An energy-scaling-based digital twin parameter tuning method and system
Patent Information
- Application Number
- CN202610779866.X
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-06-02
- Publication Date
- 2026-09-25
AI Technical Summary
[0002]现有数字孪生参数调优技术未围绕能量传输特性构建专属的数字孪生能量流动拓扑，无法完成能量节点与能量连接边的梳理标定，也无法精准采集物理实体在标准工况下稳定运行的能量守恒基准偏差，导致参数调优过程缺乏统一的能量参照标准，参数调整逻辑与物理实体实际能量流动规律完全脱节，数字孪生仿真能量状态与物理实体真实能量状态难以匹配，参数调优缺少客观稳定的基准支撑
1.本发明通过构建数字孪生模型专属的能量流动拓扑结构，完整梳理能量节点与能量连接边的关联指向，同步采集物理实体在标准工况下稳定运行的能量守恒基准偏差，为数字孪生参数调优搭建起统一且精准的能量参照体系，使参数调整全程贴合物理实体真实的能量传输与转化规律，大幅提升数字孪生仿真能量状态与物理实体实际能量状态的契合度，让参数调优拥有稳定客观的基准支撑，保障调优依据的准确性与可靠性。
Smart Images
Figure CN122818612A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of parameter optimization technology, and in particular to a method and system for tuning digital twin parameters based on energy-scaling. Background Technology
[0002] Existing digital twin parameter tuning technologies do not construct a dedicated digital twin energy flow topology based on energy transmission characteristics. They cannot complete the sorting and calibration of energy nodes and energy connection edges, nor can they accurately collect the energy conservation benchmark deviation of physical entities under standard operating conditions. As a result, the parameter tuning process lacks a unified energy reference standard, the parameter adjustment logic is completely disconnected from the actual energy flow law of the physical entity, the simulated energy state of the digital twin is difficult to match with the real energy state of the physical entity, and the parameter tuning lacks objective and stable benchmark support.
[0003] Existing parameter tuning methods cannot analyze the energy equipotential surface distribution of the digital twin's output, cannot effectively identify local minima of energy in the parameter space, and cannot determine the precise direction of energy gradient flow. Parameter exploration lacks scientific guidance, and there is no verification mechanism for the ratio of the exploration energy deviation to the baseline deviation. Nor has a parameter scaling iteration logic with matching tolerance range been formed. The deviation ratio is very likely to exceed the preset tolerance range. The parameter iteration convergence speed is slow and the tuning accuracy is insufficient. It is impossible to output a stable and reliable optimal parameter set, which is difficult to meet the actual application requirements of efficient and accurate parameter tuning for digital twins. Summary of the Invention
[0004] This invention provides a method and system for tuning digital twin parameters based on energy scaling to solve the problems mentioned in the background art.
[0005] To achieve the above objectives, the present invention provides a digital twin parameter tuning method based on energy-scaling, comprising: Construct the energy flow topology of the digital twin model and collect the energy conservation benchmark deviation of the physical entity under standard operating conditions; The initial parameter values are substituted into the digital twin model for operation, and local minima of the energy equipotential surface distribution map during operation are identified to determine the direction of energy gradient flow. The parameter movement direction vector is superimposed with the current parameter value to obtain the trial parameter value, which is then substituted into the digital twin model to obtain the total trial energy. The difference between the total energy of the trial and the total energy of the physical entity's input is taken as the trial energy deviation. When the ratio of the trial energy deviation to the energy conservation benchmark deviation exceeds the tolerance range, the initial parameter value is scaled and the parameter value is output as the optimal parameter set when the ratio is stabilized within the tolerance range.
[0006] In a preferred embodiment, the energy flow topology for constructing the digital twin model includes: Traverse all adjustable parameters in the digital twin model and mark the function controlled by each adjustable parameter as an energy node; Extract the data transmission paths between each energy node, mark each data transmission path as an energy connection edge, and determine the direction of each energy connection edge by adjusting the positive and negative correlation between the direction and the change in output energy according to the parameters. The energy flow topology is formed by combining the directions of all energy nodes and energy connection edges.
