Simulation method, system, electronic device and program product

Abstract

The application discloses an emulation method, system, electronic equipment and program product, and relates to the technical field of emulation. The method comprises the following steps: acquiring multi-modal data, wherein the multi-modal data comprises an emulation instruction, an emulation object design document and historical emulation data, the emulation instruction is a natural language instruction, the emulation object design document and the historical emulation data are acquired from a preset data source by a software robot, and the software robot is used for executing a task based on a preset rule; inputting the multi-modal data into a preset large language model to generate a script command, wherein the script command comprises all execution steps of an emulation task; calling a script execution interface of a preset emulation program, executing the script command, and obtaining an emulation result; and storing the emulation result into a target server by using the software robot. The application solves the technical problems that the existing emulation method is time-consuming and has a high technical threshold.

Description

Technical Field

[0001] This application relates to the field of simulation technology, and more specifically, to a simulation method, system, electronic device, and program product. Background Technology

[0002] Engineering simulation, especially CAE (Computer Aided Engineering) technology, has long relied on the deep involvement of senior engineers as one of the pillars of modern manufacturing. From understanding the physical problems to executing complex solutions, every step tests the engineer's expertise and patience, making it a knowledge-intensive and extremely time-consuming process. With the continuous increase in product complexity and iteration speed, the technical problems of traditional simulation methods—long processing times and high technical barriers—urgently need to be addressed. Summary of the Invention

[0003] This application provides a simulation method, system, electronic device, and program product to at least solve the technical problems of existing simulation methods being time-consuming and having high technical barriers.

[0004] According to one aspect of the embodiments of this application, a simulation method is provided, comprising: acquiring multimodal data, wherein the multimodal data includes simulation instructions, simulation object design documents, and historical simulation data, the simulation instructions being natural language instructions, the simulation object design documents and historical simulation data being acquired from a preset data source via a software robot, the software robot being used to execute tasks based on preset rules; inputting the multimodal data into a preset large language model to generate script commands, wherein the script commands contain all execution steps of the simulation task; calling the script execution interface of a preset simulation program to execute the script commands and obtain simulation results; and using the software robot to store the simulation results in a target server.

[0005] Optionally, the multimodal data is input into a preset large language model to generate script commands, including: inputting the multimodal data into the preset large language model to determine multiple simulation sub-tasks; performing parameter mapping based on the multiple simulation sub-tasks to obtain simulation configuration parameters; and generating script commands based on the simulation configuration parameters and the multiple simulation sub-tasks.

[0006] Optionally, the script execution interface of the preset simulation program is called to execute script commands and obtain simulation results, including: verifying the interface status of the script execution interface and obtaining verification results; in response to the verification result indicating that the interface status is a preset standard state, the computing cluster is called through the script execution interface to execute script commands and obtain simulation results; in response to the verification result indicating that the interface status is a preset abnormal state, the operation of the preset simulation program is controlled by a software robot under the guidance of script commands to obtain simulation results.

[0007] Optionally, the simulation method further includes: inputting simulation results and analysis instructions into a preset large language model to generate anomaly analysis results, wherein the analysis instructions are generated by a software robot; generating optimization suggestions based on the anomaly analysis results; and using the software robot to update the simulation object design document under the guidance of the optimization suggestions.

[0008] Optionally, a software robot is used to store the simulation results to the target server, including: using the software robot to extract key conclusions from the simulation results based on preset rules; and storing the key conclusions in the target directory of the product lifecycle management system, wherein the target directory is the directory corresponding to the target object, and the product lifecycle management system is deployed on the target server.

[0009] Optionally, the multimodal data is input into a preset large language model to generate script commands, including: inputting the multimodal data into the preset large language model, performing data type identification on the multimodal data, and obtaining identification results, wherein the identification results include multiple data types; calling data processing models corresponding to multiple data types to process the multimodal data and obtain data processing results, wherein the data type of the data processing results is text type; and inputting the data processing results into the large language model to generate script commands.

[0010] Optionally, the script execution interface of the preset simulation program is called to execute script commands and obtain simulation results, including: calling the script execution interface of the preset simulation program to execute script commands and obtain simulation calculation results; using a preset large language model to extract data from the simulation calculation results and obtain key calculation results; using a preset large language model to analyze the key calculation results and generate simulation results.

[0011] According to one aspect of the embodiments of this application, a simulation system is provided, comprising:

[0012] The acquisition module is used to acquire multimodal data, which includes simulation instructions, simulation object design documents, and historical simulation data. The simulation instructions are in natural language. The simulation object design documents and historical simulation data are acquired from a preset data source by a software robot, which executes tasks based on preset rules. The generation module is used to input the multimodal data into a preset large language model to generate script commands, which contain all execution steps of the simulation task. The invocation module is used to invoke the script execution interface of the preset simulation program, execute the script commands, and obtain the simulation results. The storage module is used to store the simulation results to the target server using the software robot.

[0013] According to one aspect of the embodiments of this application, an electronic device is provided, including a memory and a processor, wherein the memory stores a computer program and the processor is configured to run the computer program to perform the simulation method described in any of the embodiments.

[0014] According to one aspect of the embodiments of this application, a computer program product is provided, including a computer program that, when executed by a processor, implements the simulation method described in any of the embodiments.

