane.xml) is generated.

[0095] Step 5: Dynamic semantic analysis

[0096] According to the model interface information in the aircraft entity model description file plane.xml generated in the fourth step, extract and dynamically call the test aircraft entity model external interface function, input the initial parameters that conform to the model interface data format, test run the model file and view the model output data, and analyze and check the output data.

[0097] The input interface parameters and output interface parameters of the aircraft entity model are shown in Table 1 and Table 2 respectively.

[0098] Table 1 Aircraft platform model input interface function list

[0099]

[0100] Table 2 Aircraft platform model output interface function list

[0101]

[0102]

[0103] In particular, the input interface parameters of the aircraft entity model in Table 1 have the following requirements, and the custom data types in the aircraft entity model are as follows:

[0104] CSMapPoint defines the waypoint data structure that the aircraft model needs to input, including the three-dimensional position coordinates composed of longitude, latitude and altitude on the map. CSMapPath defines the cruise route data structure that the aircraft model needs to input, including a list of waypoints. CSMapRegion defines the cruise region data structure that the aircraft model needs to input, including the airspace surrounded by a series of cruise routes. PlaneManeuver defines the maneuvers that the aircraft model needs to input through maneuver control instructions. In addition to basic acceleration, deceleration, circling, turning, etc., there are also special somersaults, separation S, rolls, J-turns, leaf drifts, barrel rolls, etc. CommandParameters defines the additional parameters of the above maneuvers, such as acceleration, turning angle, circling altitude, etc. PlanePerformance defines the performance parameters that the aircraft model needs to input, including speed, altitude, range, flight time, load, etc.

[0105] The dynamic semantic analysis function is mainly responsible for online testing of the joint simulation model integration tool for the dynamic library file (plane.dll) after the aircraft entity model is imported and before integration. The input parameters of the aircraft entity model interface function are filled in through the joint simulation model integration tool interface, and the aircraft entity model dynamic library file is run to verify whether the result returned by the aircraft entity model interface function is semantically correct (the output parameter here is not necessarily the function return value. It is possible that the model interface file will indicate that a certain input parameter of the function is limited to out. In this case, the parameter is also an output parameter and also needs to be verified).

[0106] In this embodiment, the input parameters required by the entry point function SetInput (input interface function) of the aircraft entity model are listed in Table 1 and described in the interface description file plane.csv. The function of this entry point function is to initialize the aircraft entity model. When the SetInput function returns a value of -1, it means that the initialization has failed, and a return value of 0 means that the initialization has succeeded. When the performance parameters, waypoints and other information input by the SetInput function are reasonable, the return value is 0, otherwise the return value is -1, which verifies that the semantic description of the aircraft entity model is correct. Repeat this step to verify all entry point functions, and pass the dynamic semantic analysis if it conforms to the actual situation. On the contrary, if the actual situation of calling the aircraft entity model entry point function does not conform to the description in the interface description file plane.csv, the dynamic semantic analysis fails, the aircraft entity model is invalid, and the subsequent integration steps cannot be performed.

[0107] After the dynamic semantic analysis process, the aircraft entity model can realize maneuvering functions such as take-off, cruising, circling, and climbing after being given waypoint information and maneuvering instructions.

[0108] Step 6: Model interface association mapping

[0109] For the model interface function information in the plane.xml file of the aircraft entity model description, automatic association mapping or manual association mapping is performed to the standard interface in the target simulation system model specification framework. The model interface association mapping process may include logic such as adding, deleting, modifying, merging, decomposing, and re-arranging of interface functions. The list of standard interface function events in the target simulation system model specification framework is shown in Table 3.

[0110] Table 3 List of standard interfaces of simulation system model specification framework

[0111] method Semantics parameter Remark Creat Create an instance none For registering models Init initialization Json Initialize the model interface Subscribe Subscription Data Any<any<string,map>> Passing model data before single-step simulation Tick Stepping Double t, doubledelta_t Simulation time and step time Publish Publishing Data Any<any<string,map>> Publish data after single-step simulation Destroy Deregistering an instance none Model destruction

[0112] Among them, the possible function mapping strategies include "one-to-one", "one-to-many", "many-to-many", "many-to-one", "one-to-zero", and "zero-to-one". The specific mapping logics used based on these strategies include adding functions, deleting functions, renaming functions, merging functions, decomposing functions, and rearranging APIs.