[0007] In a preferred embodiment, the energy conservation benchmark deviation of the acquired physical entity under standard operating conditions includes: The difference between the total input energy and the total output energy of a physical entity when it is running stably under standard operating conditions is defined as the energy conservation benchmark deviation.
[0008] In a preferred embodiment, the step of substituting the initial parameter values into the digital twin model for operation and identifying local minima in the energy equipotential surface distribution map during operation includes: Map the energy values of each energy node output after running the digital twin model to the corresponding positions in the parameter space, and calculate the energy difference between each parameter position and its neighboring parameter positions; The locations of parameters where all energy differences are positive and the absolute value exceeds a preset threshold are marked as local minima.
[0009] In a preferred embodiment, determining the energy gradient flow direction includes: Starting from the current parameter position, calculate the direction vector pointing from the current parameter position to each local minimum region, and select the direction with the fastest energy decrease rate as the energy gradient flow direction. The energy decrease rate is determined by dividing the energy difference between the current parameter position and the local minimum region by the parameter spatial distance.
[0010] In a preferred embodiment, the step of superimposing the parameter movement direction vector with the current parameter value to obtain the trial parameter value, and substituting it into the digital twin model to obtain the total trial energy, includes: Starting from the current parameter value, move one step distance along the direction indicated by the parameter movement direction vector, and take the parameter space position reached after the movement as the trial parameter value. Write the trial parameter values into the corresponding parameter interface of the digital twin model to drive the digital twin model to complete a full simulation; Extract the output values of each energy node from the simulation results and sum them up. Use the summation result as the total trial energy.
[0011] In a preferred embodiment, the difference between the total test energy and the total input energy of the physical entity is used as the test energy deviation. When the ratio of the test energy deviation to the energy conservation reference deviation exceeds the tolerance range, the initial parameter value is scaled, including: Calculate the ratio of the experimental energy deviation to the energy conservation baseline deviation, and compare the ratio with the preset tolerance range; If the ratio is greater than the upper limit of the tolerance range or less than the lower limit of the tolerance range, it is judged as exceeding the tolerance range. The scaling factor is determined according to the degree of deviation between the ratio and the tolerance range boundary. The current parameter value is multiplied by the scaling factor to obtain the adjusted parameter value.
[0012] In a preferred embodiment, scaling the initial parameter value further includes: When the experimental energy deviation is less than the energy conservation reference deviation, the scaling factor is set to a positive number greater than 1 to expand the initial parameter value; When the experimental energy deviation is greater than the energy conservation reference deviation, the scaling factor is set to a positive number less than 1 to reduce the initial parameter value; The adjusted parameter values are used as the current parameter values for the next iteration, and the steps from determining the direction of the energy gradient flow to calculating the trial energy deviation are re-executed.
[0013] In a preferred embodiment, the step of outputting parameter values whose ratios are stabilized within the tolerance range as the optimal parameter set includes: After each iteration completes parameter scaling and updates the current parameter value, the current ratio of the trial energy deviation to the energy conservation baseline deviation is recalculated, and it is determined whether the current ratio falls within the tolerance range multiple times consecutively. If the current ratio is within the tolerance range for multiple consecutive iterations, the ratio is determined to be stable, and the parameter value used in the last iteration is output as the optimal parameter set. If the current ratio does not fall within the tolerance range multiple times consecutively, the iteration continues until the stability condition is met.
[0014] To address the aforementioned problems, the present invention also provides a digital twin parameter tuning system based on energy-scaling, the system comprising: The topology benchmark modeling module is used to construct the energy flow topology of the digital twin model and collect the energy conservation benchmark deviation of the physical entity under standard operating conditions. The gradient situation analysis module is used to substitute the initial parameter values into the digital twin model for operation and identify local minima of the energy equipotential surface distribution map during operation in order to determine the direction of energy gradient flow. The parameter trial calculation module is used to superimpose the parameter movement direction vector with the current parameter value to obtain the trial parameter value, and substitute it into the digital twin model to obtain the total trial energy; The deviation optimization and tuning module is used to take the difference between the total test energy and the total input energy of the physical entity as the test energy deviation. When the ratio of the test energy deviation to the energy conservation benchmark deviation exceeds the tolerance range, the initial parameter value is scaled and the parameter value is output as the optimal parameter set when the ratio is stabilized within the tolerance range.