[0015] According to one aspect of the embodiments of this application, a computer-readable storage medium is provided, the computer-readable storage medium including a stored program, wherein, when the program is executed, it controls the device where the storage medium is located to perform the simulation method described in any of the embodiments.

[0016] In this embodiment, multimodal data is acquired, including simulation instructions, simulation object design documents, and historical simulation data. The simulation instructions are natural language instructions. The simulation object design documents and historical simulation data are obtained from a preset data source through a software robot, which executes tasks based on preset rules. The multimodal data is input into a preset large language model to generate script commands, which contain all execution steps of the simulation task. The script execution interface of a preset simulation program is called to execute the script commands and obtain simulation results. The simulation results are then stored on a target server using the software robot. Thus, this embodiment employs integrated multimodal input processing, combining the efficient data acquisition capabilities of natural language processing and software robots, enabling seamless integration of simulation instructions, design documents, and historical data to form a deep understanding of the simulation task. The preset large language model can intelligently generate accurate script commands based on this multimodal data, significantly improving the speed of simulation setup and eliminating the need for simulation engineers to perform simulation analysis. Furthermore, the software robot not only participates in the initial data collection but also handles the automated management and storage of simulation results, ensuring the integrity and efficiency of the data chain. This fully automated and intelligent operation reduces the manual workload of engineers, shortens the simulation cycle, and achieves a high degree of automation from requirement understanding to result analysis of simulation tasks, thereby solving the technical problems of long time consumption and high technical threshold of existing simulation methods. Attached Figure Description

[0017] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments of this application and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0018] Figure 1 This is a hardware structure block diagram of an optional terminal device for implementing a simulation method according to an embodiment of this application;

[0019] Figure 2 This is a flowchart of a simulation method according to an embodiment of this application;

[0020] Figure 3 This is a flowchart illustrating an optional simulation method according to an embodiment of this application;

[0021] Figure 4 This is a structural block diagram of a simulation system according to an embodiment of this application. Detailed Implementation

[0022] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present application.

[0023] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0024] According to an embodiment of this application, an embodiment of a simulation method is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0025] Figure 1 This is a hardware structure block diagram of an optional terminal device for implementing a simulation method according to an embodiment of this application, such as... Figure 1As shown, the terminal device may include one or more processors 102 (processor 102 may include, but is not limited to, a microprocessor (MCU) or a field-programmable gate array (FPGA), etc.), a memory 104 for storing data, and a transmission device 106 for communication functions. In addition, it may also include: a display device 110, an input / output device 108, a Universal Serial Bus (USB) port (which may be included as one of the ports of a computer bus, not shown in the figure), a network interface (not shown in the figure), a power supply (not shown in the figure), and / or a camera (not shown in the figure). Those skilled in the art will understand that... Figure 1 The structure shown is for illustrative purposes only and does not limit the structure of the terminal device described above. For example, the terminal device may also include components that are more... Figure 1 The more or fewer components shown, or having the same Figure 1 The different configurations shown.

[0026] It should be noted that the aforementioned one or more processors 102 and / or other data processing circuits may be embodied, in whole or in part, as software, hardware, firmware, or any other combination thereof. Furthermore, the data processing circuitry may be a single, independent processing module, or may be integrated, in whole or in part, into any other element within the terminal device (or mobile device).

[0027] The memory 104 can be used to store software programs and modules of application software, such as the program instructions / data storage device corresponding to the simulation method in this embodiment. The processor 102 executes various functional applications and data processing by running the software programs and modules stored in the memory 104, thereby realizing the above-mentioned simulation method. The memory 104 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 104 may further include memory remotely located relative to the processor 102, and these remote memories can be connected to terminal devices via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0028] The transmission device 106 is used to receive or send data via a network. Specific examples of the network described above may include a wireless network provided by the communication provider of the terminal device. In one example, the transmission device 106 includes a Network Interface Controller (NIC), which can connect to other network devices via a base station to communicate with the Internet. In another example, the transmission device 106 may be a Radio Frequency (RF) module, used for wireless communication with the Internet.

[0029] Under the above operating environment, the embodiments of this application provide the following: Figure 2 The simulation method shown, Figure 2 This is a flowchart of a simulation method according to an embodiment of this application, such as... Figure 2 As shown, the method includes the following implementation steps:

[0030] Step S201: Obtain multimodal data, wherein the multimodal data includes simulation instructions, simulation object design documents and historical simulation data. The simulation instructions are natural language instructions. The simulation object design documents and historical simulation data are obtained from a preset data source through a software robot. The software robot is used to execute tasks based on preset rules.

[0031] Step S202: Input multimodal data into a preset large language model to generate script commands, wherein the script commands contain all execution steps of the simulation task;

[0032] Step S203: Call the script execution interface of the preset simulation program, execute the script commands, and obtain the simulation results;

[0033] Step S204: Use a software robot to store the simulation results to the target server.

[0034] Multimodal data is a collection of data containing multiple types of information. In the embodiments of this application, it mainly includes simulation instructions in natural language, design documents such as CAD drawings or PDF technical specifications, and historical simulation data from past simulation cases.

[0035] Simulation instructions are engineering simulation requirements given in natural language, such as "analyze the thermal deformation of this mechanical component at high temperature." These instructions are easy for engineers to understand and generate, but they need to be converted into computer-executable instructions before they can be used in the simulation process.

