Power electronic converter digital twin parameter correction method based on uncertainty quantization

Abstract

The invention belongs to but is not limited to the technical field of power electronic simulation, and discloses a method for correcting digital twin parameters of a power electronic converter based on uncertainty quantization. And performing parameter uncertainty representation of the power electronic converter simulation model, parameter sensitivity analysis based on forward uncertainty propagation and parameter correction based on reverse uncertainty propagation in sequence. Under the framework, the simulation model parameters of the power electronic converter are no longer regarded as fixed unknown numbers, but have probability distribution of statistical characteristics such as mean values and variances, and the correction result of the simulation model parameters can more accurately and comprehensively represent the device parameters of the power electronic converter, so that the prediction accuracy of the simulation model is improved. The method is simple and clear in steps and easy to implement. According to the method, with the purpose that model response and test data tend to be consistent, the random uncertainty of simulation model parameters is quantified, and the cognitive uncertainty of the simulation model parameters is reduced.

Description

Technical Field

[0001] The present invention belongs to, but is not limited to, the technical field of power electronics simulation, and particularly relates to a method for correcting the parameters of a digital twin of a power electronic converter based on uncertainty quantification. Background Art

[0002] In the research and development process of power electronic converters, controller hardware-in-the-loop (CHIL), as a key simulation technology, combines numerical simulation and hardware testing, bringing a significant boost to the entire research and development process. The CHIL simulation technology can use a hardware-in-the-loop test platform to conduct detailed testing and verification of the controller, enabling developers to perform code debugging in a safe and controllable environment, simulate various fault scenarios, and deeply explore the performance and limits of control strategies, thus greatly improving the overall efficiency of controller development.

[0003] The accuracy of the simulation model is crucial for the performance of the CHIL simulation technology. If the accuracy of the simulation model is insufficient, it will directly lead to a decline in the performance of CHIL. However, when faced with complex power electronic converters, it is a very challenging task to construct a high-fidelity simulation model at one time. Therefore, model correction work is essential.

[0004] Model correction refers to the process of continuously correcting the parameters of the simulation model to make the model response tend to be consistent with the experimental data. The goal of traditional model correction, that is, the deterministic model correction method, is to minimize the gap between the model response and the experimental data. This process is usually achieved through intelligent optimization algorithms such as particle swarm optimization algorithm, genetic algorithm, and simulated annealing algorithm. These algorithms can search for the optimal parameter set to make the model response match the experimental data.

[0005] However, there are a large number of uncertainties in physical modeling, numerical simulation, and experiments: inevitable simplification and approximation in the physical modeling process; due to the limitations of knowledge, it is impossible to accurately characterize the inherent parameters of the physical system in numerical simulation; the experimental data, as the basis for model correction, is affected by random factors (such as environmental noise, systematic error, and subjective judgment) during the acquisition process. The model corrected by the deterministic model correction method is very likely to have a large prediction deviation.

[0006] At present, all existing methods for correcting the parameters of power electronic converter simulation models do not consider the uncertainty factors of the simulation model parameters, which are called deterministic correction methods. That is, the simulation model parameters are regarded as fixed and unknown physical quantities, and a set of parameters are found through intelligent optimization algorithms to minimize the gap between the simulation model response and the experimental data. However, uncertainty is ubiquitous in the construction of the physical model of power electronic converters, numerical simulation, and the experimental process. In the process of constructing the physical model of power electronic converters, due to the necessity of simplification and approximation, it is impossible to fully capture all details. In numerical simulation, due to the limitations of our understanding of power electronic converters, it is impossible to accurately describe their internal parameters. As the key reference for correcting the parameters of the simulation model, the collection process of experimental data is easily affected by various random factors, including environmental noise, systematic errors, and subjective evaluations. If the above uncertainties from physical modeling, numerical simulation, and experiments are not considered, the results of correcting the simulation model parameters may lead to fluctuations in the response of the power electronic converter simulation model, and then lead to design failures, bringing catastrophic consequences.

[0007] In view of the above analysis, the technical problems urgently to be solved in the existing technology are as follows: To sum up, the current methods for correcting the parameters of power electronic converter simulation models are all carried out without considering parameter uncertainty, and correction deviations may occur due to the influence of parameter uncertainty. Therefore, it is particularly important to correct the parameters of the simulation model under the condition of considering parameter uncertainty. Summary of the Invention

[0008] Aiming at the problems existing in the prior art, the present invention provides a method for correcting the digital twin parameters of a power electronic converter based on uncertainty quantification.

[0009] The present invention is implemented as follows. A method for correcting the digital twin parameters of a power electronic converter based on uncertainty quantification includes the following steps: Step 1, parameter uncertainty characterization of the power electronic converter simulation model; Step 2, parameter sensitivity analysis based on forward uncertainty propagation; Step 3, parameter correction based on reverse uncertainty propagation.

[0010] Furthermore, in Step 1, the parameter uncertainty of the power electronic converter simulation model is represented in the form of a probability distribution; it is set that the device parameters follow a normal distribution with a mean of the rated value of the device parameters and a standard deviation of half of the device parameter tolerance.