[0015] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention constructs a unique energy flow topology for the digital twin model, comprehensively analyzes the associations between energy nodes and energy connection edges, and synchronously collects the energy conservation benchmark deviation of the physical entity under standard operating conditions. This establishes a unified and accurate energy reference system for digital twin parameter optimization, ensuring that parameter adjustments closely match the actual energy transmission and conversion patterns of the physical entity. This significantly improves the consistency between the simulated energy state of the digital twin and the actual energy state of the physical entity, providing stable and objective benchmark support for parameter optimization and guaranteeing the accuracy and reliability of the optimization basis.
[0016] 2. This invention can accurately identify local minima of energy equipotential surface distribution during the operation of a digital twin, scientifically determine the direction of energy gradient flow to guide parameter adjustment paths, obtain the total energy of the trial through parameter trial calculation and complete deviation verification, and rely on the successive energy scaling iteration mechanism to stably constrain the deviation ratio within the preset tolerance range, effectively accelerating the parameter iteration convergence speed, comprehensively improving the efficiency and accuracy of parameter tuning, and stably outputting the optimal parameter set with strong adaptability and high reliability, ensuring that the entire process of digital twin parameter tuning is efficient and controllable, and the results are accurate and stable. Attached Figure Description
[0017] Figure 1 This is a flowchart illustrating a digital twin parameter tuning method based on energy-scaling according to an embodiment of the present invention. Figure 2 A functional block diagram of a digital twin parameter tuning system based on energy scaling is provided in one embodiment of the present invention. The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation
[0018] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
[0019] This application provides a method for tuning digital twin parameters based on energy-scaled scaling. The executing entity of this energy-scaled digital twin parameter tuning method includes, but is not limited to, at least one of the following electronic devices that can be configured to execute the method provided in this application: a server, a terminal, etc. In other words, the energy-scaled digital twin parameter tuning method can be executed by software or hardware installed on a terminal device or a server device. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster. The server can be an independent server or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDN), and big data and artificial intelligence platforms.
[0020] Reference Figure 1 The diagram shown is a flowchart illustrating a digital twin parameter tuning method based on energy-scaled scaling according to an embodiment of the present invention. In this embodiment, the digital twin parameter tuning method based on energy-scaled scaling includes: Construct the energy flow topology of the digital twin model and collect the energy conservation benchmark deviation of the physical entity under standard operating conditions; In this embodiment of the invention, the energy flow topology for constructing the digital twin model includes: Traverse all adjustable parameters in the digital twin model and mark the function controlled by each adjustable parameter as an energy node; Extract the data transmission paths between each energy node, mark each data transmission path as an energy connection edge, and determine the direction of each energy connection edge by adjusting the positive and negative correlation between the direction and the change in output energy according to the parameters. The energy flow topology is formed by combining the directions of all energy nodes and energy connection edges.
[0021] The energy conservation benchmark deviation of the physical entity under standard operating conditions includes: The difference between the total input energy and the total output energy of a physical entity when it is running stably under standard operating conditions is defined as the energy conservation benchmark deviation.
[0022] The digital twin model is disassembled layer by layer according to its functional modules. All adjustable control items within the model are thoroughly examined. The actual control function, target, and execution logic of each adjustable control item during model operation are clearly defined. Each independent control function is separately designated and marked as a dedicated energy node. The complete calibration of all energy nodes within the model is completed in sequence.
[0023] Based on the functional affiliation and operational interaction of energy nodes, a comprehensive review of the information transmission and energy interaction paths between all identified energy nodes is conducted. Each independent information and energy transmission path is marked as an energy connection edge. For each energy connection edge, the association attribute is determined one by one. When the adjustment direction of the adjustable control item is consistent with the output energy change of the corresponding energy node, the direction of the energy connection edge is set to positive. When the adjustment direction of the adjustable control item is opposite to the output energy change of the corresponding energy node, the direction of the energy connection edge is set to negative. The precise determination of the direction of all energy connection edges is completed in this way.