[0036] Simulation object design documents include, but are not limited to, CAD drawings, 3D model files, and technical specification PDF documents, which describe in detail the geometry, material properties, and structural configuration of the simulation object and provide necessary contextual information for simulation commands.

[0037] Historical simulation data is a record of real-world engineering simulation cases conducted previously, including simulation settings, operating parameters, result analysis, and conclusions. It forms the basis for large language model learning and prediction, and helps improve simulation accuracy and efficiency.

[0038] Software robot (RPA robot): A software tool based on RPA (Robotic Process Automation) capable of executing preset rules and processes, automatically transferring data, performing operations, and monitoring status between different software systems and platforms. In this embodiment, the software robot is used to automatically obtain the required simulation object design documents and historical simulation data from a product data management system or test database, and also to receive and store simulation results.

[0039] The pre-defined large language model is a trained artificial intelligence model capable of understanding and generating human language, and performing complex tasks such as code generation and strategy planning. In this embodiment, the large language model is used to parse multimodal data and convert it into script commands that guide the simulation process.

[0040] Script commands are a set of specific instructions generated from a pre-defined large language model, used to guide simulation software in executing specific simulation processes. These commands encompass various aspects of the simulation settings, such as material properties, boundary conditions, and solution methods, and are crucial for connecting multimodal data with simulation execution.

[0041] The script execution interface of the preset simulation program is an interface provided by the simulation program for executing external scripts, allowing users to operate and control the software by writing scripts to achieve automated simulation processes. Through this interface, the script commands in this embodiment are directly passed to the preset simulation program to automatically complete the simulation tasks.

[0042] The target server is a server or cloud storage space used to store simulation results data. Storing simulation results on the target server facilitates subsequent analysis, archiving, and cross-departmental sharing.

[0043] In this embodiment, the software robot automatically accesses a preset data source, such as a CAD system, product data management system, or test database, to extract simulation object design documents and historical simulation data. Simultaneously, engineers input simulation commands in natural language; together, these constitute multimodal data, providing comprehensive information for subsequent simulation processes.

[0044] Multimodal data is input into a pre-defined large language model for processing. The large language model first parses the simulation instructions to understand their physical background and technical requirements, then analyzes design documents and historical data to extract key information. Finally, the large language model generates specific script commands based on this information to guide the simulation program in performing specific simulation tasks.

[0045] The generated script commands are executed through the script execution interface of the preset simulation program. The simulation program automatically configures simulation parameters, creates models, generates meshes, sets boundary conditions, performs the solution, and generates simulation results according to the commands.

[0046] The software robot is responsible for uploading the simulation results data obtained from the aforementioned steps to the target server for archiving and management. This ensures the systematic nature and traceability of the results data, facilitating subsequent analysis and utilization by engineers.

[0047] In this embodiment, multimodal data is acquired, including simulation instructions, simulation object design documents, and historical simulation data. The simulation instructions are natural language instructions. The simulation object design documents and historical simulation data are obtained from a preset data source via a software robot, which executes tasks based on preset rules. The multimodal data is input into a preset large language model to generate script commands, which guide the simulation process of the simulation object. The script execution interface of a preset simulation program is called to execute the script commands and obtain simulation results. The simulation results are then stored on a target server using the software robot. Thus, this embodiment employs integrated multimodal input processing, combining the efficient data acquisition capabilities of natural language processing and software robots, enabling seamless integration of simulation instructions, design documents, and historical data to form a deep understanding of the simulation task. The preset large language model can intelligently generate precise script commands based on this multimodal data, significantly improving the speed of simulation setup and eliminating the need for simulation engineers to perform simulation analysis. Furthermore, the software robot not only participates in the initial data collection but also handles the automated management and storage of simulation results, ensuring the integrity and efficiency of the data chain. This fully automated and intelligent operation reduces the manual workload of engineers, shortens the simulation cycle, and achieves a high degree of automation from requirement understanding to result analysis of simulation tasks, thereby solving the technical problems of long time consumption and high technical threshold of existing simulation methods.

[0048] In this embodiment, a seamless transition from natural language simulation requirements to automated execution is achieved through efficient data acquisition by a software robot and precise instruction parsing and script generation by a pre-set large language model. This technical feature greatly simplifies the simulation process, reduces the operational burden on engineers, and improves the execution speed and accuracy of simulation tasks. The collaboration between the software robot and the large language model ensures the automatic storage and management of simulation results, strengthens data integration and sharing, and provides a solid data foundation for subsequent analysis and optimization. Overall, this embodiment significantly optimizes the workflow of engineering simulation, promotes the democratization and intelligentization of simulation technology, and provides strong support for accelerating innovation in product design and materials science.

[0049] Optionally, the multimodal data is input into a preset large language model to generate script commands, including: inputting the multimodal data into the preset large language model to determine multiple simulation sub-tasks; performing parameter mapping based on the multiple simulation sub-tasks to obtain simulation configuration parameters; and generating script commands based on the simulation configuration parameters and the multiple simulation sub-tasks.

[0050] Defining simulation subtasks involves breaking down complex simulation requirements into several specific, executable smaller tasks. For example, an overall CFD simulation command can be broken down into several specific subtasks, such as cleaning the geometric model, meshing, and setting fluid properties and boundary conditions, to facilitate more accurate parameter mapping and script command generation.

