Digital twin modeling method for power module

Abstract

The invention relates to the technical field of digital twinning, in particular to a digital twinning modeling method for a power module. Data parameters needing to be collected are determined according to core requirements, historical data corresponding to the data parameters in a power module are obtained, a data driving model and a three-dimensional visualization model are established according to the historical data, and a mechanism model is established in combination with physical laws and design parameters. Fusing the mechanism model and the data driving model, integrating a three-dimensional visualization model to generate a digital twinborn model of the power module, judging generation of the digital twinborn model based on an aggregation mismatch index, judging a reason for non-conforming to a standard, and generating a corresponding processing instruction; and adjusting the corresponding parameters or sending out corresponding notifications based on the corresponding processing instructions. According to the method, accurate generation of the digital twinborn model is effectively guaranteed, and the generation efficiency of the digital twinborn model is effectively improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of digital twinning, in particular to a digital twinning modeling method for a power module. BACKGROUND

[0002] For power enterprises, handheld PDA is a customized terminal for digitalization of power field work. This device is different from ordinary consumer terminals and has industrial protection performance such as dustproof, waterproof, and drop resistance. The power module is a core functional component that gives handheld PDA professional power detection capability, such as an electric energy metering module, a power consumption inspection module, and an infrared temperature measurement module. Digital twinning modeling of the power module can achieve deeper cognition, prediction, and optimization. The digital twinning model of the power module not only reflects the appearance of the power module, but also real-time maps the running state and health state of the power module, thereby achieving predictive maintenance and timely fault detection. The digital twinning model of the power module is a key cornerstone supporting the stability, efficiency, and sustainability of modern life. Therefore, it is crucial to research the digital twinning modeling direction of the power module.

[0003] Chinese Patent Publication No. CN118096117A discloses a power equipment fault detection method and system based on digital twinning, which relates to the technical field of digital twinning. The method includes: performing simulation modeling according to intermediate values of multiple environment neighborhoods and power equipment basic information to construct an equipment twinning model library; matching to obtain an adaptive equipment twinning model for synchronous operation, comparing the actual operation data with the synchronous simulation operation data, predicting equipment faults based on the deviation comparison result, generating a predicted fault type and a predicted occurrence node; obtaining an associated fault type set according to the predicted fault type; determining a device maintenance window according to the predicted occurrence node, and performing device fault maintenance.

[0004] As can be seen from the above, the above-mentioned scheme obtains the equipment twinning model for synchronous operation, compares the actual operation data with the synchronous simulation operation data to predict the equipment fault, improves the accuracy and comprehensiveness of fault prediction, and achieves the effect of ensuring the stability of equipment operation. However, the above-mentioned scheme cannot guarantee the accurate generation of the digital twinning model, thereby cannot guarantee the generation efficiency of the digital twinning model. SUMMARY

[0005] Therefore, the present application provides a digital twinning modeling method for a power module to overcome the problem that the prior art cannot guarantee the accurate generation of the digital twinning model, thereby resulting in low generation efficiency of the digital twinning model.

[0006] To achieve the above-mentioned purpose, the present application provides a digital twinning modeling method for a power module, comprising: determining the core requirements of power module modeling and determining the data parameters to be collected according to the core requirements; acquire historical data corresponding to the data variables in the power module and perform noise reduction processing on the historical data; create a data-driven model and a three-dimensional visualization model according to the historical data, create a mechanism model in combination with physical laws and design parameters; fuse the mechanism model and the data-driven model and integrate the three-dimensional visualization model to generate a digital twin model of the power module; drive the mechanism model and the data-driven model by real-time data for simulation and calculation, and display the calculation results through the three-dimensional visualization model; continuously correct and update the parameters of the digital twin model using real-time data; determine the generation of the digital twin model based on the aggregate mismatch index; determine whether the generation of the digital twin model meets the standard based on the digital twin model, or, generate corresponding processing instructions, or, complete the determination of the digital twin model of the power module and trigger the digital twin model to execute subsequent instructions; adjust the corresponding parameters based on the corresponding processing instructions, and issue a corresponding notification.