[0011] Furthermore, in step 2, based on the forward propagation of parameter uncertainty, the influence degrees of different uncertain input parameters on the response of the power electronic converter simulation model are explored to evaluate the sensitivity of the simulation model response to the changes in input parameters, and the uncertain input parameters with less influence on the simulation model response are found and given fixed and known values, thereby effectively reducing the computational amount of the simulation model parameter correction.

[0012] Furthermore, based on the effective characterization of parameter uncertainty, the parameter uncertainty is transmitted through the power electronic converter simulation model, and then its influence on the simulation model response is analyzed and the parameter uncertainty of the simulation model is quantified.

[0013] Furthermore, in step 2, the cumulative probability density function is used to quantify the parameter uncertainty of the power electronic converter simulation model, that is, a suitable sampling method is selected according to the distribution of input parameters to sample the input parameters to obtain several groups of different input parameter combinations, then each group of input parameters is input into the simulation model to simulate the simulation model response, and then the cumulative probability density function of the simulation model response is obtained. At this time, the parameter uncertainty can be represented by the cumulative probability density function.

[0014] Furthermore, in step 3, based on the test data of the power electronic converter and the initial characterization of the parameter uncertainty of the simulation model (prior probability distribution), the optimal uncertainty characterization (posterior probability distribution) of the simulation model parameters is calculated by inverse deduction; when calculating by inverse deduction, the forward calculation of Bayesian inference is used to solve the problem of the optimal uncertainty characterization of the power electronic converter simulation model parameters.

[0015] Another object of the present invention is to provide a simulation model parameter correction system for a power electronic converter based on uncertainty quantification of a digital twin parameter correction method for a power electronic converter based on uncertainty quantification, including: A parameter uncertainty characterization module for characterizing the parameter uncertainty of the power electronic converter simulation model; A parameter sensitivity analysis module for parameter sensitivity analysis based on forward uncertainty propagation; A parameter correction module for parameter correction based on reverse uncertainty propagation.

[0016] Another object of the present invention is to provide a computer device, which includes a memory and a processor. When the computer program stored in the memory is executed by the processor, the processor executes the steps of the digital twin parameter correction method for a power electronic converter based on uncertainty quantification.

[0017] Another object of the present invention is to provide a computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to execute the steps of the method for correcting the digital twin parameters of a power electronic converter based on uncertainty quantification.

[0018] Another object of the present invention is to provide an information data processing terminal, which includes the system for correcting the simulation model parameters of a power electronic converter based on uncertainty quantification.

[0019] Combined with the above technical solutions and the technical problems solved, the advantages and positive effects of the technical solutions to be protected by the present invention are as follows: First, the method for correcting the digital twin parameters of a power electronic converter based on uncertainty quantification disclosed by the present invention fully considers the uncertainty from the simulation model parameters. The present invention introduces the parameter uncertainty of the simulation model. The simulation model parameters are no longer regarded as fixed unknowns, but probability distributions with statistical characteristics such as mean and variance. The correction results of the simulation model parameters can more accurately and comprehensively represent the device parameters of the power electronic converter, thereby improving the accuracy of the simulation model prediction.

[0020] Second, the technical problems solved by the technical solution of the present invention in industrial applications.

[0021] The simulation models of traditional power electronic converters usually regard parameters as fixed values, ignoring the randomness and uncertainty of actual component parameters caused by factors such as manufacturing tolerances and environmental impacts. This leads to a deviation between the simulation results and the experimental data, reducing the accuracy and practicality of the model.

[0022] The prior art lacks systematicness in the face of random uncertainty, making it difficult to comprehensively quantify and correct the uncertainty of model parameters, which affects the reliability of the simulation model in complex application scenarios.

[0023] The parameter optimization process lacks effective screening means and cannot distinguish between high-sensitivity and low-sensitivity parameters, resulting in a significant increase in the optimization calculation amount and reducing the efficiency of parameter correction.

[0024] The existing model calibration methods usually make adjustments based on fixed parameters, ignoring the parameter distribution characteristics brought by randomness, resulting in the calibrated model still unable to effectively reflect the dynamic characteristics of the real circuit.

[0025] The lack of means for dealing with epistemic uncertainty (such as the lack of accurate distribution descriptions for some parameters) leads to insufficient reliability of the model correction results.

[0026] The present invention regards the parameters of the power electronic converter simulation model as probability distributions, comprehensively characterizing the uncertainty of the parameters, including random uncertainty and epistemic uncertainty.

[0027] Use a probability distribution (such as a normal distribution) to accurately describe the mean and variance of device parameters, quantify the random characteristics of model parameters, and thus significantly improve the practical adaptability of the simulation model.

[0028] Through forward uncertainty propagation analysis, systematically evaluate the impact of input parameters on the response of the simulation model, providing data support for sensitivity analysis.

[0029] Combined with reverse uncertainty propagation, adjust the model parameters according to the experimental data feedback, making the simulation model response closer to the actual system performance, and thus effectively reducing the epistemic uncertainty.

[0030] Sensitivity analysis optimizes the model correction efficiency The present invention adopts advanced methods such as Sobol' sensitivity index to quickly identify high-sensitivity parameters with a greater impact on the simulation model response and low-sensitivity parameters with a smaller impact.