[0051] Parameter mapping refers to converting the physical concepts and parameters described in multimodal data into specific values ​​and settings that simulation software can recognize and execute. This process requires a pre-defined large language model that can understand and convert the physical properties and conditions described in natural language, such as mapping "gypsum board" to a material model, and mapping "dryer zone 2" to the boundary conditions of a specific region.

[0052] For example, the "dryer CFD simulation" is automatically mapped to the corresponding fan outlet air volume and temperature and set as the boundary condition. The "gypsum board" is automatically mapped to the corresponding material model and its parameters (such as thermal conductivity, specific heat capacity, etc.) in the material library.

[0053] Simulation configuration parameters are obtained through parameter mapping and are detailed parameters used to guide the preset simulation program in executing specific simulation subtasks. These parameters cover all necessary settings for the simulation process, including but not limited to material properties, mesh type, boundary conditions, solver options, etc., and are an important basis for generating script commands.

[0054] The pre-defined large language model can automatically identify and decompose multiple specific simulation sub-tasks based on the simulation requirements described in natural language. It leverages AI technology's ability to understand natural language, reducing the workload of manual task planning and improving the automation level of the simulation process.

[0055] The pre-defined large language model can intelligently map parameters and generate detailed script commands based on the identified multiple simulation sub-tasks. This achieves automatic conversion from natural language descriptions to simulation software operation instructions, greatly simplifying the simulation setup process, reducing the workload for engineers, and improving the efficiency and accuracy of simulation task execution.

[0056] In this embodiment, when multimodal data is input into a preset large language model, the model first parses the simulation requirements described in natural language and decomposes them into multiple specific simulation sub-tasks, such as thermodynamic analysis and structural mechanics analysis. Next, based on these sub-tasks, the model maps the physical concepts and values ​​from design documents and historical data related to the simulation object into configuration parameters recognizable by the simulation software. Finally, based on the mapped simulation configuration parameters and the determined simulation sub-tasks, the model generates directly executable script commands to guide the preset simulation program in completing the simulation process.

[0057] By inputting multimodal data into a pre-defined large language model, determining multiple simulation sub-tasks, mapping parameters to obtain simulation configuration parameters, and finally generating script commands, this embodiment of the application achieves an intelligent transformation from simulation requirements described in natural language to automated execution of engineering simulations. This technical feature simplifies the complexity of simulation setup, reduces reliance on senior engineers, and improves the execution efficiency and accuracy of simulation tasks. Simultaneously, it promotes the accumulation and reuse of simulation knowledge, providing efficient and intelligent simulation solutions for engineering design and materials science, significantly shortening the time cycle from requirement formulation to simulation result acquisition, and accelerating product development and optimization processes.

[0058] Optionally, the script execution interface of the preset simulation program is called to execute script commands and obtain simulation results, including: verifying the interface status of the script execution interface and obtaining verification results; in response to the verification result indicating that the interface status is a preset standard state, the computing cluster is called through the script execution interface to execute script commands and obtain simulation results; in response to the verification result indicating that the interface status is a preset abnormal state, the operation of the preset simulation program is controlled by a software robot under the guidance of script commands to obtain simulation results.

[0059] A computing cluster is a cluster system composed of multiple computers used to perform large-scale, computationally intensive tasks, such as complex engineering simulations, which can significantly improve simulation speed and processing power.

[0060] The verification result is obtained after checking the interface status of the script execution interface, and is used to determine whether the interface is in a state where it can normally receive and execute script commands.

[0061] The default standard state refers to the state in which the script execution interface is in normal working condition, and can directly receive and process script commands transmitted through the interface.

[0062] Preset abnormal states refer to non-standard working conditions encountered by the interface, such as network interruption, software update, insufficient resources, etc. These states may prevent the normal execution of script commands.

[0063] For the step of calling the script execution interface of the preset simulation program, executing script commands, and obtaining simulation results, the embodiments of this application ensure the smooth progress of the simulation task and improve the stability and reliability of the overall system by verifying the interface status and intelligently handling abnormal situations.

[0064] Interface status verification ensures that it is in a preset standard state, meaning it can normally receive and execute external scripts. This verification process avoids simulation failures caused by interface anomalies, thus improving the success rate of simulation tasks.

[0065] When the interface is in a preset abnormal state, a software robot is used to replace script execution and control the operation of the preset simulation program. The innovation of this feature is that, through the intervention of the software robot, it can complete the simulation task without relying on the script execution interface, thus enhancing the system's flexibility and fault tolerance.

[0066] In this embodiment of the invention, before calling the script execution interface, the status of the script execution interface of the preset simulation program is verified to ensure that it can normally receive and execute script commands. This verification may include checking network connectivity, software version, resource availability, etc. When the interface status verification result indicates that the interface is in a preset standard state, the computing cluster is directly called through the script execution interface to execute the script commands. The computing cluster utilizes its powerful computing capabilities to quickly complete the simulation task and generate simulation results. When the interface status verification result indicates that the interface is in a preset abnormal state, this embodiment of the application will not simply abandon execution, but will intelligently utilize a software robot to manually operate the user interface of the preset simulation program according to the instructions in the script commands to execute the corresponding simulation process. This process may include operations such as logging in to the software robot, opening files, setting parameters, and submitting calculations, ensuring that the simulation task is completed at the software interface operation level.

