Digital twin actuator model generation method
By constructing and calibrating actuator models, a digital twin model with real-time operation capability is generated, solving the problem of modeling accuracy under the influence of nonlinear factors, realizing efficient simulation testing and virtual-real fusion, and reducing aircraft design costs and cycle time.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- XIAN AIRCRAFT DESIGN INST OF AVIATION IND OF CHINA
- Filing Date
- 2025-12-31
- Publication Date
- 2026-04-28
AI Technical Summary
Existing technologies make it difficult to accurately establish digital twin models that take into account the influence of nonlinear factors during actuator operation, resulting in insufficient modeling accuracy. Furthermore, the reliance on hardware verification leads to long development cycles and high costs.
By constructing an actuator model, conducting closed-loop operation test data testing, building a simulation environment, performing simulation testing and calibration, adjusting key parameters using parameter estimation tools, performing order reduction or proxy processing, and generating a digital twin actuator model with real-time operation capabilities.
It enables highly realistic hardware-in-the-loop and software-in-the-loop simulation tests, supports virtual-real fusion and virtual testing of aircraft systems, reduces design costs and cycle time, and improves test coverage and reliability.
Smart Images
Figure CN121936136A_ABST
Abstract
Description
Technical Field
[0001] This application belongs to the field of aircraft flight control test technology, specifically relating to a method for generating digital twin actuator models. Background Technology
[0002] The aircraft flight control system is a critical system for aircraft safety and high reliability, and it is also one of the most complex control systems. In the conventional aircraft design process, the experimental verification work of the flight control system design process basically depends on the hardware environment. The design, production, and verification often require a long development cycle. If there are new design technologies, they need to undergo high-cost and long-term verification. Twin models can effectively solve these problems.
[0003] In the process of digital equipment development, establishing highly realistic digital twin models is a key technology supporting MIL (Model in the Loop SIL), SIL (Software in the Loop), and HIL (Hardware in the Loop) system simulation and verification. The increasing realism and operational condition coverage of digital twin models mean that aircraft design iterations are no longer entirely dependent on the hardware environment, thus alleviating some of the constraints of hardware development cycles. Twin model generation can effectively reduce system verification costs, accelerate aircraft design iteration cycles, and improve test coverage. Especially for fault simulation that is difficult to achieve with aircraft system hardware, and for flight test evaluation of high-risk faults, twin models applicable to different scenarios can effectively reduce the economic cost of aircraft design and improve aircraft reliability.
[0004] Actuators are a crucial component of flight control system testing. They are complex, containing numerous nonlinear elements, and modeling them often struggles to fully account for the nonlinear factors affecting actuator operation. However, nonlinear factors are a major contributor to modeling accuracy. Furthermore, actuator operation is influenced by factors such as energy system flow, pressure, and temperature, making it difficult to accurately establish digital twin models of actuators in practice. Therefore, this application is submitted. Summary of the Invention
[0005] The purpose of this application is to provide a method for generating digital twin actuator models, which can generate digital twin models of actuators, conduct hardware-in-the-loop and software-in-the-loop system simulation tests, support equipment digitization, and realize the implementation of virtual and real fusion tests of aircraft systems as well as virtual tests.
[0006] The technical solution of this application is:
[0007] A method for generating a digital twin actuator model includes:
[0008] Step 1: Model Preparation: Construct an actuator model according to its mechanism and physical composition;
[0009] Step 2, Test Data Preparation: Using the actual actuator, perform closed-loop operation of the actuator and test the test data;
[0010] Step 3: Set up the simulation environment: Set up a simulation environment for the actuator to run according to the actuator's functional and performance testing requirements;
[0011] Step 4: Simulation Test: Conduct actuator simulation tests under the same working conditions as the experimental data test, and record the test data;
[0012] Step 5: Actuator Model Calibration: Calibrate the actuator model;
[0013] Step 6: Actuator Model Processing: Process the actuator model to enable it to run in real time;
[0014] Step 7: Actuator Model Export: Use software tools to export the actuator model as an FMU or other model format that can run on a real-time system.
[0015] According to at least one embodiment of this application, in the above-described digital twin actuator model generation method, step one involves constructing an actuator model, which includes a servo control valve, a logic control valve, a valve core position sensor, an actuator displacement sensor, and an actuator cylinder.
[0016] According to at least one embodiment of this application, in the above-described digital twin actuator model generation method, in step two, the actuator is used to perform closed-loop operation under given pressure, flow rate, and no-load and loaded conditions to test the test data of stroke, accuracy, maximum output force, hysteresis, threshold, speed, time-domain dynamic response, and bandwidth.
[0017] According to at least one embodiment of this application, in the above-described digital twin actuator model generation method, in step two, the experimental data sampling rate is 1 KSa / s.
[0018] According to at least one embodiment of this application, in the above-described digital twin actuator model generation method, step three involves building a simulation environment for actuator operation, including closed-loop control parameters, energy system, and input signal commands.