[0113] For the aircraft entity model, the result of model interface association mapping is as follows: Figure 4 As shown. Finally, through the model interface association mapping, the aircraft entity model integration file planeCAE.dll is generated. The model library file contains the generated function mapping relationship. By calling planeCAE.dll, it can automatically associate and call the corresponding interface function in plane.dll. This is equivalent to encapsulating the original interface of plane.dll.

[0114] Step 7: Model integration configuration

[0115] Parameter configuration unit: Since the system environment windows (.dll dynamic library) for generating the aircraft entity model is not compatible with the system environment linux for running the model, it is necessary to set the model integration parameters for model integration configuration. In this embodiment, the integration target type is .so, the target interface format is xml, the target language is C++, the target bit number is 64 bits, the communication interface type is restful API, the target operating environment is linux, the integration model name is planeCAE, the virtualization technology is VM virtual machine, the host environment is ubuntu18.04, etc. Fill in the above information on the interface, and finally generate the integration configuration file planeConfigCAE.xml through the tool.

[0116] Virtualization environment deployment unit: Since it relies on the dll library files of the windows operating system, it cannot be run directly on the linux operating system, so a virtualization environment deployment unit is required. In this embodiment, the virtualization technology selects the VM virtual machine, and requires a functional communication module inside and outside the virtual machine. The functional communication module here is implemented by the model service bus technology (the aircraft entity model is a third-party black box model, and the model provider cannot provide the source code and cannot recompile).

[0117] In this embodiment, the overall structure of the virtualized operating environment deployment includes two parts: the Linux system model service module and the Windows virtual machine model service module (both implemented by the model service bus). The Linux system model service module is mainly responsible for data interaction with the simulation system and data communication with the virtual machine service module, including service call, service release, protocol adaptation, and protocol conversion. The Windows virtual machine model service module is mainly responsible for managing and calling the dll model dynamic library (planeCAE.dll) under the virtual machine environment, including service interface and DLL call program. First, the simulation system sends a call interface to execute the dll method message to the Linux system model service module. Secondly, the module parses the request parameters and converts them into a message in an agreed format and sends it to the Windows virtual machine model service module. Then, the Windows virtual machine model service module parses the request parameters, obtains information such as the dll file name, method name, method parameters, and loads the dll file to request related methods. Finally, the obtained dll dynamic library return value is sent back to the Windows virtual machine model service module, and then sent back to the Linux system model service module, and finally the return result is sent to the simulation system to complete the entire operation process.

[0118] The model service bus is mainly responsible for the data communication and service operation management between the host (Linux) simulation system and the dynamic library in the virtualization environment. The entire virtualization deployment and operation process is as follows: Figure 5 Show.

[0119] Step 8: Model integration into the database

[0120] In the tool interface, select the above-mentioned model integration file planeCAE.dll, integration configuration file planeConfigCAE.xml, and the polymorphic heterogeneous model original files plane.dll and plane.xml for packaging and integration, and finally send them to the model resource library for storage, or directly provide them to the simulation system for link calling. For component / physical models, package and integrate the original model library files and integration configuration files. For system / semi-physical models, package and integrate the original model proxy model and integration configuration files.

[0121] The present invention provides a model standardization integration method and tool for simulation personnel to use, targeting the four types of multi-morphological heterogeneous models of components, entities, semi-physical objects and systems that need to be integrated and called in joint simulation experiments. A large number of multi-morphological heterogeneous models that have been formed are integrated to enable various models to be integrated and called by the target simulation system. The method and tool help to expand the types of joint simulation experiment resources, introduce mature simulation models into the system, improve the resource construction efficiency of joint simulation experiments, free experimenters from programming and complex modeling, and devote more energy to simulation experiments. Support the interconnection, interoperability and reusability of models, maximize the value of model resources, expand the scope of model application, and avoid duplication.