[0007] Further, the process of determining the generation of the digital twin model based on the aggregate mismatch index includes: when the aggregate mismatch index is less than a preset aggregate mismatch index, determine whether the generation of the digital twin model meets the standard based on the digital twin model; when the aggregate mismatch index is greater than or equal to the preset aggregate mismatch index, determine that the generation of the digital twin model does not meet the standard, and adjust the physical calibration parameters of the mechanism model based on the system mismatch.

[0008] Further, the process of determining whether the generation of the digital twin model meets the standard based on the digital twin model includes: determine the number of models in the digital twin model whose normalized error is greater than or equal to the preset aggregate mismatch index; determine whether the generation of the digital twin model meets the standard based on the number of models; when the number of models is less than a preset number of models, determine that the generation of the digital twin model meets the standard, complete the determination of the digital twin model of the power module, and trigger the digital twin model to execute subsequent instructions; when the number of models is greater than or equal to the preset number of models, determine that the generation of the digital twin model does not meet the standard, and adjust the corresponding parameters based on the type of digital twin model that does not meet the standard.

[0009] Further, the process of adjusting the corresponding parameter based on the digital twin model type that does not meet the standard comprises: When the digital twin model type that does not meet the standard is the mechanism model, adjusting the physical calibration parameter of the mechanism model based on the system mismatch; When the digital twin model type that does not meet the standard is the data-driven model, adjusting the equivalent noise bandwidth of the lock-in amplifier based on the signal-to-noise ratio; When the digital twin model type that does not meet the standard is the three-dimensional visualization model, issuing a notification of rework to correct the geometric size of the three-dimensional visualization model.

[0010] Further, the process of adjusting the physical calibration parameter of the mechanism model based on the system mismatch comprises: Determining the system adaptation, and inversely deducing the real gain value and the real zero drift value in the mechanism model according to the system mismatch; Replacing the original theoretical gain value and the theoretical zero drift value in the mechanism model with the real gain value and the real zero drift value.

[0011] Further, the process of determining the generation of the digital twin model based on the aggregate mismatch index after the adjustment of the physical calibration parameter is completed comprises: When the aggregate mismatch index is less than the preset aggregate mismatch index, adjusting the corresponding parameter based on whether the digital twin model type that does not meet the standard is still the mechanism model; When the aggregate mismatch index is greater than or equal to the preset aggregate mismatch index, determining that the generation of the digital twin model does not meet the standard, and adjusting the equivalent noise bandwidth of the lock-in amplifier based on the signal-to-noise ratio; When the digital twin model type that does not meet the standard is still the mechanism model, issuing a notification of mechanism model review.

[0012] Further, the process of reducing the equivalent noise bandwidth of the lock-in amplifier based on the signal-to-noise ratio comprises: Reducing the equivalent noise bandwidth of the lock-in amplifier based on the signal-to-noise ratio, and the reduction amplitude of the equivalent noise bandwidth of the lock-in amplifier is inversely proportional to the signal-to-noise ratio.

[0013] Further, the process of determining the generation of the digital twin model based on the aggregate mismatch index after the reduction of the equivalent noise bandwidth of the lock-in amplifier is completed comprises: When the aggregate mismatch index is less than the preset aggregate mismatch index, adjusting the corresponding parameter based on whether the digital twin model type that does not meet the standard is still the data-driven model; determining that the generation of the digital twin model does not meet the standard when the aggregation mismatch index is greater than or equal to the preset aggregation mismatch index, and adjusting the total amount of historical data based on the model generalization entropy; when the type of the digital twin model that does not meet the standard is still the data-driven model, adjusting the total amount of historical data based on the model generalization entropy.

[0014] Further, the process of increasing the total amount of historical data based on the model generalization entropy comprises: The total amount of historical data is increased based on the model generalization entropy, and the increase amplitude of the total amount of historical data is proportional to the model generalization entropy.

[0015] Further, the process of determining the generation of the digital twin model based on the aggregation mismatch index after the total amount of historical data is increased comprises: adjusting the corresponding parameters based on whether the type of the digital twin model that does not meet the standard is still the data-driven model when the aggregation mismatch index is less than the preset aggregation mismatch index; determining that the generation of the digital twin model does not meet the standard when the aggregation mismatch index is greater than or equal to the preset aggregation mismatch index, and issuing a digital twin model systematic mismatch notification; issuing a data-driven model review notification when the type of the digital twin model that does not meet the standard is still the data-driven model.