[0031] By fixing the low-sensitivity parameters to reduce the computational amount and focusing on correcting the high-sensitivity parameters, the efficiency of correcting the simulation model is greatly improved.

[0032] The present invention iteratively corrects the simulation model parameters, making the simulation response tend to be consistent with the experimental results in statistical characteristics, thereby greatly improving the accuracy of the model.

[0033] By reducing the epistemic uncertainty of the model and quantifying the stochastic uncertainty of the model, the reliability of the simulation model in industrial applications is significantly improved.

[0034] The technical solution of the present invention is applicable to various power electronic converters, including DC-DC converters, inverters, AC-DC rectifiers, etc., and can effectively cope with the influence of different circuit parameter uncertainties on the simulation model.

[0035] By improving the model accuracy and correction efficiency, it can be widely applied in fields such as new energy, electric vehicles, and industrial automation, providing strong support for the optimal design of complex power electronic systems.

[0036] The present invention effectively solves the problems in the prior art such as fixed simulation model parameters, low calibration efficiency, and inability to quantify parameter randomness by comprehensively using uncertainty quantification, forward and reverse uncertainty propagation, and sensitivity analysis. Its significant technological progress is reflected in improving the accuracy of the simulation model, enhancing the consistency between the model and experimental data, and optimizing the correction efficiency, providing more reliable and efficient technical support for the simulation modeling and application of power electronic converters. Brief Description of the Drawings

[0037] Figure 1 It is a flowchart of the digital twin parameter correction method for power electronic converters based on uncertainty quantification provided by an embodiment of the present invention; Figure 2 is a schematic diagram of forward parameter uncertainty propagation provided by an embodiment of the present invention; Figure 3 is a schematic diagram of reverse parameter sensitivity analysis (parameter correction) provided by an embodiment of the present invention; Figure 4 is a structural diagram of a parameter correction system for a power electronic converter simulation model based on uncertainty quantification provided by an embodiment of the present invention; Figure 5 is a physical diagram of a buck converter provided by an embodiment of the present invention; Figure 6 is a circuit topology diagram of a buck converter provided by an embodiment of the present invention; Figure 7 is a discrete circuit topology diagram of a buck converter provided by an embodiment of the present invention; Figure 8 is a schematic diagram of the MH sampling process provided by an embodiment of the present invention; Figure 9 is a schematic diagram of a Markov chain of the traditional MCMC algorithm provided by an embodiment of the present invention; Figure 10 is a schematic diagram of the CDF curve provided by an embodiment of the present invention; Figure 11 is a schematic diagram of parameter uncertainty quantification of the simulation model before correction provided by an embodiment of the present invention; Figure 12 is a schematic diagram of the probability box of the total uncertainty of the simulation model before correction; Figure 13 is a schematic diagram of parameter uncertainty quantification of the simulation model after correction provided by an embodiment of the present invention; Figure 14 is a schematic diagram of the probability box of the total uncertainty of the simulation model after correction provided by an embodiment of the present invention. Detailed implementation manners

[0038] In order to make the objectives, technical solutions and advantages of the present invention clearer and more understandable, the present invention will be further described in detail below with reference to embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0039] As Figure 1As shown in the figure, the embodiment of the present invention provides a method for correcting the parameters of a digital twin of a power electronic converter based on uncertainty quantification. The method for correcting the parameters of a digital twin of a power electronic converter based on uncertainty quantification is a comprehensive technical framework. Based on uncertainty quantification, it sequentially performs parameter uncertainty characterization of the power electronic converter simulation model, parameter sensitivity analysis based on forward uncertainty propagation, and parameter correction based on reverse uncertainty propagation. Under this framework, the parameters of the power electronic converter simulation model are no longer regarded as fixed unknowns, but probability distributions with statistical characteristics such as mean and variance. The present invention aims to make the response of the power electronic converter simulation model consistent with the experimental data, quantify the random uncertainty of the simulation model parameters, and reduce the epistemic uncertainty of the simulation model parameters.

[0040] 1 Parameter uncertainty characterization Uncertainty quantification consists of uncertainty characterization and uncertainty propagation. Before correcting the parameters of the simulation model, the present invention first characterizes the parameter uncertainty of the power electronic converter simulation model.

[0041] According to the different attributes of uncertainty, uncertainty can be divided into random uncertainty and epistemic uncertainty. Random uncertainty represents the randomness existing in nature or physical phenomena, and people cannot control or reduce this kind of randomness. Epistemic uncertainty is caused by insufficient subjective understanding of people, lack of knowledge and data, resulting in the inability to accurately construct a physical model, or the inability to accurately describe the uncertainty of certain factors / parameters with an accurate probability distribution.

[0042] The parameter uncertainty of the power electronic converter simulation model is caused by the fact that the input parameters of the simulation model are not uniquely determined values but a range, and it includes both random uncertainty and epistemic uncertainty. The uncertainty of the input parameters is transmitted through the simulation model, resulting in uncertainty in the simulation model response.