[0067] By intelligently verifying interface status and controlling software robots in abnormal situations, this application's embodiments achieve stable and efficient execution of natural language description simulation requirements under different conditions. This technical feature not only ensures a high success rate for simulation tasks but also enhances the system's adaptability and robustness in the face of unpredictable situations, improving the automation level and overall efficiency of engineering simulation. When the interface is normal, the efficient execution of the computing cluster significantly shortens the simulation time; when the interface is abnormal, the intervention of the software robot avoids task interruption, ensuring the continuity of the simulation process, thus providing strong and reliable technical support for product development and material optimization in the building materials industry.

[0068] Optionally, the simulation method further includes: inputting simulation results and analysis instructions into a preset large language model to generate anomaly analysis results, wherein the analysis instructions are generated by a software robot; generating optimization suggestions based on the anomaly analysis results; and using the software robot to update the simulation object design document under the guidance of the optimization suggestions.

[0069] Analysis instructions are a set of commands generated by a software robot to guide a pre-defined large language model in performing in-depth analysis and anomaly detection of simulation results. These instructions cover the checking and evaluation of key indicators of the simulation results, as well as possible abnormal scenarios.

[0070] Anomaly analysis results refer to simulation performance that does not meet expectations or standards after analyzing the simulation results using a pre-set large language model, including possible errors, defects, performance bottlenecks, etc.

[0071] Optimization suggestions refer to specific design modification suggestions generated from the pre-defined large language model based on the anomaly analysis results. These suggestions aim to improve product performance or solve problems found in simulation, such as modifying material properties, adjusting structural dimensions, and optimizing mesh generation.

[0072] In this embodiment, the preset large language model can not only handle simulation requirements, but also analyze anomalies in the simulation results. This feature introduces an AI-assisted anomaly detection mechanism, which can intelligently identify problems in the simulation results and provide precise directions for subsequent design optimization.

[0073] The pre-defined large language model automatically generates optimization suggestions based on anomaly analysis results. This feature enables the automatic transformation from simulation result analysis to design improvement, reducing the time for manual analysis and decision-making and accelerating the optimization process.

[0074] The software robot automatically updates the simulation object design document based on optimization suggestions. This feature utilizes software robot technology to automate design modifications, reducing reliance on engineers and ensuring the accuracy and timeliness of design updates.

[0075] In this embodiment, simulation results and analysis instructions generated by the software robot are input into a preset large language model. The model performs in-depth analysis of the simulation results based on the analysis instructions, identifying abnormal behaviors. For example, if the analysis instruction requires checking for "excessive thermal stress," the preset large language model will analyze the simulation results and identify areas where stress values ​​exceed the safe range. Based on the anomaly analysis results, the preset large language model generates optimization suggestions, such as "increasing the thickness of the gypsum board to reduce thermal stress." These optimization suggestions are based on the preset large language model's learning and understanding of engineering domain knowledge, aiming to solve problems and improve product performance. After receiving the optimization suggestions, the software robot automatically logs into the CAD system, finds the design document of the simulation object, and updates the design parameters according to the optimization suggestions, such as changing the gypsum board thickness from 9.5mm to 11.5mm, thereby completing the automatic update of the design document.

[0076] In this embodiment, by inputting simulation results and analysis commands into a preset large language model, anomaly analysis results and optimization suggestions are automatically generated. A software robot, guided by these optimization suggestions, updates the simulation object design document, achieving an automated closed loop from simulation result analysis to design optimization. This technical feature significantly improves the efficiency and quality of engineering design, reduces the subjectivity and error rate of manual analysis, and accelerates design iteration and product development cycles through AI-based intelligent decision-making and automated software robot operations. It can provide the building materials industry or other industries with efficient and intelligent design optimization tools.

[0077] Optionally, a software robot is used to store the simulation results to the target server, including: using the software robot to extract key conclusions from the simulation results based on preset rules; and storing the key conclusions in the target directory of the product lifecycle management system, wherein the target directory is the directory corresponding to the target object, and the product lifecycle management system is deployed on the target server.

[0078] Preset rules are a series of logical rules or algorithms pre-defined in a software robot to guide the robot in identifying and extracting key conclusions from simulation results. These rules can be based on engineering knowledge, statistical analysis, or specific threshold conditions to ensure that the extracted information is accurate and has engineering significance.

[0079] Key conclusions refer to information extracted from simulation results that has direct guiding significance for engineering design and optimization. These may include maximum stress values, fluid drag coefficients, and heat transfer efficiency, and are crucial data for product iteration and material selection.

[0080] A Product Lifecycle Management (PLM) system is a software system used to manage all information throughout a product's lifecycle, from conceptual design to end-of-life recycling, including design data, simulation results, test reports, and production information. In this embodiment, the PLM system is deployed on the target server to store and manage key conclusions.

[0081] The target server refers to the designated server or cloud storage space used to store simulation results and key conclusions. It is the infrastructure for the operation of the product lifecycle management system and ensures the security, accessibility and traceability of data.