[0016] Compared with the prior art, the beneficial effects of the present application are that the generation of the digital twin model is determined based on the aggregation mismatch index, whether the generation of the digital twin model meets the standard is determined based on the digital twin model, the determination of whether the generation of the digital twin model meets the standard can be timely and accurately completed, the reason is determined when the generation of the digital twin model does not meet the standard, the corresponding processing instructions are generated based on the determined reason, the corresponding parameters are adjusted based on the corresponding processing instructions, and the corresponding notifications are issued, which effectively guarantees the accurate generation of the digital twin model and effectively improves the generation efficiency of the digital twin model.

[0017] Further, the generation of the digital twin model is determined based on the aggregation mismatch index, it is timely determined whether the generation of the digital twin model needs to be determined based on the digital twin model or the physical calibration parameters of the mechanism model need to be adjusted based on the system mismatch, the misjudgment is avoided, the accurate generation of the digital twin model is further guaranteed, and the generation efficiency of the digital twin model is further improved.

[0018] Furthermore, this invention determines whether the generated digital twin model conforms to the standard based on the number of models, promptly confirms that the generated digital twin model conforms to the standard and triggers the digital twin model to execute subsequent instructions, and promptly determines whether it is necessary to adjust the corresponding parameters based on the type of digital twin model that does not conform to the standard, so as to ensure the accuracy of the determination and avoid misjudgment. While further ensuring the accurate generation of digital twin models, it also further improves the generation efficiency of digital twin models.

[0019] Furthermore, this invention adjusts the corresponding parameters based on the type of non-compliant digital twin model, accurately determining whether physical calibration parameters need to be adjusted based on the system mismatch adjustment mechanism model, whether the equivalent noise bandwidth of the lock-in amplifier needs to be adjusted based on the signal-to-noise ratio, or whether a notification to rework and correct the geometric dimensions of the 3D visualization model needs to be issued. This ensures the generation of targeted processing instructions for the reasons for non-compliance, further guaranteeing the accurate generation of the digital twin model while also improving the generation efficiency of the digital twin model.

[0020] Furthermore, the present invention, based on the physical calibration parameters of the system mismatch adjustment mechanism model, effectively avoids the situation where the generation of digital twin models does not meet the standards due to the non-compliance of the physical calibration parameters of the mechanism model. While further ensuring the accurate generation of digital twin models, it also further improves the generation efficiency of digital twin models.

[0021] Furthermore, after the physical calibration parameters are adjusted, the present invention makes a judgment on the generation of the digital twin model based on the aggregate mismatch index, and further determines whether the equivalent noise bandwidth of the lock-in amplifier needs to be adjusted based on the signal-to-noise ratio or whether a mechanism model review notice needs to be issued. This not only further ensures the accurate generation of the digital twin model, but also further improves the generation efficiency of the digital twin model.

[0022] Furthermore, this invention reduces the equivalent noise bandwidth of the lock-in amplifier based on the signal-to-noise ratio, effectively avoiding the situation where the generated digital twin model does not meet the standard due to the equivalent noise bandwidth of the lock-in amplifier not meeting the standard. This further ensures the accurate generation of the digital twin model while improving the generation efficiency of the digital twin model.

[0023] Furthermore, after the equivalent noise bandwidth of the lock-in amplifier is reduced, the present invention makes a judgment on the generation of the digital twin model based on the aggregation mismatch index, and determines in a timely and accurate manner whether the total amount of historical data needs to be adjusted based on the model generalization entropy. This not only further ensures the accurate generation of the digital twin model, but also further improves the generation efficiency of the digital twin model.

[0024] Further, the application increases the total amount of historical data based on model generalization entropy, effectively avoiding the situation that the generation of the digital twin model does not meet the standard due to the total amount of historical data not meeting the standard, further ensuring the accurate generation of the digital twin model, and further improving the generation efficiency of the digital twin model.