[0043] The present invention uses the form of probability distribution to represent the parameter uncertainty of the power electronic converter simulation model. When purchasing devices such as resistors, inductors, and capacitors, rated values and tolerances are given. According to the law of large numbers and the central limit theorem, if the device samples are sufficient, the distribution of device parameters approaches a normal distribution. Therefore, the present invention sets the device parameters to follow a normal distribution with a mean equal to the rated value of the device parameters and a standard deviation equal to half of the tolerance of the device parameters.

[0044] 2 Parameter sensitivity analysis The parameter uncertainty propagation of the power electronic converter simulation model includes two aspects: (1) Forward parameter uncertainty propagation, such as Figure 2As shown in; (2) Reverse parameter uncertainty propagation. Forward parameter uncertainty propagation is based on the effective characterization of parameter uncertainty. The parameter uncertainty is passed through the simulation model of the power electronic converter, and then its impact on the response of the simulation model is analyzed and the parameter uncertainty of the simulation model is quantified. The present invention uses the cumulative probability density function to quantify the parameter uncertainty of the simulation model.

[0045] Based on the forward propagation of the parameter uncertainty of the power electronic converter simulation model, the present invention conducts input parameter sensitivity analysis to deeply explore the influence degree of different uncertain input parameters on the response of the power electronic converter simulation model. The purpose is to evaluate the sensitivity of the simulation model response to the change of input parameters, find the uncertain input parameters with less influence on the simulation model response, and assign fixed and known values to them, so as to effectively reduce the computational amount of simulation model parameter correction.

[0046] The traditional global sensitivity analysis methods for dealing with random uncertainty can generally be divided into regression-based methods and variance-based methods. Among them, the variance analysis method based on the Sobol' sensitivity index has been widely used due to its simple and effective characteristics.

[0047] 3 Parameter correction As Figure 3 shown, as part of the parameter uncertainty propagation of the power electronic converter simulation model, reverse parameter uncertainty propagation (parameter correction) is based on the test data of the power electronic converter and the initial characterization (prior probability distribution) of the parameter uncertainty of the simulation model, and calculates the best uncertainty characterization (posterior probability distribution) of the simulation model parameters by inverse deduction, so as to reduce the cognitive uncertainty of the simulation model parameters and quantify the random uncertainty of the simulation model parameters.

[0048] As can be seen from the above, the key to the parameter correction of the power electronic converter simulation model lies in calculating the posterior distribution of the simulation model parameters by inverse deduction. As a mature technology, Bayesian inference can solve the complex mathematical model of reverse simulation model parameter uncertainty propagation through forward calculation, and can obtain the posterior distribution of the simulation model parameters under the prior distribution of the simulation model parameters and the given test data.

[0049] Using Bayesian inference for parameter correction of the power electronic converter simulation model can be expressed as: (1); Among them, is the prior distribution of the simulation model parameter , which is obtained by collecting expert information and historical data before obtaining the test data ; is the likelihood function of the simulation model parameter , which can be understood as under the simulation model parameter Under the condition of, the degree of coincidence between the simulation model response and the test data Degree of agreement; is the evidence function, which, as a normalization constant, ensures that the integral of the posterior distribution is always 1. Its value is fixed and independent of the simulation model parameters Therefore, the following proportional relationship can be obtained: (2); As can be seen from Equation (2), the posterior distribution of the simulation model parameters is only implicitly known. When using Bayesian inference to correct the simulation model parameters, the key lies in understanding the relationship between the simulation model parameters and the test data Therefore, the present invention ignores the evidence function for the time being.

[0050] The prior distribution of the simulation model parameters is confirmed before obtaining the test data, and it combines the prior information and initial assumptions of the simulation model parameters. Although theoretically any type of distribution can be used as the prior distribution, in actual operation, the uniform distribution and the normal distribution are the most commonly selected types. According to the parameter uncertainty characterization described above, the present invention knows the mean and standard deviation of the simulation model parameters, so its prior distribution is set as a normal distribution, then The probability density function of the - dimensional parameter to be corrected can be expressed as: (3); where, is the mean of , and is the standard deviation of .

[0051] Assume that are independent of each other, then the prior distribution can be expressed as: (4); When constructing the likelihood function, two elements must be covered: the difference between the test data and the simulation model response, and the simulation model parameters. When obtaining n N - dimensional test data it is considered that each dimension of the test data is independent of each other and each dimension follows the same distribution, then the likelihood function can be expressed as follows: (5); Assume that the error between the test data and the simulation model response result follows a zero - mean normal distribution with a mean of zero and a fixed variance, then the likelihood function follows the following normal distribution: (6); wherein, represents the standard deviation of the error between the j - th dimensional test data and the model output result; of; represents the j - th dimensional model output result.

[0052] The Markov Chain Monte Carlo (MCMC) algorithm is a commonly used algorithm for solving the posterior distribution of Bayesian inference. Its idea is: to obtain a sample sequence, that is, a Markov chain, by sampling the posterior distribution of the simulation model parameters. This chain will eventually converge, and the posterior distribution of the simulation model parameters is estimated using the Markov chain after removing the burn - in period.