[0024] Based on the energy transmission logic and node interaction rules of the digital twin model, all the marked energy nodes and all the determined positive and negative energy connection edges are integrated and arranged according to their corresponding relationships, so that all energy nodes and energy connection edges form a closed and complete associated interaction system, and finally combine to construct the complete energy flow topology of the digital twin model.
[0025] The physical entity is placed in a preset standard operating environment and continuously started and run. The core operating indicators of the physical entity, such as speed, power, and temperature, are monitored in real time until the core operating indicators remain stable without numerical fluctuations or state deviations for a preset continuous stable duration of 300 seconds. This completes the formal confirmation of the stable operating state of the physical entity under standard operating conditions.
[0026] Throughout the entire cycle of stable operation of the physical entity under standard operating conditions, a dedicated energy acquisition device is used to collect and accumulate all forms of energy connected to the input end of the physical entity in real time. This process fully records all the energy at the input end throughout the entire cycle and summarizes it to obtain the total energy at the input end of the physical entity under stable operating conditions.
[0027] Throughout the entire cycle of stable operation of the physical entity under standard working conditions, a dedicated energy acquisition device is used to collect and accumulate all forms of energy output from the physical entity's output terminal in real time. This process fully records all the energy output from the output terminal throughout the entire cycle and summarizes the total energy output from the physical entity under stable operation under standard working conditions.
[0028] The total energy value at the input end of a physical entity operating stably under standard conditions is subtracted from the total energy value at the output end under the same conditions. The resulting unique value is directly determined as the energy conservation benchmark deviation of the physical entity under standard conditions.
[0029] The initial parameter values are substituted into the digital twin model for operation, and local minima of the energy equipotential surface distribution map during operation are identified to determine the direction of energy gradient flow. In this embodiment of the invention, the step of substituting the initial parameter values into the digital twin model for operation and identifying local minima in the energy equipotential surface distribution map during operation includes: Map the energy values of each energy node output after running the digital twin model to the corresponding positions in the parameter space, and calculate the energy difference between each parameter position and its neighboring parameter positions; The locations of parameters where all energy differences are positive and the absolute value exceeds a preset threshold are marked as local minima.
[0030] Determining the direction of the energy gradient flow includes: Starting from the current parameter position, calculate the direction vector pointing from the current parameter position to each local minimum region, and select the direction with the fastest energy decrease rate as the energy gradient flow direction. The energy decrease rate is determined by dividing the energy difference between the current parameter position and the local minimum region by the parameter spatial distance.
[0031] After the digital twin model completes its full operation, the energy value corresponding to each energy node is precisely placed into a specific and fixed coordinate position in the parameter space according to the unique correspondence between the energy node and the parameter space coordinates. This completes the mapping and arrangement of all energy node energy values to their corresponding coordinate positions in the parameter space without any omissions.
[0032] For each independent coordinate position within the parameter space, the energy value mapped to that coordinate position is accurately read. At the same time, the energy values mapped to all directly adjacent coordinate positions of the same dimension around that coordinate position are read. The energy value of the current coordinate position is subtracted from the energy value of each directly adjacent coordinate position in turn to obtain the energy difference between the current coordinate position and each neighboring coordinate position.
[0033] A threshold value of 0.05 energy standard units is preset to determine the absolute value of the energy difference in a local minimal region. All neighboring energy differences corresponding to each coordinate position in the parameter space are checked one by one to confirm that all neighboring energy differences corresponding to the coordinate position are positive and that the absolute value of each neighboring energy difference is greater than or equal to the preset threshold value of 0.05 energy standard units.
[0034] The spatial coordinates of the parameters that simultaneously satisfy the condition that all neighboring energy differences are positive and the absolute values of all neighboring energy differences are greater than or equal to the preset absolute value threshold of energy difference are uniformly marked as local minima in the energy equipotential surface distribution map.
[0035] Starting from the current coordinate position in the iterative process within the parameter space, a straight path is generated sequentially toward each marked local minima coordinate position. Based on the spatial arrangement and orientation relationship between the current coordinate position and the coordinate positions of each local minima, the direction vector corresponding to each pointing path is accurately determined.