[0025] Further, the application determines the generation of the digital twin model based on the aggregation mismatch index after the total amount of historical data is increased, determines in a timely manner whether to issue a digital twin model systematic mismatch notification or to issue a data-driven model review notification, avoids the occurrence of misjudgment, further ensures the accurate generation of the digital twin model, and further improves the generation efficiency of the digital twin model. BRIEF DESCRIPTION OF DRAWINGS

[0026] Figure 1 The structural block diagram of the system for the digital twin modeling method for the power module is used for the embodiment of the application; Figure 2 The flowchart of the digital twin modeling method for the power module is used for the embodiment of the application; Figure 3 The flowchart of determining whether the generation of the digital twin model meets the standard and determining the reason for not meeting the standard is used for the embodiment of the application; Figure 4 The flowchart of determining the reason for the generation of the digital twin model not meeting the standard is used for the embodiment of the application. DETAILED DESCRIPTION

[0027] In order to make the objects and advantages of the present application clearer, the following further describes the present application with reference to the embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application, and do not limit the present application.

[0028] The preferred embodiments of the present application are described below with reference to the accompanying drawings. Those skilled in the art should understand that these embodiments are only used to explain the technical principles of the present application, and are not intended to limit the protection scope of the present application.

[0029] In addition, it should be further pointed out that, in the description of the present application, unless otherwise explicitly specified and limited, the terms "mounting", "connection" and "connection" should be understood in a broad sense, for example, it can be fixed connection, or detachable connection, or integral connection; it can be mechanical connection, or electrical connection; it can be direct connection, or indirect connection through an intermediate medium, or internal communication of two elements. Those skilled in the art can understand the specific meaning of the above terms in the present application according to the specific circumstances.

[0030] Please refer to Figure 1As shown, it is a structural block diagram of the system using the digital twin modeling method for the power module in the embodiment of the application. The structure of the embodiment of the application includes a collection module, a creation module, a driving module, a correction module, a management module and an adjustment module; wherein, The collection module is used to determine the core requirements of power module modeling and determine the data parameters to be collected according to the core requirements; The collection module is also used to obtain the historical data corresponding to the data parameters in the power module and to perform noise reduction processing on the historical data; The creation module is connected with the collection module, used to create a data-driven model and a three-dimensional visualization model according to the historical data, and to create a mechanism model combined with physical laws and design parameters; The creation module is also used to fuse the mechanism model and the data-driven model and integrate the three-dimensional visualization model to generate the digital twin model of the power module; The driving module is connected with the creation module, used to drive the mechanism model and the data-driven model through real-time data for simulation and calculation, and to display the calculation results through the three-dimensional visualization model; The correction module is connected with the collection module, the creation module and the driving module respectively, used to continuously correct and update the parameters of the digital twin model using real-time data; The management module is connected with the creation module, used to determine the generation of the digital twin model based on the aggregation mismatch index; The management module is also used to determine whether the generation of the digital twin model meets the standard based on the digital twin model, Or, generate corresponding processing instructions, Or, complete the determination of the power module digital twin model, and trigger the digital twin model to execute subsequent instructions; The adjustment module is connected with the collection module, the creation module, the driving module, the correction module and the management module respectively, used to adjust the corresponding parameters based on the corresponding processing instructions, and to issue corresponding notifications; Specifically, triggering the digital twin model to execute subsequent instructions includes: triggering the driving module to drive the mechanism model and the data-driven model through real-time data for simulation and calculation, and displaying the calculation results through the three-dimensional visualization model, and triggering the correction module to continuously correct and update the parameters of the digital twin model using real-time data; The noise reduction processing can remove random noise and interference in the data to ensure the authenticity of the data; The historical data includes external measurement data, internal state data and environmental data; The mechanism model error weight is 50%, the data-driven model error weight is 40%, and the three-dimensional visualization model error weight is 10%, the qualified threshold of the mechanism model is 0.5V, the qualified threshold of the data-driven model is 0.15%, and the qualified threshold of the three-dimensional visualization model is 1.0mm, the normalized error of the mechanism model is the ratio of the actual error of the mechanism model to the qualified threshold of the mechanism model, the normalized error of the data-driven model is the ratio of the actual error of the data-driven model to the qualified threshold of the data-driven model, and the normalized error of the three-dimensional visualization model is the ratio of the actual error of the three-dimensional visualization model to the qualified threshold of the three-dimensional visualization model; The process for determining the actual error of the mechanism model is that the difference between the measured value of the calculation module and the output value of the model is calculated as a system mismatch, the squares of each difference are calculated, the average value of each square is calculated, and the average value is squared to obtain the actual error of the mechanism model. The process for determining the actual error of the data-driven model is that the predicted drift value and the actual drift value are determined, the absolute value of the difference between the predicted drift value and the corresponding actual drift value is calculated, and the average value of the absolute values of each difference is calculated to obtain the actual error of the data-driven model. The process for determining the actual error of the three-dimensional visualization model is that the absolute value of the difference between the size of the three-dimensional model and the size of the drawing is calculated, and the maximum value of the absolute values of each difference is selected as the actual error of the three-dimensional visualization model. The normalized error of the mechanism model is multiplied by the mechanism model error weight, the normalized error of the data-driven model is multiplied by the data-driven model error weight, and the normalized error of the three-dimensional visualization model is multiplied by the three-dimensional visualization model error weight, the sum of the mechanism model product, the data-driven model product and the three-dimensional visualization model product is calculated, and the calculated sum is taken as the aggregation mismatch index.