[0053] In the MCMC algorithm, the sampling methods that are more widely used are: Metropolis - Hastings (MH) sampling and Gibbs sampling. In Gibbs sampling, it is difficult to select a suitable conditional probability distribution to represent the posterior distribution of the parameter to be corrected for high - dimensional and complex posterior distributions, which limits the scope of application of Gibbs sampling. However, MH sampling can sample from any probability distribution as long as the function is proportional to its normalized probability density function and the value of the function can be calculated.

[0054] As Figure 4 shown, the embodiment of the present invention provides a power electronic converter simulation model parameter correction system based on uncertainty quantification, including: A parameter uncertainty characterization module for characterizing the parameter uncertainty of the power electronic converter simulation model; A parameter sensitivity analysis module for parameter sensitivity analysis based on forward uncertainty propagation; A parameter correction module for parameter correction based on reverse uncertainty propagation.

[0055] Taking the buck converter as an example, the present invention introduces the whole process of correcting the parameter uncertainty of the power electronic converter model. The buck converter is as Figure 5 shown. The operating frequency of this buck converter is 200 kHz. When the steady - state inductor current in the closed - loop control is 4 A, the peak - to - peak value of the inductor current and the output voltage are selected as the steady - state response quantities of the buck converter; the response time (rise time) when the inductor current steps from 1 A to 4 A and the response time (fall time) when the inductor current steps from 4 A to 1 A in the closed - loop control are selected as the transient response quantities of the buck converter.

[0056] 1 Converter Modeling The circuit topology of the BUCK converter is as Figure 6 shown. The present invention selects the node voltage method to construct the BUCK mathematical model, which is divided into the following three steps: (1) Model the switching devices in the circuit using the binary resistance method. When the switching device is on, it is equivalent to a small resistance, and when the switching device is off, it is equivalent to a large resistance.

[0057] (2) Discretize the continuous model of the non-linear devices in the circuit using the implicit Euler method. The discretization results of the inductor and capacitor are as follows: (7); Among them, is the simulation time step, are the inductor currents at the th and th steps respectively, is the inductor voltage at the th step, is the capacitor current at the th step, are the capacitor voltages at the th and th steps respectively, and: (8); Obtain Figure 7 the discrete circuit topology diagram of the buck converter.

[0058] (3) Establish the node voltage equation of the discrete circuit using the node voltage method. The node voltage equation is expressed as follows: (9); Among them, represents the admittance matrix, as shown in Equation (10); represents the node voltage vector, as shown in Equation (11); represents the vector of the current sources injected into the nodes, as shown in Equation (12).

[0059] (10); Among them, is the equivalent conductance of the switch , is the equivalent conductance of the switch , is the equivalent conductance of the input resistance , is the equivalent conductance of the input capacitor , is the equivalent conductance of the capacitor , is the equivalent conductance of the inductor , is the equivalent conductance of the resistor , is the equivalent conductance of the output capacitor ​ is the output resistance of the equivalent conductance, is the load resistance of the equivalent conductance.

[0060] (11); (12); Among them, is the input voltage, are all equivalent model parameters of inductors and capacitors.

[0061] The node voltage is obtained by solving Equation (9) and is used to update the entire network.

[0062] 2 Parameter Sensitivity Analysis The present invention uses the variance analysis method based on the Sobol' sensitivity index to perform parameter sensitivity analysis on the simulation model of the power electronic converter. This method expands the variance of the simulation model response through the simulation model of the power electronic converter, and then uses the ratio of the local variance to the total variance of the simulation model response as the Sobol' sensitivity index, so as to quantify the contribution of each uncertainty parameter to the global sensitivity of the simulation model response and evaluate its relative importance to the uncertainty influence of the simulation model response. Monte Carlo simulation is a common method for calculating the Sobol' sensitivity index at present. This method is carried out in three steps: the first step is to characterize the uncertainty of the parameters of the power electronic converter simulation model in the form of a probability distribution; the second step is to generate a set of parameter samples with the distribution of the uncertainty parameters of the simulation model; the third step is to input the parameter samples into the power electronic converter simulation model for simulation, and then calculate the variance of the simulation model response, and finally obtain the Sobol' sensitivity index.

[0063] The present invention briefly introduces the Sobol' sensitivity index. For a dimensional uncertainty parameter of the power electronic converter simulation model response , it can be expressed as: (13); Among them, represents the mean value of the system response y, represents the simulation model response related to the parameters and .

[0064] (14); Among them, is the domain of definition; is Joint probability density function. If the simulation model response is square integrable, the uncertainty parameters are independent of each other, and the mean of each term in expansion (14) is 0, then all terms in expansion (13) are pairwise orthogonal to each other, and thus each term can be uniquely determined.

[0065] Simulation model response The total variance can be expressed as: (15); Furthermore, according to equation (13) and the orthogonality between its components, the total variance D of the simulation model response can be decomposed into the sum of variances of each term: (16); Among them, the variance of each term can be expressed as follows: (17); In equation (16), represents the influence of the single parameter in the corresponding dimension on the simulation model response ; represents the influence of the interaction between multi-parameters on the simulation model response ; represents the influence of the interaction between multi-parameters on the simulation model response ; represents the influence of the interaction between multi-parameters on the simulation model response ;

[0066] It can be seen from equations (13) and (16) that when the input parameters are independent of each other, the variance decomposition formula can reflect the structure of the simulation model of the power electronic converter. Moreover, the variance of the simulation model response can be decomposed into: variance terms related to a single input parameter, variance terms related to two input parameters, and variance terms related to multiple input parameters, so that the uncertainty (variance) of the simulation model response can be allocated to each input parameter and the interaction terms between different input parameters.