[0036] Read the energy value mapped to the current coordinate position in the parameter space, and at the same time read the energy value mapped to the coordinate position of each marked local minima. Subtract the energy value of the corresponding local minima from the energy value of the current coordinate position to obtain the energy difference between the current coordinate position and the local minima.
[0037] Accurately measure the straight-line spatial length from the current coordinate position to the coordinate position of each local minima within the parameter space, and directly use this straight-line spatial length as the parameter space distance between the current coordinate position and the corresponding local minima.
[0038] Divide the energy difference between the current coordinate position and the local minima by the parameter space distance between the current coordinate position and the local minima to obtain the energy decrease rate corresponding to that direction. Check the energy decrease rate values corresponding to all directions pointing to the local minima one by one, and finally determine the direction with the largest energy decrease rate value as the energy gradient flow direction in the operation of the digital twin model.
[0039] The parameter movement direction vector is superimposed with the current parameter value to obtain the trial parameter value, which is then substituted into the digital twin model to obtain the total trial energy. In this embodiment of the invention, the step of superimposing the parameter movement direction vector with the current parameter value to obtain the trial parameter value, and substituting it into the digital twin model to obtain the total trial energy, includes: Starting from the current parameter value, move one step distance along the direction indicated by the parameter movement direction vector, and take the parameter space position reached after the movement as the trial parameter value. Write the trial parameter values into the corresponding parameter interface of the digital twin model to drive the digital twin model to complete a full simulation; Extract the output values of each energy node from the simulation results and sum them up. Use the summation result as the total trial energy.
[0040] Using the current coordinate position determined by the iterative process in the parameter space as the displacement starting point, strictly follow the fixed spatial orientation pointed to by the parameter movement direction vector, and perform a single linear spatial displacement operation according to the pre-set reference step length of 0.02 parameter space units. The unique parameter space coordinate position reached after the displacement ends is directly used as the trial parameter value corresponding to this iteration process.
[0041] The trial parameter values determined in this iteration are accurately written through the dedicated configuration port reserved in the digital twin model to ensure that the trial parameter values are completely transmitted to the corresponding functional modules inside the model. Relying on the stable information transmission of the configuration port, the digital twin model is driven to perform a full-cycle simulation operation without interruption according to the preset complete operation process.
[0042] From the complete output of the full-cycle simulation of the digital twin model, the energy output values corresponding to each of the calibrated energy nodes are extracted one by one without omission. All the extracted energy output values are accumulated in the order of energy node calibration. The final value obtained after the accumulation is directly used as the total test energy corresponding to this simulation.
[0043] The difference between the total energy of the trial and the total energy of the physical entity's input is taken as the trial energy deviation. When the ratio of the trial energy deviation to the energy conservation benchmark deviation exceeds the tolerance range, the initial parameter value is scaled and the parameter value is output as the optimal parameter set when the ratio is stabilized within the tolerance range.
[0044] In this embodiment of the invention, the difference between the total test energy and the total input energy of the physical entity is used as the test energy deviation. When the ratio of the test energy deviation to the energy conservation benchmark deviation exceeds the tolerance range, the initial parameter value is scaled, including: Calculate the ratio of the experimental energy deviation to the energy conservation baseline deviation, and compare the ratio with the preset tolerance range; If the ratio is greater than the upper limit of the tolerance range or less than the lower limit of the tolerance range, it is judged as exceeding the tolerance range. The scaling factor is determined according to the degree of deviation between the ratio and the tolerance range boundary. The current parameter value is multiplied by the scaling factor to obtain the adjusted parameter value.
[0045] The scaling of the initial parameter value further includes: When the experimental energy deviation is less than the energy conservation reference deviation, the scaling factor is set to a positive number greater than 1 to expand the initial parameter value; When the experimental energy deviation is greater than the energy conservation reference deviation, the scaling factor is set to a positive number less than 1 to reduce the initial parameter value; The adjusted parameter values are used as the current parameter values for the next iteration, and the steps from determining the direction of the energy gradient flow to calculating the trial energy deviation are re-executed.