[0031] Referring to Figure 2 The flowchart of the digital twin modeling method for the power module according to the embodiment of the present application is shown in the figure. The method according to the embodiment of the present application comprises: Step S1, determining the core requirements of the power module modeling and determining the data parameters to be collected according to the core requirements; Step S2, obtaining the historical data corresponding to the data parameters in the power module and performing noise reduction processing on the historical data; Step S3, creating a data-driven model and a three-dimensional visualization model according to the historical data, and creating a mechanism model combining physical laws and design parameters; Step S4, fusing the mechanism model and the data-driven model and integrating the three-dimensional visualization model to generate a digital twin model of the power module; Step S5, simulate and calculate by driving the mechanism model and the data-driven model by the real-time data, and display the calculation result by the three-dimensional visualization model; continuously correct and update the digital twin model parameters by using real-time data; Step S6, determine the generation of the digital twin model based on the aggregated mismatch index; Step S7, determine whether the generation of the digital twin model meets the standard based on the digital twin model, or, generate corresponding processing instructions, Step S8, or, complete the determination of the power module digital twin model, and trigger the digital twin model to execute subsequent instructions; Step S9, adjust the corresponding parameters based on the corresponding processing instructions, and issue a corresponding notification.

[0032] Please refer to Figure 3 , which is a flowchart of the process of determining whether the generation of the digital twin model meets the standard and determining the reason for not meeting the standard according to the embodiment of the application. The process of determining the generation of the digital twin model based on the aggregated mismatch index according to the embodiment of the application includes: When the aggregated mismatch index is less than the preset aggregated mismatch index Y, determine whether the generation of the digital twin model meets the standard based on the digital twin model, wherein the preset aggregated mismatch index Y = 1 in this embodiment; When the aggregated mismatch index is greater than or equal to the preset aggregated mismatch index Y, determine that the generation of the digital twin model does not meet the standard, and adjust the physical calibration parameters of the mechanism model based on the system mismatch; Specifically, the preset aggregated mismatch index is set to 1, which is an empirical threshold.

[0033] Please continue to refer to Figure 3 , the process of determining whether the generation of the digital twin model meets the standard based on the digital twin model according to the embodiment of the application includes: Determine the number of models in the digital twin model whose normalized error is greater than or equal to the preset aggregated mismatch index; Determine whether the generation of the digital twin model meets the standard based on the number of models; When the number of models is less than the preset number of models Q, determine that the generation of the digital twin model meets the standard, complete the determination of the power module digital twin model, and trigger the digital twin model to execute subsequent instructions, wherein the preset number of models Q = 1 in this embodiment; When the number of models is greater than or equal to the preset number of models Q, determine that the generation of the digital twin model does not meet the standard, and adjust the corresponding parameters based on the type of digital twin model that does not meet the standard; Specifically, the embodiment presets the number of models to be 1, and can determine whether the normalized error of the mechanism model, the data-driven model and the three-dimensional visualization model is greater than or equal to the preset aggregated mismatch index.