[0122] Those skilled in the art may further appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented with electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel may use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

Claims

1. A joint simulation model integration method, characterized in that: The integration method is used to integrate and call four types of multi-modal heterogeneous models: component, entity, semi-physical and system; The integration steps are as follows: Step 1: semantic analysis of model parameters: semantic analysis is performed on the interface parameters of the polymorphic heterogeneous model. The process of semantic analysis is divided into the following stages: Metamodel data analysis, conceptual model description analysis, static semantic analysis, dynamic semantic analysis; S1. Metamodel data analysis includes: extracting and analyzing basic model information such as model ID, model name, model version number, model description, model discipline, model creator, model creation time, model modifier, model modification time, model application scope, usage restrictions, model confidentiality level, etc.; S2. Conceptual model description and analysis includes: extracting and analyzing model interaction patterns, describing the model's internal interaction process in the form of event sequences between entities; the elements involved in this process include model state transitions, entity types, and event types; S3, static semantic analysis includes: extracting the model interface function list and analyzing the input and output parameter ID, input and output parameter name, input and output parameter data type, input and output parameter unit, input and output parameter value upper limit, input and output parameter value lower limit and input and output parameter row number and column number in the input and output data file; S4, dynamic semantic analysis includes: obtaining the pointer of the model external interface function, then calling each model external interface function and checking the model output data, and analyzing and checking the output data. When calling the model external interface function, it is necessary to obtain the data type, unit and value range of the corresponding external interface input parameter based on the results of static semantic analysis, and generate the specific value of the input parameter of the model external interface function; Step 2: Model interface association mapping: Perform event association mapping on the interface functions of the multi-modal heterogeneous model. For the model interface function list in the semantic analysis result described in step 1, event association is mapped to the final required model specification framework; The association mapping process refers to: mapping the interface function list in the original model to the model interface function list after integration, wherein the model interface function list after integration refers to the standard interface required by the final model specification framework; Step 3: Model integration configuration: Integrate and configure the multi-modal heterogeneous models Configure the imported multi-modal heterogeneous model integration environment, including: configuring the model running environment, setting model integration parameters, generating model catalog files, generating model assembly files, and generating model scheduling files; Step 4: Model integration and storage: The multi-modal heterogeneous models are integrated and stored in the database; the model integration and storage means that the models are packaged and sent to the model resource library for unified management after integration; According to the integration configuration file described in step three, all relevant files of the multi-modal heterogeneous model are packaged and integrated as needed, and finally sent to the model resource library for storage, or directly provided to the simulation system for link calling.

2. The integration method according to claim 1, characterized in that: When the method integrates a semi-physical model or a system model, a digital proxy model needs to be created before step 1, and the digital proxy model is used to replace the semi-physical model or the system model.

3. The integration method according to claim 2, characterized in that: In step four, when the polymorphic heterogeneous model is a component or a physical model, the model original file and the integration configuration file are packaged and integrated; when the polymorphic heterogeneous model is a system or a semi-physical model, the digital proxy model and the integration configuration file are packaged and integrated.

4. The integration method according to claim 2, characterized in that: When the physical model is an aircraft entity model, the custom data types in the aircraft entity model are as follows: Define the waypoint data structure that the aircraft model needs to input, including the three-dimensional position coordinates composed of latitude, longitude and altitude on the map; define the cruise route data structure that the aircraft model needs to input, including a list of waypoints; Define the cruising area data structure that the aircraft model needs to input, including the airspace surrounded by the cruising route; Define the maneuvers that the aircraft model can perform by maneuvering control commands, including basic acceleration, deceleration, circling, and turning, as well as somersaults, separation S, rolls, J-turns, leaf flutters, and barrel rolls. Define additional parameters for the above maneuvers, such as acceleration, turning angle, circling altitude, etc. Define the performance parameters that the aircraft model needs to input, including speed, altitude, range, flight time, and load.

5. The integration method according to claim 1 or 2, characterized in that: The model parameter semantic analysis process also includes: intelligent semantic analysis and manual semantic analysis. The intelligent semantic analysis refers to: automatically analyzing the extracted semantic parameters by searching the semantic analysis knowledge base; The manual semantic analysis means that in the process of extracting and analyzing the semantic parameters and outputting the result data, manual participation is required to select or set the configuration parameters.