[0082] A target catalog is a directory within a product lifecycle management system that organizes and stores key conclusions based on target objects (such as specific products, components, or design versions). This catalog structure facilitates engineers' quick location and retrieval of simulation results related to a specific design or product, promoting knowledge management and collaborative work.

[0083] In this embodiment, the software robot first identifies and extracts key conclusions from the simulation results according to preset rules. This process may involve statistical analysis of the simulation data, threshold comparison, and pattern recognition to ensure that the extracted information has engineering value and clarity. The extracted key conclusions are then stored by the software robot in the target directory of the Product Lifecycle Management System. The software robot logs into the PLM system, finds the directory corresponding to the target object, uploads and archives the key conclusions, and may also update design parameters, generate design change records, and associate with other design documents to form a complete engineering data chain.

[0084] By utilizing software robots to store key conclusions from simulation results in a product lifecycle management system on a target server, this embodiment of the application achieves automated management and efficient distribution of simulation data. This technical feature not only ensures the accurate extraction of key conclusions, avoiding the subjectivity and errors of manual identification, but also ensures the orderly storage and rapid retrieval of data within the product lifecycle management system. This promotes information sharing and collaborative work among design teams, providing timely and reliable decision-making support for design iterations and product optimization. Furthermore, the automated storage process enhances data security and traceability, reduces data management costs, and optimizes engineering design and R&D management processes. This aspect improves the efficiency of simulation data management and the collaborative nature of the design process, providing strong support for subsequent R&D decisions and product iterations.

[0085] Optionally, the multimodal data is input into a preset large language model to generate script commands, including: inputting the multimodal data into the preset large language model, performing data type identification on the multimodal data, and obtaining identification results, wherein the identification results include multiple data types; calling data processing models corresponding to multiple data types to process the multimodal data and obtain data processing results, wherein the data type of the data processing results is text type; and inputting the data processing results into the large language model to generate script commands.

[0086] Data processing models are automated processing tools built for specific data types, used to convert multimodal data into a unified text-based data type. For example, image processing models can parse CAD drawing images and extract geometric features and dimensional parameters; 3D model file processing models can read and convert model data to adapt to the input requirements of a pre-defined large language model.

[0087] The data processing results are uniformly formatted text data obtained after processing multimodal data, which is convenient for pre-defined large language models to understand and operate. These results include engineers' natural language instructions, key information extracted from design documents and 3D models, and textual descriptions of historical simulation data.

[0088] The pre-defined large language model can recognize and process inputs containing multiple data types. By calling the corresponding data processing model, it converts non-text data into text, achieving intelligent parsing of complex inputs. This original implementation brings a more comprehensive and accurate understanding of simulation requirements to the embodiments of this application, significantly improving the accuracy and flexibility of simulation settings.

[0089] By processing multimodal data into a unified text type, this constraint ensures that the pre-defined large language model can effectively handle all information without misunderstandings due to data type incompatibility or formatting issues. This improvement simplifies the data input process, enhances the processing capabilities of the large language model, and makes simulation setups more efficient and accurate.

[0090] After multimodal data is input into a pre-defined large language model, data type recognition is first performed to obtain recognition results. For example, the model recognizes different data types such as natural language instructions, PDF technical specifications, and 3D model files. Based on the recognition results, the corresponding data processing model is invoked to process the multimodal data, converting non-text data into text data to obtain data processing results. This processing may involve image recognition, 3D model data reading and format conversion, and textual description of historical data. The text data processing results are then input into the pre-defined large language model. Based on this information, the model understands the specific details of the simulation requirements and generates script commands. For example, based on the engineer's natural language instruction "Please analyze the thermodynamic properties of this gypsum board dryer," and combined with information extracted from design documents and historical data, the model generates a Python script for finite element analysis software to set material properties, boundary conditions, and solution parameters.

[0091] By inputting multimodal data into a pre-defined large language model, and after data type identification and processing, finally generating script commands, this embodiment of the application achieves intelligent conversion from natural language requirements to automated execution of simulation settings. This technical effect is independent of other parts of the solution, significantly improving the accuracy and efficiency of simulation requirement processing, enabling non-professionals to participate in the setup process of complex engineering simulations through intuitive natural language descriptions. Simultaneously, the introduction of the data processing model ensures the effective utilization of all types of information, enhances the understanding capability of the large language model and the adaptability of the simulation process, and provides strong technical support for the rapid iteration of engineering design and materials science.

[0092] Optionally, the script execution interface of the preset simulation program is called to execute script commands and obtain simulation results, including: calling the script execution interface of the preset simulation program to execute script commands and obtain simulation calculation results; using a preset large language model to extract data from the simulation calculation results and obtain key calculation results; using a preset large language model to analyze the key calculation results and generate simulation results.

[0093] After invoking the script execution interface of the preset simulation program, the program performs simulation calculations based on the received script commands, yielding simulation results including detailed physical performance analysis. The preset large-scale language model then extracts data from the received simulation results, leveraging its deep understanding of the engineering field to identify and extract key calculation results, such as the location of maximum stress, fluid drag coefficient, and temperature distribution maps. These results are presented in structured text format. Next, the preset large-scale language model further analyzes the key calculation results to generate simulation results. This process may include statistical analysis, trend prediction, and anomaly detection, ultimately summarizing the performance of the simulated object and providing optimization suggestions in natural language, offering clear and actionable guidance for engineering decisions.