[0067] Through equations (16) and (17), the total variance of the simulation model response and the local variances when each simulation model uncertainty parameter acts alone or interactively can be obtained. Furthermore, the influence degree of each simulation model uncertainty parameter on the simulation model response can be quantified by the ratio of the local variance to the total variance, that is, the Sobol' global sensitivity index: (18); Among them, are the uncertain parameters of the simulation model The global sensitivity index of the individual action. The global sensitivity index of the uncertain parameters of the simulation model can be used as a key index to judge the sensitivity of the response of the power electronic converter simulation model to the change of the uncertain parameters.

[0068] The present invention uses the variance analysis method based on the Sobol’ sensitivity index for the inductor of the BUCK converter , load resistance and output capacitance to perform sensitivity analysis. The parameters of the inductor, resistor and capacitor selected by the present invention are shown in Table 1. According to the parameter uncertainty characterization described above, the present invention sets the inductor, resistor and capacitor to follow a normal distribution, and the distribution parameters are given below: (19); (20); (21); Table 1 Main device parameters Components Code number Rating Tolerance Inductance L 15 15% Output capacitance <![CDATA[C out > 242 20% Load resistance <![CDATA[R load > 2.5 Ω 5% In the process of sampling the parameter distribution, a key issue is the sampling error. The more sampling groups, the smaller the sampling error, and the sample group can better represent the parameter probability distribution, but the model execution requires a certain amount of time. After weighing the sampling error and time cost, it is decided to extract 400 groups of samples each time. Table 2 gives the sensitivity calculated by extracting 400 groups each time. It can be seen that the uncertainty of the output capacitance C out has little impact on the responses of the four simulation models, while the inductor L and load resistance R load have a greater impact on the responses of the four simulation models. Therefore, it is concluded that the inductor L and load resistance R load should be corrected.

[0069] Table 2 Results of parameter sensitivity analysis Sensitivity index Peak-to-peak inductor current Output voltage Rise time Fall time <![CDATA S L > 0.238460692 0.002641453 0.162603572 0.092090224 <![CDATA S R > 0.002865478 0.248218852 0.167870139 0.185932278 <![CDATA S C > 7.36716E-09 1.54267E-10 9.16689E-06 2.21501E-05 <![CDATA S LR > 0.250428747 0.244766835 0.176450132 0.184493712 <![CDATA S LC > 0.267255999 0.002107876 0.146028017 0.098062024 <![CDATA S RC > 0.002785802 0.252367927 0.174542059 0.178423477 <![CDATA S LRC > 0.238203275 0.249897057 0.172496914 0.260976135 3 Model parameter uncertainty correction The present invention uses the MCMC algorithm to solve the posterior distribution of the simulation model parameters. The present invention uses the Metropolis-Hastings (MH) sampling method to obtain the Markov chain. This method is a random walk algorithm, which generates candidate samples of the Markov chain through the proposal distribution and the acceptance or rejection of the candidate samples is determined by the acceptance probability function. The definition of the acceptance function is shown below: where (22); Among them, Denote the candidate sample The estimated value on the posterior distribution Denote the current sample The estimated value on the posterior distribution Denote that under the condition of the candidate sample The probability of obtaining the current sample Denote that under the condition of the current sample The probability of obtaining the candidate sample

[0070] In order to reduce the computational amount, the present invention suggests that the proposed distribution be a symmetric distribution, then there is , so formula (22) can be simplified to: (23); Substituting formula (1) into formula (23) gives: (24); Wherein Is the value of the candidate sample On the likelihood function Is the value of the current sample On the likelihood function Is the value of the candidate sample On the prior distribution Is the value of the current sample on the prior distribution

[0071] The present invention selects the normal distribution, a symmetric distribution, as the proposed distribution for generating candidate samples, whose mean is the current sample , and the standard deviation is used as the adjustment parameter for MH sampling, which can be adjusted accordingly according to specific situations. The standard deviation of the proposed distribution affects the sampling efficiency of the MCMC algorithm. If the standard deviation of the proposed distribution is small, it will lead to a small difference between adjacent samples in the Markov chain, so more iteration times are required to achieve convergence. If the standard deviation is large, the generated candidate samples may exceed the range of the posterior distribution, thus unable to effectively search the entire posterior distribution space, which will lead to a high rejection rate of candidate samples. Therefore, setting the standard deviation of the proposed distribution too large or too small will result in low sampling efficiency.

[0072] Figure 8 Shows the MH sampling process. Combining with formula (23), it can be seen that when the candidate sample Generated by the proposed distribution Moves towards the high probability density direction of the posterior distribution, it is always accepted. Otherwise, the candidate sample Is substituted into formula (23) to calculate the acceptance probability , and then compared with the uniform distribution​​​ Generated random value Compare, if , then accept the candidate sample , if , then reject the candidate sample .