[0019] According to at least one embodiment of this application, in the above-described digital twin actuator model generation method, in step five, adjustment parameters are selected and set, and parameters that can directly affect the causal relationship between system input and output are adjusted by means of parameter estimation software tools, and are used as objects of parameter estimation. The function and performance of the system are changed by adjusting such parameters.
[0020] The selected adjustment parameter is observable, and its observation or measurement can be obtained from the available data.
[0021] According to at least one embodiment of this application, in the above-described digital twin actuator model generation method, in step five, the valve core position of the servo valve is set as an adjustment parameter, and the flow rate is parametrically analyzed.
[0022] According to at least one embodiment of this application, in the above-described digital twin actuator model generation method, in step five, the actuator model calibration result is evaluated, the calibration error under different working conditions is calculated, and the calculation is evaluated to see if it meets the simulation requirements. If not, the calibration is performed again.
[0023] According to at least one embodiment of this application, in the above-described digital twin actuator model generation method, in step five, the actuator model is downgraded using software tools to enable the model to run in real time.
[0024] According to at least one embodiment of this application, in the above-described digital twin actuator model generation method, in step five, the actuator model is processed by a proxy using software tools, so that the model has the ability to run in real time.
[0025] This application has at least the following beneficial technical effects:
[0026] A method for generating digital twin actuator models is provided, which can generate digital twin models of actuators, conduct high-fidelity and real-time simulation tests of hardware-in-the-loop and software-in-the-loop systems, support equipment digitization, and realize the conduct of virtual and real fusion tests of aircraft systems. Attached Figure Description
[0027] Figure 1 This is a schematic diagram of the actuator provided in the embodiments of this application;
[0028] Figure 2 This is a schematic diagram of setting structural variables and boundary conditions provided in the embodiments of this application.
[0029] To better illustrate this embodiment, some content in the accompanying drawings may be omitted, enlarged, or reduced. They are for illustrative purposes only and should not be construed as limiting the scope of this application. Detailed Implementation
[0030] To make the technical solution and advantages of this application clearer, the technical solution of this application will be described in a clearer and more complete manner below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are only some embodiments of this application, and are only used to explain this application, not to limit this application. It should be noted that, for ease of description, only the parts related to this application are shown in the accompanying drawings, and other related parts can be referred to the general design.
[0031] Furthermore, unless otherwise defined, the technical or scientific terms used in this application description shall have the ordinary meaning understood by one of ordinary skill in the art to which this application pertains. The word "comprising" as used in this application description indicates that the concept preceding the word encompasses the concepts listed following the word and their equivalents, without excluding other related concepts.
[0032] A method for generating digital twin actuator models.
[0033] Step 1: Model preparation: Construct an actuator model according to its mechanism and physical composition.
[0034] Actuator models typically include servo control valves, logic control valves, valve spool position sensors, actuator displacement sensors, actuator cylinders, etc. Figure 1 As shown.
[0035] Step 2: Test Data Preparation: Using the actual actuator, conduct closed-loop operation of the actuator and test the test data.
[0036] Using the actual actuator, closed-loop operation is performed under given pressure, flow rate, and different input signals, typically including no-load and loaded conditions, to test data such as stroke, accuracy, maximum output force, hysteresis, threshold, speed, time-domain dynamic response, and bandwidth.
[0037] The recommended sampling rate for experimental data is 1 kSa / s to ensure complete collection of data information for the corresponding feature points in the system. The operating condition coverage depends on the operating conditions of the twin model.
[0038] Step 3: Set up the simulation environment: According to the actuator's functional and performance testing requirements, set up a simulation environment for the actuator's operation, including closed-loop control parameters, energy system, input signal commands, etc.
[0039] Step 4: Simulation Test: Conduct actuator simulation tests under the same working conditions as the experimental data test, record the test data, and use it for actuator model calibration.
[0040] Step 5: Actuator Model Calibration: Calibrate the actuator model.
[0041] Selecting and setting adjustment parameters, and using parameter estimation software tools, involves adjusting parameters that directly affect the causal relationship between system input and output. These parameters are the main objects of parameter estimation. By adjusting these parameters, the function and performance of the system can be changed. On the one hand, it is necessary to ensure that the selected parameters are observable, that is, that observations or measurements of these parameters can be obtained from available data. On the other hand, it is necessary to evaluate the impact of the parameters on the model output. The valve core position of the servo valve is set as the adjustment parameter, and the flow rate is parametrically analyzed.
[0042] Establish and configure the experiment, determine the operating conditions of the model, and ensure that the model operating conditions meet the requirements of the experimental task. Associate the output of the simulation model with the measured data. The optimization algorithm needs to define an objective function to evaluate the merits of parameter combinations. Associating the simulation model output with the measured data provides a baseline value for the objective function, allowing the optimization process to be adjusted and verified based on actual measurement data, thus ensuring the final parameter estimation results are practically feasible in the actual system.
[0043] Set the structural variables and boundary conditions as follows: Figure 2 Fixed parameters are optional. Once set, they will not be changed during the optimization of related parameters. Fixed parameters are needed to observe changes in output by changing certain parameters in the actuator model during actuator model calibration.