[0094] By calling the script execution interface of a pre-defined simulation program to execute script commands and utilizing a pre-defined large language model to extract and analyze key calculation results, this embodiment of the application achieves an automated and intelligent process from simulation command execution to simulation result generation. This technical feature significantly improves the efficiency of simulation result acquisition and analysis quality, ensuring the timeliness and accuracy of engineering design decisions. The intelligent analysis capabilities of the pre-defined large language model enable simulation results to not only include intuitive physical performance indicators but also provide in-depth performance interpretations and optimization suggestions, greatly promoting the design team's utilization of simulation data and accelerating product development iterations.

[0095] Optionally, refer to Figure 3 The simulation method provided in this application is implemented as follows:

[0096] First, engineers submit natural language requirements (corresponding to simulation instructions). A large language model understands these requirements and decomposes them into tasks. Then, based on the simulation object design document, it generates a CAE (Computer-Aided Engineering) simulation script, calls the CAE software command line to execute simulation calculations, and generates corresponding result data and simulation reports. A software robot extracts data from the simulation results and writes it to a database or / and a PLM (Product Lifecycle Management) system. Simultaneously, the software robot automatically generates and sends email notifications to engineers based on their email addresses. The PLM system updates product design parameters based on the data written by the software robot, forming a closed-loop design simulation process.

[0097] It should be noted that when the script cannot be executed in the CAE software command line, the large language model generates software robot control instructions based on the script. These instructions then control the software robot to operate the CAE software's user interface to perform simulation calculations.

[0098] In this embodiment, a simulation system is also provided, which is used to implement the above embodiments and preferred embodiments; details already described will not be repeated. As used below, a "module" is a combination of software and / or hardware that can perform a predetermined function. Although the apparatus described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0099] Figure 4 This is a structural block diagram of a simulation system according to an embodiment of this application, such as... Figure 4 As shown, the system includes:

[0100] The acquisition module 401 is used to acquire multimodal data, which includes simulation instructions, simulation object design documents, and historical simulation data. The simulation instructions are natural language instructions. The simulation object design documents and historical simulation data are acquired from a preset data source by a software robot, which is used to execute tasks based on preset rules. The generation module 402 is used to input the multimodal data into a preset large language model to generate script commands, which contain all execution steps of the simulation task. The calling module 403 is used to call the script execution interface of the preset simulation program, execute the script commands, and obtain simulation results. The storage module 404 is used to store the simulation results to the target server using the software robot.

[0101] It should be noted that the above modules can be implemented by software or hardware. For the latter, they can be implemented in the following ways, but are not limited to: all the above modules are located in the same processor; or, the above modules are located in different processors in any combination.

[0102] According to another aspect of the embodiments of this application, an electronic device is also provided, including a memory and a processor, wherein the memory stores a computer program and the processor is configured to run the computer program to perform the simulation method described in any of the foregoing embodiments.

[0103] Optionally, in this embodiment, the processor can be configured to perform the following steps via a computer program:

[0104] Step S201: Obtain multimodal data, wherein the multimodal data includes simulation instructions, simulation object design documents and historical simulation data. The simulation instructions are natural language instructions. The simulation object design documents and historical simulation data are obtained from a preset data source through a software robot. The software robot is used to execute tasks based on preset rules.

[0105] Step S202: Input multimodal data into a preset large language model to generate script commands, wherein the script commands contain all execution steps of the simulation task;

[0106] Step S203: Call the script execution interface of the preset simulation program, execute the script commands, and obtain the simulation results;

[0107] Step S204: Use a software robot to store the simulation results to the target server.

[0108] Optionally, specific examples in this embodiment can refer to the examples described in the above embodiments and their optional implementations, and will not be repeated here.

[0109] According to another aspect of the embodiments of this application, a computer program product is also provided, including a computer program that, when executed by a processor, implements the simulation method described in any of the foregoing embodiments.

[0110] Optionally, in this embodiment, the computer program, when executed by the processor, performs the following steps:

[0111] Step S201: Obtain multimodal data, wherein the multimodal data includes simulation instructions, simulation object design documents and historical simulation data. The simulation instructions are natural language instructions. The simulation object design documents and historical simulation data are obtained from a preset data source through a software robot. The software robot is used to execute tasks based on preset rules.

[0112] Step S202: Input multimodal data into a preset large language model to generate script commands, wherein the script commands contain all execution steps of the simulation task;

[0113] Step S203: Call the script execution interface of the preset simulation program, execute the script commands, and obtain the simulation results;

[0114] Step S204: Use a software robot to store the simulation results to the target server.

[0115] According to another aspect of the embodiments of this application, a computer-readable storage medium is also provided, the storage medium including a stored program, wherein, when the program is running, it controls the device where the storage medium is located to execute the simulation method described in any of the foregoing embodiments.

[0116] Optionally, in embodiments of this application, the storage medium may be configured to store a computer program for performing the following steps:

[0117] Step S201: Obtain multimodal data, wherein the multimodal data includes simulation instructions, simulation object design documents and historical simulation data. The simulation instructions are natural language instructions. The simulation object design documents and historical simulation data are obtained from a preset data source through a software robot. The software robot is used to execute tasks based on preset rules.