[0046] The step of stabilizing the ratio within the tolerance range and outputting the parameter values as the optimal parameter set includes: After each iteration completes parameter scaling and updates the current parameter value, the current ratio of the trial energy deviation to the energy conservation baseline deviation is recalculated, and it is determined whether the current ratio falls within the tolerance range multiple times consecutively. If the current ratio is within the tolerance range for multiple consecutive iterations, the ratio is determined to be stable, and the parameter value used in the last iteration is output as the optimal parameter set. If the current ratio does not fall within the tolerance range multiple times consecutively, the iteration continues until the stability condition is met.
[0047] The total energy value obtained by complete simulation using a digital twin model is subtracted from the total energy value collected by the physical entity during a stable operating cycle under standard conditions. The difference between the two sets of energy values under the same period and operating conditions is calculated, and the unique difference result is directly determined as the experimental energy deviation corresponding to this iteration.
[0048] The precise ratio between the experimental energy deviation and the energy conservation benchmark deviation under the standard working conditions of the physical entity, calculated using the experimental energy deviation value obtained in this iteration, is obtained by dividing it by the previously calibrated energy conservation benchmark deviation value of the physical entity.
[0049] The lower limit of the tolerance range for determining the compliance of deviations is preset to 0.95 and the upper limit of the tolerance range is preset to 1.05. The calculated ratio is first compared with the lower limit of the tolerance range, and then compared with the upper limit of the tolerance range to complete a precise comparison one by one.
[0050] When the ratio is greater than the upper limit of the tolerance range of 1.05, or less than the lower limit of the tolerance range of 0.95, it is officially determined that the ratio exceeds the preset deviation compliance tolerance range.
[0051] Calculate the absolute difference between the ratio and the corresponding boundary value of the tolerance range, and determine the corresponding scaling factor based on the specific value of the absolute difference. The magnitude of the difference and the value of the scaling factor maintain a predefined fixed correspondence.
[0052] Multiply the value corresponding to the current coordinate position in the parameter space by the determined scaling factor, and use the result of multiplying the two sets of values directly as the adjusted parameter space position value.
[0053] The experimental energy deviation value obtained in this iteration is directly compared with the calibrated energy conservation benchmark deviation value. When it is confirmed that the experimental energy deviation value is less than the energy conservation benchmark deviation value, the scaling factor is set to a fixed positive number greater than 1. This positive number is then multiplied with the initial parameter spatial position value to make the initial parameter spatial position value proportionally expanded according to the scaling factor.
[0054] The experimental energy deviation value obtained in this iteration is directly compared with the calibrated energy conservation benchmark deviation value. When it is confirmed that the experimental energy deviation value is greater than the energy conservation benchmark deviation value, the scaling factor is set to a fixed positive number less than 1. This positive number is then multiplied with the initial parameter spatial position value to make the initial parameter spatial position value scaled down proportionally according to the scaling factor.
[0055] The parameter space position value after scaling is set as the current parameter space position value for the next iteration, and all continuous operation steps from determining the energy gradient flow direction to calculating the trial energy deviation are re-executed according to the established process.
[0056] After each iteration completes the scaling of the parameter space position value and updates the current parameter space position value, the current trial energy deviation value is divided by the current energy conservation baseline deviation value to obtain the current ratio of the trial energy deviation to the energy conservation baseline deviation corresponding to this iteration.
[0057] The threshold for consecutive compliance counts used to determine a stable state is set to 5 times. The current ratio is compared with the lower and upper limits of the tolerance range in turn, and the number of consecutive compliance counts when the current ratio is within the tolerance range is recorded one by one.
[0058] When the number of consecutive successful tests reaches the preset threshold of 5 consecutive successful tests, it is determined that the current ratio of the trial energy deviation to the energy conservation benchmark deviation has reached a stable state. All parameter spatial position values used in the last iteration are retrieved, and this set of values is combined as the optimal parameter set for formal output.
[0059] When the number of consecutive successful attempts recorded does not reach the preset threshold of 5 consecutive successful attempts, the loop operation of parameter scaling, ratio calculation and consecutive success determination is continuously executed until the number of consecutive successful attempts meets the stability condition corresponding to the preset threshold of consecutive successful attempts.