[0034] Please continue to refer to Figure 3 As shown in the figure, the process of adjusting the corresponding parameters based on the digital twin model type that does not meet the standard in the embodiment of the application includes: When the digital twin model type that does not meet the standard is the mechanism model, adjusting the physical calibration parameter of the mechanism model based on the system mismatch; When the digital twin model type that does not meet the standard is the data-driven model, adjusting the equivalent noise bandwidth of the lock-in amplifier based on the signal-to-noise ratio; When the digital twin model type that does not meet the standard is the three-dimensional visualization model, issuing a notification of reworking and correcting the geometric size of the three-dimensional visualization model.

[0035] Please continue to refer to Figure 3 As shown in the figure, the process of adjusting the physical calibration parameter of the mechanism model based on the system mismatch in the embodiment of the application includes: Determining the system adaptation, and inversely deducing the real gain value and the real zero drift value in the mechanism model according to the system mismatch; Replacing the original theoretical gain value and the theoretical zero drift value in the mechanism model with the real gain value and the real zero drift value; Specifically, a parameter identification algorithm is used when inversely deducing the real gain value and the real zero drift value in the mechanism model according to the system mismatch.

[0036] Please refer to Figure 4 As shown in the figure, it is a flowchart for determining the reason why the generation of the digital twin model does not meet the standard in the embodiment of the application. The process of determining the generation of the digital twin model based on the aggregated mismatch index after the adjustment of the physical calibration parameter in the embodiment of the application includes: When the aggregated mismatch index is less than the preset aggregated mismatch index Y, adjusting the corresponding parameters based on whether the digital twin model type that does not meet the standard is still the mechanism model; When the aggregated mismatch index is greater than or equal to the preset aggregated mismatch index Y, determining that the generation of the digital twin model does not meet the standard, and adjusting the equivalent noise bandwidth of the lock-in amplifier based on the signal-to-noise ratio; When the digital twin model type that does not meet the standard is still the mechanism model, issuing a mechanism model review notification.

[0037] Please continue to refer to Figure 4As shown, the process of reducing the equivalent noise bandwidth of the phase-locked amplifier based on the signal-to-noise ratio in the embodiment of the application includes: When the signal-to-noise ratio is greater than or equal to a second preset signal-to-noise ratio T2, the equivalent noise bandwidth of the phase-locked amplifier is reduced to 0.52 times the equivalent noise bandwidth of the initial phase-locked amplifier, wherein the second preset signal-to-noise ratio T2 in the embodiment is 40 dB; When the signal-to-noise ratio is less than the second preset signal-to-noise ratio T2 and greater than or equal to a first preset signal-to-noise ratio T1, the equivalent noise bandwidth of the phase-locked amplifier is reduced to 0.39 times the equivalent noise bandwidth of the initial phase-locked amplifier, wherein the first preset signal-to-noise ratio T1 in the embodiment is 20 dB; When the signal-to-noise ratio is less than the first preset signal-to-noise ratio T1, the equivalent noise bandwidth of the phase-locked amplifier is reduced to 0.21 times the equivalent noise bandwidth of the initial phase-locked amplifier. Specifically, the lower the signal-to-noise ratio, the more the equivalent noise bandwidth of the phase-locked amplifier is reduced, and the values of the signal-to-noise ratio and the reduction multiple of the equivalent noise bandwidth of the phase-locked amplifier are obtained from actual debugging results.

[0038] Please continue to refer to Figure 4 As shown, the process of determining the generation of the digital twin model based on the aggregation mismatch index after the reduction of the equivalent noise bandwidth of the phase-locked amplifier in the embodiment of the application includes: When the aggregation mismatch index is less than the preset aggregation mismatch index Y, it is determined whether the digital twin model type that does not meet the standard is still the data-driven model, and the corresponding parameters are adjusted based on the determination result; When the aggregation mismatch index is greater than or equal to the preset aggregation mismatch index Y, it is determined that the generation of the digital twin model does not meet the standard, and the total amount of historical data is adjusted based on the model generalization entropy; When the digital twin model type that does not meet the standard is still the data-driven model, the total amount of historical data is adjusted based on the model generalization entropy.