[0073] The present invention uses 300 inductors with the same parameters and 300 resistors implemented by electronic loads with a given distribution to access the physical buck converter, measures the test data, and uses the test data to verify the method for correcting the simulation model of the power electronic converter proposed above.

[0074] After 914 iterations, the Markov chain of the traditional MCMC algorithm as shown in Figure 9 is obtained. It can be seen that the Markov chain gradually converges after about 300 iterations.

[0075] The present invention uses the form of the cumulative probability density function to quantify the parameter uncertainty of the simulation model of the power electronic converter, that is, according to the distribution of the input parameters, a suitable sampling method is selected to sample the input parameters to obtain several groups of different input parameter combinations. Then, each group of input parameters is input into the simulation model for simulation to obtain the response of the simulation model. Then, the cumulative probability density function of the response of the simulation model is obtained. At this time, the parameter uncertainty can be represented by the cumulative probability density function. Figure 10 The cumulative probability density function curves drawn from the simulation model responses of the test data, prior distribution, and posterior distribution are given. It can be seen that the method proposed by the present invention effectively corrects the simulation model parameters, making the simulation model response more consistent with the test data.

[0076] Quantification of the total uncertainty before and after correction of the parameters of the 4 BUCK converter simulation model The parameter uncertainty of the simulation model of the power electronic converter can be measured in the form of the cumulative probability density function, and the form uncertainty of the simulation model of the power electronic converter can be measured by the improved area measurement verification method. In order to comprehensively obtain the total uncertainty of the simulation model by combining the parameter uncertainty of the simulation model and the form uncertainty of the simulation model, the present invention adopts the form of the probability box. In the probability box, the parameter uncertainty of the simulation model is characterized in the form of a probability distribution, and the form uncertainty of the simulation model is characterized in the form of an interval.

[0077] (1) Before correction: Figure 11 The CDF curve for quantifying the parameter uncertainty of the simulation model before correction is given.

[0078] According to the improved area measurement verification method, the of the four groups of output responses of the peak-to-peak value of the inductor current, the output voltage value, the rise time of the inductor current from 1A to 4A, and the fall time of the inductor current from 4A to 1A are obtained. and As shown in Table 3

[0079] Table 3 Before correction and calculate Output response <![CDATA[d - > <![CDATA d + > Steady state Peak-to-peak inductor current / A 0.0241 0.0323 Steady state Output voltage / V 2.1221 0 Transient Rise time / s 5.11051e-3 0 Transient Fall time / s 8.9025e-3 0 The safety factor of the present invention is 1.74295. Further, the form uncertainties of the simulation models before correction for the four groups of response quantities, namely the peak-to-peak value of the inductor current, the output voltage value, the rise time of the inductor current from 1 A to 4 A, and the fall time of the inductor current from 4 A to 1 A, are shown in Table 4

[0080] Table 4 Form uncertainties of the simulation model before correction Target response quantity <![CDATA[Model form uncertainty F(x) - F S d - , F(x) + F S d + > Steady state Peak-to-peak inductor current / A [-0.042005，+0.056297] Steady state Output voltage / V [-3.698714，] Transient Rise time / s [-8.9074-3，] Transient Fall time / s [-15.5166-3，] The simulation total uncertainty is represented by integrating the simulation uncertainties of the two forms in the form of a probability box into one graph. The probability boxes of the simulation total uncertainties before correction for the four system response quantities are as Figure 12 shown

[0081] (2) After correction: Figure 13 The CDF curve for quantifying the parameter uncertainty of the corrected simulation model is given

[0082] Table 5 gives the corrected and calculate Output response <![CDATA d - > <![CDATA d + > Steady state Peak-to-peak inductor current / A 0.00879 0.0132 Steady state Output voltage / V 0.04039 0.0184 Transient Rise time / s 3.0759e-5 0.0013 Transient Fall time / s 6e-4 0 The safety factor of the present invention is 1.74295. Further, the form uncertainties of the simulation models after correction for the four groups of response quantities, namely the peak-to-peak value of the inductor current, the output voltage value, the rise time of the inductor current from 1 A to 4 A, and the fall time of the inductor current from 4 A to 1 A, are shown in Table 6

[0083] Table 6 Form uncertainties of the simulation model after correction Target response quantity <![CDATA[Model form uncertainty F(x) - F S d - , F(x) + F S d + <!-- 12 -->]]> ​ Steady state Peak-to-peak inductor current / A [-0.015321，+0.023007] Steady state Output voltage / V [-0.070398，0.03207] Transient Rise time / s [-5.3611-5，0.002266] Transient Fall time / s [-1.04577-3，] The simulation total uncertainty is represented by integrating the simulation uncertainties of the two forms in the form of a probability box into one graph. The probability boxes of the simulation total uncertainties after correction for the four system response quantities are as Figure 14 shown

[0084] Table 7 gives the comparison of the simulation total uncertainties before and after parameter correction at a 95.45% confidence level. It can be seen that after applying the method for correcting the parameters of the power electronic converter simulation model described in the present invention to the BUCK converter simulation model, the simulation total uncertainty is significantly reduced