[0060] like Figure 2 The diagram shown is a functional block diagram of a digital twin parameter tuning system based on energy scaling according to an embodiment of the present invention.
[0061] The energy-scale-based digital twin parameter tuning system 100 described in this invention can be installed in an electronic device. Depending on the functions implemented, the energy-scale-based digital twin parameter tuning system 100 may include a topology baseline modeling module 101, a gradient situation analysis module 102, a parameter trial calculation module 103, and a deviation optimization tuning module 104. The modules described in this invention can also be referred to as units, which are a series of computer program segments that can be executed by the processor of an electronic device and perform a fixed function, stored in the memory of the electronic device.
[0062] In this embodiment, the functions of each module / unit are as follows: The topology benchmark modeling module 101 is used to construct the energy flow topology of the digital twin model and collect the energy conservation benchmark deviation of the physical entity under standard operating conditions. The gradient situation analysis module 102 is used to substitute the initial parameter values into the digital twin model for operation, and identify the local minima of the energy equipotential surface distribution map during operation, so as to determine the direction of energy gradient flow. The parameter trial calculation module 103 is used to superimpose the parameter movement direction vector with the current parameter value to obtain the trial parameter value, and substitute it into the digital twin model to obtain the total trial energy. The deviation optimization and tuning module 104 is used to take the difference between the total test energy and the total input energy of the physical entity as the test energy deviation. When the ratio of the test energy deviation to the energy conservation benchmark deviation exceeds the tolerance range, the initial parameter value is scaled and the parameter value is output as the optimal parameter set when the ratio is stabilized within the tolerance range.
[0063] In the several embodiments provided by this invention, it should be understood that the disclosed methods and systems can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and other division methods may be used in actual implementation.
[0064] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment, depending on actual needs.
[0065] Furthermore, the functional modules in the various embodiments of the present invention 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 in the form of hardware plus software functional modules.
[0066] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.
[0067] This application embodiment can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence is the theory, method, technology, and application system that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.
[0068] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.
Claims
1. A method for tuning digital twin parameters based on energy-scaling, characterized in that, The method includes: Construct the energy flow topology of the digital twin model and collect the energy conservation benchmark deviation of the physical entity under standard operating conditions; The initial parameter values are substituted into the digital twin model for operation, and local minima of the energy equipotential surface distribution map during operation are identified to determine the direction of energy gradient flow. The parameter movement direction vector is superimposed with the current parameter value to obtain the trial parameter value, which is then substituted into the digital twin model to obtain the total trial energy. The difference between the total energy of the trial and the total energy of the physical entity's input is taken as the trial energy deviation. When the ratio of the trial energy deviation to the energy conservation benchmark deviation exceeds the tolerance range, the initial parameter value is scaled and the parameter value is output as the optimal parameter set when the ratio is stabilized within the tolerance range.
2. The method for tuning digital twin parameters based on energy-scaling as described in claim 1, characterized in that, The energy flow topology for constructing the digital twin model includes: Traverse all adjustable parameters in the digital twin model and mark the function controlled by each adjustable parameter as an energy node; Extract the data transmission paths between each energy node, mark each data transmission path as an energy connection edge, and determine the direction of each energy connection edge by adjusting the positive and negative correlation between the direction and the change in output energy according to the parameters. The energy flow topology is formed by combining the directions of all energy nodes and energy connection edges.
3. The method for tuning digital twin parameters based on energy-scaling as described in claim 2, characterized in that, The energy conservation benchmark deviation of the physical entity under standard operating conditions includes: The difference between the total input energy and the total output energy of a physical entity when it is running stably under standard operating conditions is defined as the energy conservation benchmark deviation.
4. The method for tuning digital twin parameters based on energy-scaling as described in claim 1, characterized in that, The step of substituting the initial parameter values into the digital twin model for operation and identifying local minima in the energy equipotential surface distribution map during operation includes: Map the energy values of each energy node output after running the digital twin model to the corresponding positions in the parameter space, and calculate the energy difference between each parameter position and its neighboring parameter positions; The locations of parameters where all energy differences are positive and the absolute value exceeds a preset threshold are marked as local minima.