6. The integration method according to claim 1 or 2, characterized in that: The model interface association mapping in step 2 includes automatic association mapping or manual association mapping. The automatic association mapping refers to mapping the model interface function list to the standard interface of the required model specification framework one by one according to the automatic analysis results of the semantic analysis knowledge base. The manual association mapping refers to manually mapping the model interface function list to the standard interface of the required model specification framework one by one through the experience and knowledge of the integrator; it also includes manually correcting the results of automatic association mapping.

7. The integration method according to claim 1 or 2, characterized in that: The model specification framework described in step 2 includes FMI specifications, BOM specifications, SMP specifications, etc.

8. The integration method according to claim 1 or 2, characterized in that: The event association mapping strategies described in step 2 include "one-to-one", "one-to-many", "many-to-many", "many-to-one", "one-to-zero", and "zero-to-one"; the association mapping logics used based on the strategies include adding functions, deleting functions, renaming functions, merging functions, decomposing functions, and rearranging APIs.

9. The integration method according to claim 1 or 2, characterized in that: In step three, during the integration configuration of multi-modal heterogeneous models, it is necessary to set the model integration parameters, including the integration target type, target interface format, target language, target bit number, communication interface type, target operating environment, dependencies, integration model name, integration description, host environment, etc.

10. The integration method according to claim 1 or 2, characterized in that: The simulation system refers to a set of software and hardware systems that implements the model of the simulated system in accordance with the requirements of the simulation task to conduct simulation experiments, including simulation equipment, operators and models.

11. A joint simulation model integration tool, The integration method according to claim 1 or 2 is characterized in that: the integration tool includes a model parameter semantic analysis function module, a model interface association mapping function module, a model integration configuration function module and a model integration warehousing function module; The model parameter semantic analysis function module includes: a metamodel data analysis unit, a concept model description analysis unit, a static semantic analysis unit and a dynamic semantic analysis unit; The model interface association mapping function module maps the interface functions in the model to be integrated with the standard interface required by the final model specification framework, and automatically generates mapping code; the module inputs the model interface function list and the mapping connection relationship input on the interface, and outputs the relationship mapping code file; The model integration configuration function module is responsible for configuring the imported multi-modal heterogeneous model integration environment; the model integration configuration function module includes a virtualization environment deployment unit and a parameter configuration unit; The model integration warehousing function module will package and integrate all relevant files of the multi-modal heterogeneous model as needed, and finally send them to the model resource library for storage, or directly provide them to the target simulation system for link calling.

12. The integrated tool according to claim 11, characterized in that: The metamodel data analysis unit analyzes and describes the model from a pragmatic level. The unit converts the basic information of the model to be integrated into an analysis file in XML format and outputs it, providing partial node knowledge for the subsequent semantic analysis knowledge base; The conceptual model description and analysis unit analyzes and describes the model interaction mode, interaction process, state transition, interaction-related entity types and event types from a semantic level. The unit allows the user to input the model interaction mode through the UI interface, define the event type and state transition conditions of each model interaction mode, define the model entity type associated with each event, and output an analysis file in XML format, which can describe the interaction process inside or outside the model; The static semantic analysis unit enables the user to input relevant information of the model to be integrated through the UI interface, including the input and output parameter ID, input and output parameter name, input and output parameter data type, input and output parameter unit, input and output parameter value upper limit, input and output parameter value lower limit and input and output parameter row number and column number in the input and output data file, and outputs an analysis file in XML format to provide basic data for the implementation of subsequent model encapsulation of various functional units; The dynamic semantic analysis unit supports extracting and dynamically calling the model's external interface functions, inputs the model interface data format, inputs initial parameters, then runs the model file for trial operation, outputs the model's calling interface return data, and implements the model interface verification function.

13. The integrated tool according to claim 11, characterized in that: The parameter configuration unit enables the user to input parameters to configure the model operating environment and set model integration parameters through the UI interface, and outputs integrated configuration files such as model catalog files, model assembly files, and model scheduling files in XML format; The virtualized environment deployment unit deploys the model to be integrated in the virtual model running environment by encapsulating it, and then communicates with the target simulation system in the heterogeneous model running system environment through the model service bus; The virtual model operating environment includes multiple implementation modes such as virtual machines, containers, etc. The input of this unit is the model library to be packaged and the integration configuration file, and the output is the deployment directory and startup mode of the virtualized environment.

14. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the integrated tool as claimed in any one of claims 10 to 13 is implemented.

Images