[0085] Table 7 Comparison of the total simulation uncertainties before and after parameter correction at a 95.45% confidence level Target response quantity Total uncertainty before correction Total uncertainty after correction Steady state Peak-to-peak inductor current / A 1.97923-1.42957=0.54966 1.83479-1.55513=0.27966 Steady state Output voltage / V 10.3941-5.89555=4.49855 8.50652-7.36426=1.14226 Transient Rise time / s 0.0399775-0.0279551=0.0120225 0.0359635-0.0303789=0.0055846 Transient Fall time / s 0.0434275-0.0245159=0.0189116 0.0358775-0.0304617=0.0054158 It should be noted that the embodiments of the present invention can be implemented by hardware, software, or a combination of software and hardware. The hardware part can be implemented using dedicated logic; the software part can be stored in a memory and executed by a suitable instruction execution system, such as a microprocessor or dedicated designed hardware. Those of ordinary skill in the art can understand that the above devices and methods can be implemented using computer-executable instructions and / or included in processor control code, for example, such code is provided on a carrier medium such as a disk, CD, or DVD-ROM, a programmable memory such as a read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. The devices and their modules of the present invention can be implemented by hardware circuits of programmable hardware devices such as very large scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, etc., or field programmable gate arrays, programmable logic devices, etc., can also be implemented by software executed by various types of processors, or can be implemented by a combination of the above hardware circuits and software such as firmware.

[0086] As described above, the above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention, any modifications, equivalent replacements, and improvements made within the spirit and principle of the present invention shall be covered by the protection scope of the present invention.

Claims

1. A method for correcting parameters of digital twins of power electronic converters based on uncertainty quantification, characterized in that: The following steps are involved: Step 1, parameter uncertainty characterization of power electronic converter simulation model; Step 2: Parameter sensitivity analysis based on forward uncertainty propagation; Step 3: Parameter correction based on reverse uncertainty propagation.

2. The method for correcting parameters of a digital twin of a power electronic converter based on uncertainty quantification according to claim 1, characterized in that: In step 1, the parameter uncertainty characterization of the power electronic converter simulation model uses the form of probability distribution to represent the parameter uncertainty of the power electronic converter simulation model; the device parameters are set to obey a normal distribution with a mean of the device parameter rated value and a standard deviation of half the device parameter tolerance.

3. The method for correcting parameters of a digital twin of a power electronic converter based on uncertainty quantification according to claim 1, characterized in that: In step 2, based on forward parameter uncertainty propagation, the influence of different uncertain input parameters on the response of the power electronic converter simulation model is explored to evaluate the sensitivity of the simulation model response to changes in input parameters, find the uncertain input parameters that have less impact on the simulation model response, and assign them fixed and known values, thereby effectively reducing the computational complexity of simulation model parameter correction.

4. The method for correcting parameters of a digital twin of a power electronic converter based on uncertainty quantification according to claim 3, characterized in that: On the basis of effective characterization of parameter uncertainty, the parameter uncertainty is transferred through the power electronic converter simulation model, and then its influence on the simulation model response is analyzed and the simulation model parameter uncertainty is quantified.

5. The method for correcting parameters of a digital twin of a power electronic converter based on uncertainty quantification according to claim 1, characterized in that: In step 2, the parameter sensitivity analysis based on forward uncertainty propagation uses the form of a cumulative probability density function curve to quantify the parameter uncertainty of the power electronic converter simulation model, that is, according to the distribution of the input parameters, a suitable sampling method is selected to sample the input parameters to obtain several groups of different input parameter combinations, and then each group of input parameters is input into the simulation model to simulate the simulation model response, and then the simulation model response is plotted as a cumulative probability density function curve. At this time, the parameter uncertainty can be represented by the cumulative probability density function curve.

6. The method for correcting parameters of a digital twin of a power electronic converter based on uncertainty quantification according to claim 1, characterized in that: The parameter correction based on reverse uncertainty propagation in step three is based on the initial characterization of the uncertainty of the power electronic converter test data and the simulation model parameters, that is, the prior probability distribution, and reversely calculates the optimal uncertainty characterization of the simulation model parameters, that is, the posterior probability distribution; during the reverse calculation, Bayesian reasoning forward calculation is used to solve the problem of the optimal uncertainty characterization of the power electronic converter simulation model parameters.

7. A power electronic converter digital twin parameter correction system based on uncertainty quantification according to any one of claims 1 to 6, characterized in that: include: Parameter uncertainty characterization module, parameter uncertainty characterization of power electronic converter simulation model; Parameter sensitivity analysis module, parameter sensitivity analysis based on forward uncertainty propagation; Parameter correction module, parameter correction based on reverse uncertainty propagation.

8. A computer device, characterized in that: The computer device includes a memory and a processor, the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the method for correcting parameters of a digital twin of a power electronic converter based on uncertainty quantification as described in any one of claims 1 to 6.

9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program. When the computer program is executed by a processor, the processor executes the steps of the method for correcting parameters of a digital twin of a power electronic converter based on uncertainty quantification as described in any one of claims 1 to 6.

10. An information data processing terminal, characterized in that: The information data processing terminal includes the power electronic converter simulation model parameter correction system based on uncertainty quantification as described in claim 7.

Images