5. The method for tuning digital twin parameters based on energy-scaling as described in claim 4, characterized in that, Determining the direction of the energy gradient flow includes: Starting from the current parameter position, calculate the direction vector pointing from the current parameter position to each local minimum region, and select the direction with the fastest energy decrease rate as the energy gradient flow direction. The energy decrease rate is determined by dividing the energy difference between the current parameter position and the local minimum region by the parameter spatial distance.
6. The method for tuning digital twin parameters based on energy-scaling as described in claim 1, characterized in that, The step of superimposing the parameter movement direction vector with the current parameter value to obtain the trial parameter value, and substituting it into the digital twin model to obtain the total trial energy, includes: Starting from the current parameter value, move one step distance along the direction indicated by the parameter movement direction vector, and take the parameter space position reached after the movement as the trial parameter value. Write the trial parameter values into the corresponding parameter interface of the digital twin model to drive the digital twin model to complete a full simulation; Extract the output values of each energy node from the simulation results and sum them up. Use the summation result as the total trial energy.
7. The method for tuning digital twin parameters based on energy-scaling as described in claim 1, characterized in that, The difference between the total test energy and the total input energy of the physical entity is taken as the test energy deviation. When the ratio of the test energy deviation to the energy conservation benchmark deviation exceeds the tolerance range, the initial parameter value is scaled, including: Calculate the ratio of the experimental energy deviation to the energy conservation baseline deviation, and compare the ratio with the preset tolerance range; If the ratio is greater than the upper limit of the tolerance range or less than the lower limit of the tolerance range, it is judged as exceeding the tolerance range. The scaling factor is determined according to the degree of deviation between the ratio and the tolerance range boundary. The current parameter value is multiplied by the scaling factor to obtain the adjusted parameter value.
8. The method for tuning digital twin parameters based on energy-scaling as described in claim 7, characterized in that, The scaling of the initial parameter value further includes: When the experimental energy deviation is less than the energy conservation reference deviation, the scaling factor is set to a positive number greater than 1 to expand the initial parameter value; When the experimental energy deviation is greater than the energy conservation reference deviation, the scaling factor is set to a positive number less than 1 to reduce the initial parameter value; The adjusted parameter values are used as the current parameter values for the next iteration, and the steps from determining the direction of the energy gradient flow to calculating the trial energy deviation are re-executed.
9. The method for tuning digital twin parameters based on energy-scaling as described in claim 8, characterized in that, The step of stabilizing the ratio within the tolerance range and outputting the parameter values as the optimal parameter set includes: After each iteration completes parameter scaling and updates the current parameter value, the current ratio of the trial energy deviation to the energy conservation baseline deviation is recalculated, and it is determined whether the current ratio falls within the tolerance range multiple times consecutively. If the current ratio is within the tolerance range for multiple consecutive iterations, the ratio is determined to be stable, and the parameter value used in the last iteration is output as the optimal parameter set. If the current ratio does not fall within the tolerance range multiple times consecutively, the iteration continues until the stability condition is met.
10. A digital twin parameter tuning system based on energy-scaled scaling, used to implement the digital twin parameter tuning method based on energy-scaled scaling as described in claim 1, the system comprising: The topology benchmark modeling module is used to construct the energy flow topology of the digital twin model and collect the energy conservation benchmark deviation of the physical entity under standard operating conditions. The gradient situation analysis module is used to substitute the initial parameter values into the digital twin model for operation and identify local minima of the energy equipotential surface distribution map during operation in order to determine the direction of energy gradient flow. The parameter trial calculation module is used to superimpose the parameter movement direction vector with the current parameter value to obtain the trial parameter value, and substitute it into the digital twin model to obtain the total trial energy; The deviation optimization and tuning module is used to take the difference between the total test energy and the total input energy of the physical entity as the test energy deviation. When the ratio of the test energy deviation to the energy conservation benchmark deviation exceeds the tolerance range, the initial parameter value is scaled and the parameter value is output as the optimal parameter set when the ratio is stabilized within the tolerance range.