[0118] Step S202: Input multimodal data into a preset large language model to generate script commands, wherein the script commands contain all execution steps of the simulation task;

[0119] Step S203: Call the script execution interface of the preset simulation program, execute the script commands, and obtain the simulation results;

[0120] Step S204: Use a software robot to store the simulation results to the target server.

[0121] Optionally, in this embodiment, the storage medium may include, but is not limited to, various media capable of storing computer programs, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0122] In this application, the descriptions of the various embodiments have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0123] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units can be a logical functional division, and in actual implementation, there may be other division methods. For instance, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling, direct coupling, or communication connection can be through some interfaces; the indirect coupling or communication connection between units or modules can be electrical or other forms.

[0124] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0125] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0126] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, ROM, RAM, portable hard drives, magnetic disks, or optical disks.

[0127] The above description is only a preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A simulation method, characterized in that, include: Acquire multimodal data, wherein the multimodal data includes simulation instructions, simulation object design documents, and historical simulation data, the simulation instructions are natural language instructions, the simulation object design documents and the historical simulation data are acquired from a preset data source by a software robot, and the software robot is used to execute tasks based on preset rules; The process of inputting the multimodal data into a preset large language model to generate script commands includes: inputting the multimodal data into the preset large language model to determine multiple simulation sub-tasks; mapping physical concepts and values ​​in the simulation object design document and historical simulation data related to the simulation object to obtain simulation configuration parameters recognizable by the simulation software based on the multiple simulation sub-tasks; and generating the script commands based on the simulation configuration parameters and the multiple simulation sub-tasks, wherein the script commands contain all execution steps of the simulation tasks. The script execution interface of the preset simulation program is called to execute the script commands and obtain the simulation results; The simulation results are stored on the target server using the software robot. The method further includes: inputting the simulation results and analysis instructions into the preset large language model to generate anomaly analysis results, wherein the analysis instructions are generated by the software robot; generating design optimization suggestions based on the anomaly analysis results; and using the software robot to update the simulation object design document under the guidance of the design optimization suggestions.

2. The simulation method according to claim 1, characterized in that, The step of calling the script execution interface of the preset simulation program, executing the script commands, and obtaining simulation results includes: The interface status of the script execution interface is verified to obtain the verification result; In response to the verification result indicating that the interface state is a preset standard state, the script execution interface is used to call the computing cluster to execute the script command and obtain the simulation result; In response to the verification result indicating that the interface state is a preset abnormal state, the software robot, under the guidance of the script command, controls the execution of the preset simulation program to obtain the simulation result.

3. The simulation method according to claim 1, characterized in that, The step of using the software robot to store the simulation results to the target server includes: Using a software robot, key conclusions are extracted from the simulation results based on preset rules; The key conclusions are stored in the target directory of the product lifecycle management system, wherein the target directory is the directory corresponding to the target object, and the product lifecycle management system is deployed on the target server.

4. The simulation method according to claim 1, characterized in that, The step of inputting the multimodal data into a preset large language model and generating script commands includes: The multimodal data is input into a preset large language model, and the data type of the multimodal data is identified to obtain the identification result, wherein the identification result includes multiple data types; The data processing model corresponding to the various data types is invoked to process the multimodal data and obtain the data processing result, wherein the data type of the data processing result is text. The data processing results are input into the large language model to generate the script commands.

5. The simulation method according to claim 1, characterized in that, The step of calling the script execution interface of the preset simulation program, executing the script commands, and obtaining simulation results includes: The script execution interface of the preset simulation program is called to execute the script commands and obtain the simulation calculation results; The key calculation results are obtained by extracting data from the simulation calculation results using the preset large language model; The key calculation results are analyzed using the preset large language model to generate the simulation results.

6. A simulation system, characterized in that, include: An acquisition module is used to acquire multimodal data, wherein the multimodal data includes simulation instructions, simulation object design documents, and historical simulation data. The simulation instructions are natural language instructions. The simulation object design documents and the historical simulation data are acquired from a preset data source by a software robot. The software robot is used to execute tasks based on preset rules. A generation module is used to input the multimodal data into a preset large language model and generate script commands, including: inputting the multimodal data into the preset large language model to determine multiple simulation sub-tasks; mapping physical concepts and values ​​in the simulation object design document and historical simulation data related to the simulation object to obtain simulation configuration parameters recognizable by the simulation software based on the multiple simulation sub-tasks; and generating the script commands based on the simulation configuration parameters and the multiple simulation sub-tasks, wherein the script commands contain all execution steps of the simulation tasks; The calling module is used to call the script execution interface of the preset simulation program, execute the script commands, and obtain simulation results; A storage module is used to store the simulation results to the target server using the software robot; The generation module is also used to input the simulation results and analysis instructions into the preset large language model to generate anomaly analysis results, wherein the analysis instructions are generated by the software robot; generate design optimization suggestions based on the anomaly analysis results; and update the simulation object design document under the guidance of the design optimization suggestions using the software robot.

7. An electronic device, characterized in that, It includes a memory and a processor, wherein the memory stores a computer program and the processor is configured to run the computer program to perform the simulation method of any one of claims 1 to 5.

8. A computer program product, characterized in that, It includes a computer program that, when executed by a processor, implements the simulation method of any one of claims 1 to 5.

Images