[0039] Please continue to refer to Figure 4 As shown, the process of increasing the total amount of historical data based on the model generalization entropy in the embodiment of the application includes: When the model generalization entropy is greater than or equal to a second preset model generalization entropy G2, the total amount of historical data is increased to 2.67 times the initial total amount of historical data, wherein the second preset model generalization entropy G2 in the embodiment is 3.0Hval; when the model generalization entropy is less than the second preset model generalization entropy G2 and greater than or equal to the first preset model generalization entropy G1, the total amount of historical data is increased to 1.97 times of the initial total amount of historical data, wherein the first preset model generalization entropy G1 of the embodiment = 1.5Hval; when the model generalization entropy is less than the first preset model generalization entropy G1, the total amount of historical data is increased to 1.52 times of the initial total amount of historical data. Specifically, the greater the model generalization entropy, the greater the increase multiple of the total amount of historical data, and the values of the model generalization entropy and the total amount of historical data are obtained from actual debugging results.

[0040] Please continue to refer to Figure 4 As shown in the drawings, the process of determining the generation of the digital twin model based on the aggregation mismatch index after the total amount of historical data is increased in the embodiment of the application includes: when the aggregation mismatch index is less than the preset aggregation mismatch index Y, adjusting the corresponding parameters based on whether the digital twin model type that does not meet the standard is still the data-driven model; when the aggregation mismatch index is greater than or equal to the preset aggregation mismatch index Y, determining that the generation of the digital twin model does not meet the standard, and issuing a digital twin model systematic mismatch notification; when the digital twin model type that does not meet the standard is still the data-driven model, issuing a data-driven model review notification.

[0041] So far, the technical solutions of the application have been described in combination with the preferred embodiments shown in the drawings, but those skilled in the art can easily understand that the protection scope of the application is obviously not limited to these specific embodiments. Those skilled in the art can make equivalent changes or replacements to related technical features without departing from the principles of the application, and the technical solutions after these changes or replacements will fall within the protection scope of the application.

[0042] The above only describes the preferred embodiments of the application and is not used to limit the application; for those skilled in the art, the application can have various changes and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the application shall be included in the protection scope of the application.

Claims

1. A digital twin modeling method for power modules, characterized in that, include: Determine the core requirements for power module modeling and determine the data parameters to be collected based on these core requirements; Obtain historical data corresponding to the data parameters in the power module and perform noise reduction processing on the historical data; Based on the historical data, a data-driven model and a 3D visualization model are created, and a mechanism model is created by combining physical laws and design parameters. By integrating the mechanistic model and the data-driven model, and combining them with the 3D visualization model, a digital twin model of the power module is generated; The mechanism model and the data-driven model are simulated and calculated using real-time data, and the calculation results are displayed through the three-dimensional visualization model. Continuously correct and update digital twin model parameters using real-time data; The generation of digital twin models is determined based on the aggregation mismatch index; Based on the aforementioned digital twin model, determine whether the generation of the digital twin model conforms to the standard. Alternatively, generate corresponding processing instructions. Alternatively, it can complete the determination of the digital twin model of the power module and trigger the digital twin model to execute subsequent instructions; Based on the corresponding processing instructions, the relevant parameters are adjusted, and corresponding notifications are issued.

2. The digital twin modeling method for power modules according to claim 1, characterized in that, The process of determining the generation of a digital twin model based on the aggregate mismatch index includes: When the aggregation mismatch index is less than the preset aggregation mismatch index, it is determined whether the generation of the digital twin model meets the standard based on the digital twin model; When the aggregation mismatch index is greater than or equal to the preset aggregation mismatch index, it is determined that the generation of the digital twin model does not meet the standard, and the physical calibration parameters of the mechanism model are adjusted based on the system mismatch.