[0044] Create an experiment, perform parameter simulation and residual calculation, and select a parameter optimization algorithm, including Bobyqa, Praxis, Cobyla, particle swarm optimization, genetic algorithm, multi-objective particle swarm optimization, and multi-objective genetic algorithm; select a residual function, including sum of squares residual (…). ), where x i For the measurement data points, y i For simulation data points, x max The maximum value among the measured data points is N, where N is the number of measured data points, and E is E. error (residuals), absolute values and residuals ( Normalized sum of squares residuals Simultaneously, multiple experiments are created, simulation experiments are conducted, and errors are estimated.
[0045] For actuators with complex operating conditions, digital twin models are typically only effective within a limited range.
[0046] The calibration results of the actuator model are evaluated, the calibration error under different working conditions is calculated, and the calculation is evaluated to see if it meets the simulation requirements. If not, the calibration is performed again.
[0047] Step 6: Actuator Model Processing: Clarify the relationships between model inputs, outputs, and variables, and process the actuator model to enable it to run in real time.
[0048] By using software tools to reduce the order of actuator models or perform proxy processing, the models can be made capable of real-time operation.
[0049] Step 7: Actuator Model Export: Use software tools to export the actuator model as an FMU or other model format that can run on a real-time system.
[0050] The digital twin actuator model generation method disclosed in the above embodiments provides a complete model generation process and methodology, including test condition design, model correction methods, and error estimation methods during model generation. Among numerous parameters affecting actuator operation, the valve core position parameter is selected as the primary correction parameter. While the method does not have a very high coverage of test conditions, the dynamic and static test errors of the model are guaranteed to be within ±8%. This method provides a technical approach for future equipment digitization and digital-based actuator virtual verification and virtual-real fusion testing, possessing high widespread application value and significantly promoting the development of digital twin technology.
[0051] The technical solution of this application has been described in conjunction with the preferred embodiments shown in the accompanying drawings. Those skilled in the art should understand that the scope of protection of this application is obviously not limited to these specific embodiments. Without departing from the principles of this application, those skilled in the art can make equivalent changes or substitutions to the relevant technical features, and the technical solutions after these changes or substitutions will all fall within the scope of protection of this application.
Claims
1. A method for generating a digital twin actuator model, characterized in that, include: Step 1: Model Preparation: Construct an actuator model according to its mechanism and physical composition; Step 2, Test Data Preparation: Using the actual actuator, perform closed-loop operation of the actuator and test the test data; Step 3: Set up the simulation environment: Set up a simulation environment for the actuator to run according to the actuator's functional and performance testing requirements; Step 4: Simulation Test: Conduct actuator simulation tests under the same working conditions as the experimental data test, and record the test data; Step 5: Actuator Model Calibration: Calibrate the actuator model; Step 6: Actuator Model Processing: Process the actuator model to enable it to run in real time; Step 7: Actuator Model Export: Use software tools to export the actuator model as an FMU or other model format that can run on a real-time system.
2. The method for generating a digital twin actuator model according to claim 1, characterized in that, In step one, the actuator model is constructed, including a servo control valve, a logic control valve, a valve core position sensor, an actuator displacement sensor, and an actuator cylinder.
3. The method for generating a digital twin actuator model according to claim 2, characterized in that, In step two, the actuator is tested under given pressure, flow rate, and no-load and loaded conditions in a closed loop to measure the test data of stroke, accuracy, maximum output force, hysteresis, threshold, speed, time-domain dynamic response, and bandwidth.
4. The method for generating a digital twin actuator model according to claim 3, characterized in that, In step two, the experimental data sampling rate is 1 kSa / s.
5. The method for generating a digital twin actuator model according to claim 4, characterized in that, In step three, a simulation environment for the actuator operation is built, including closed-loop control parameters, energy system, and input signal commands.
6. The method for generating a digital twin actuator model according to claim 5, characterized in that, In step five, select and set adjustment parameters. With the help of parameter estimation software tools, adjust the parameters that can directly affect the causal relationship between the system input and output, and use them as the objects of parameter estimation. By adjusting these parameters, the function and performance of the system can be changed. The selected adjustment parameter is observable, and its observation or measurement can be obtained from the available data.
7. The method for generating a digital twin actuator model according to claim 6, characterized in that, In step five, the valve core position of the servo valve is set as the adjustment parameter, and the flow rate is parametrically analyzed.
8. The method for generating a digital twin actuator model according to claim 7, characterized in that, In step five, the calibration results of the actuator model are evaluated, the calibration error under different working conditions is calculated, and the calculation is evaluated to see if it meets the simulation requirements. If not, the calibration is repeated.
9. The method for generating a digital twin actuator model according to claim 8, characterized in that, In step five, software tools are used to reduce the order of the actuator model, enabling the model to run in real time.
10. The method for generating a digital twin actuator model according to claim 9, characterized in that, In step five, software tools are used to perform proxy processing on the actuator model, enabling the model to run in real time.