3. The digital twin modeling method for power modules according to claim 2, characterized in that, The process of determining whether the generation of the digital twin model conforms to the standard based on the aforementioned digital twin model includes: Determine the number of models in the digital twin model whose normalization error is greater than or equal to the preset aggregation mismatch index; The generation of the digital twin model is determined based on the number of models; When the number of models is less than the preset number of models, it is determined that the generation of the digital twin model meets the standard, the determination of the digital twin model of the power module is completed, and the digital twin model is triggered to execute subsequent instructions. When the number of models is greater than or equal to the preset number of models, it is determined that the generation of the digital twin model does not meet the standard, and the corresponding parameters are adjusted based on the type of digital twin model that does not meet the standard.

4. The digital twin modeling method for power modules according to claim 3, characterized in that, The process of adjusting the corresponding parameters based on the non-compliant digital twin model type includes: When the non-compliant digital twin model type is the mechanistic model, the physical calibration parameters of the mechanistic model are adjusted based on the system mismatch; When the data-driven model of the digital twin model type does not conform to the standard, the equivalent noise bandwidth of the lock-in amplifier is adjusted based on the signal-to-noise ratio; When a 3D visualization model does not conform to the standard digital twin model type, a rework notification to correct the geometry of the 3D visualization model is issued.

5. The digital twin modeling method for power modules according to claim 4, characterized in that, The process of adjusting the physical calibration parameters of the mechanism model based on the system mismatch includes: The system fit is determined, and the true gain value and true zero drift value in the mechanism model are deduced from the system mismatch. The original theoretical gain and theoretical zero drift values ​​in the mechanistic model are replaced with the actual gain value and the actual zero drift value.

6. The digital twin modeling method for power modules according to claim 5, characterized in that, The process of determining the generation of the digital twin model based on the aggregate mismatch index after the physical calibration parameters have been adjusted includes: When the aggregation mismatch index is less than the preset aggregation mismatch index, adjust the corresponding parameters based on whether the digital twin model type that does not meet the standard is still the mechanism model; When the aggregation mismatch index is greater than or equal to the preset aggregation mismatch index, it is determined that the generation of the digital twin model does not meet the standard, and the equivalent noise bandwidth of the lock-in amplifier is adjusted based on the signal-to-noise ratio; If the digital twin model type that does not meet the standard is still the mechanism model, a mechanism model review notice will be issued.

7. The digital twin modeling method for power modules according to claim 6, characterized in that, The process of reducing the equivalent noise bandwidth of the lock-in amplifier based on the signal-to-noise ratio includes: The equivalent noise bandwidth of the lock-in amplifier is reduced based on the signal-to-noise ratio, and the reduction in the equivalent noise bandwidth of the lock-in amplifier is inversely proportional to the signal-to-noise ratio.

8. The digital twin modeling method for power modules according to claim 7, characterized in that, The process of determining the generation of the digital twin model based on the aggregate mismatch index after the equivalent noise bandwidth of the lock-in amplifier has been reduced includes: When the aggregation mismatch index is less than the preset aggregation mismatch index, adjust the corresponding parameters based on whether the non-compliant digital twin model type is still the data-driven model; When the aggregation mismatch index is greater than or equal to the preset aggregation mismatch index, it is determined that the generation of the digital twin model does not meet the standard, and the total amount of historical data is adjusted based on the model generalization entropy. When the digital twin model type that does not meet the standard is still the data-driven model, the total amount of historical data is adjusted based on the model generalization entropy.

9. The digital twin modeling method for power modules according to claim 8, characterized in that, The process of increasing the total amount of historical data based on the generalization entropy of the model includes: The total amount of historical data is increased based on the model generalization entropy, and the increase in the total amount of historical data is proportional to the model generalization entropy.

10. The digital twin modeling method for power modules according to claim 9, characterized in that, The process of determining the generation of the digital twin model based on the aggregated mismatch index after the total amount of historical data has been increased includes: When the aggregation mismatch index is less than the preset aggregation mismatch index, adjust the corresponding parameters based on whether the non-compliant digital twin model type is still the data-driven model; When the aggregation mismatch index is greater than or equal to the preset aggregation mismatch index, it is determined that the generation of the digital twin model does not meet the standard, and a systemic mismatch notification of the digital twin model is issued. If the digital twin model type that does not meet the standards is still the data-driven model, a data-driven model review notice will be issued.

Images