Prediction Method for Conduction Interference Test Curve Interval of Mass Production Equipment

Abstract

The present invention relates to the technical field of interval prediction, and particularly to an interval prediction method for the conducted interference test curve of mass production equipment, including: selecting some devices to be tested to conduct conducted interference tests in the frequency band to be tested to obtain test curves, dividing the prediction frequency band and the test frequency band based on the test curves, establishing a prediction data set according to the test data of the devices to be tested that have undergone conducted interference tests and industrial big data, and establishing an interval prediction model for predicting the conducted interference curve of the remaining devices to be tested in the prediction frequency band; inputting the test data of the test frequency band into the interval prediction model to predict the conducted interference curve of the prediction frequency band; evaluating the interval prediction model based on the prediction results; and correcting the division criteria of the prediction frequency band and the test frequency band based on the evaluation results, realizing the selective prediction of the conducted interference of mass-produced equipment, thereby reducing the test cost while ensuring the accuracy of the prediction results.

Description

Technical Field

[0001] The present invention relates to the technical field of interval prediction, and particularly to an interval prediction method for the conducted interference test curve of mass production equipment. Background Art

[0002] For mass-produced equipment, it is often necessary to test its electromagnetic compatibility (such as sensitivity and conducted interference). The characteristics of electromagnetism are usually complex and variable. Factors such as the type, location, signal strength, and frequency of the interference source may cause the conducted interference test curve to exhibit uncertain patterns and variations, usually with non-linear relationships and complex dependencies. Existing mathematical models are still unable to accurately capture the trends and variation laws of these interferences. In the process of capturing interference trends and laws, the quality and quantity of feature data extraction are crucial for obtaining accurate analysis results. Currently, for single-feature data analysis, statistical analysis methods are mostly used to obtain feature data, such as mean, variance, standard deviation, kurtosis, skewness, maximum value, minimum value, etc. However, such feature data extracted based on basic statistics and mathematical methods is usually not applicable to complex sequence data. The extracted features are static and difficult to capture the dynamic feature changes in the sequence. At the same time, when these features are applied to machine learning or deep learning models, it is often difficult to fully mine the important information in the data, and in the process of artificially selecting features, some key information will inevitably be lost. Summary of the Invention

[0003] The purpose of the present invention is to provide an interval prediction method for the conducted interference test curve of mass production equipment, so as to realize the selective prediction of the conducted interference of mass-produced equipment, and thus reduce the test cost while ensuring the accuracy of the prediction results.

[0004] To this end, the present invention provides an interval prediction method for the conducted interference test curve of mass production equipment, and the interval prediction method for the conducted interference test curve of mass production equipment includes:

[0005] Selecting some devices to be tested in a single batch in the frequency band to be tested to obtain a test curve through conducted interference testing;

[0006] Dividing the prediction frequency band and the test frequency band in the frequency band to be tested based on the test curves of the devices to be tested that have undergone conducted interference testing;

[0007] Establishing a prediction data set according to the test data of the devices to be tested that have undergone conducted interference testing and industrial big data;

[0008] Performing sequence feature decomposition on the prediction data set through moving average decomposition and SSA decomposition to obtain the trend term, seasonal term, short-period term, and noise term of each sequence in the prediction data set;

[0009] Perform stationarity tests on the seasonal, short-period, and noise terms of each sequence;

[0010] Based on the results of the stationarity tests, input the sequences obtained by the above-mentioned sequence feature decomposition into the corresponding models for training respectively, to obtain an interval prediction model for predicting the conducted interference curve of the remaining devices to be measured in the prediction frequency band;

[0011] For each of the remaining devices to be measured, input the test data in the test frequency band into the interval prediction model to predict the conducted interference curve in the prediction frequency band;

[0012] Evaluate the interval prediction model based on the prediction results, the interval coverage rate of the prediction results, and the interval average width;

[0013] Modify the division criteria for the prediction frequency band and the test frequency band based on the evaluation results.

[0014] As a preferred technical solution of the method for predicting the interval of the conducted interference test curve of mass-produced devices, the division of the prediction frequency band and the test frequency band specifically includes:

[0015] Extract the part where the test values in the frequency band to be measured are continuously higher than the first threshold as the test frequency band;

[0016] Divide the remaining frequency band to be measured except the test frequency band into several sub-frequency bands;

[0017] For the devices to be measured that have undergone conducted interference tests, count the number of times each sub-frequency band exceeds the limit value, and use the sub-frequency band with the proportion of the number of times exceeding the limit value greater than the second threshold as the test frequency band;

[0018] Use the frequency band to be measured except the test frequency band as the prediction frequency band;

[0019] Among them, the first threshold changes dynamically with the test curve of the devices to be measured that have undergone conducted interference tests, so that the proportion of the test frequency band in the frequency band to be measured is a preset ratio.

[0020] As a preferred technical solution of the method for predicting the interval of the conducted interference test curve of mass-produced devices, the establishment of the prediction data set according to the test data of the devices to be measured that have undergone conducted interference tests and industrial big data includes:

[0021] Obtain the historical conducted interference test data of the devices with the same power supply method and signal conduction path as the device to be measured used cumulatively as the prediction data;

[0022] Obtain the historical conducted interference test data of the devices with the same oscillation frequency and impedance characteristics as the device to be measured as the prediction data;

[0023] Obtain the historical conducted interference test data of the devices of the same model as the device to be measured as the prediction data;

[0024] Integrate the above prediction data into the prediction data set.

[0025] As a preferred technical solution of the conduction interference test curve interval prediction method for mass production equipment, the sequence feature decomposition extraction of the prediction data set by moving average decomposition and SSA decomposition includes:

[0026] Extract the trend term of each sequence in the prediction data set by moving average decomposition;

[0027] Based on the remaining terms of each sequence in the prediction data set except the trend term, extract them by SSA decomposition to obtain the seasonal term, short-period term and noise term.

[0028] As a preferred technical solution of the conduction interference test curve interval prediction method for mass production equipment, the stationarity test specifically includes:

[0029] Perform ADF test and KPSS test on the seasonal term, short-period term and noise term of the sequence respectively;

[0030] If both the ADF test and the KPSS test determine that the sequence is stationary, determine that the sequence is a stationary sequence, otherwise it is a non-stationary sequence.

[0031] As a preferred technical solution of the conduction interference test curve interval prediction method for mass production equipment, the specific process of inputting the sequences after the sequence feature decomposition into the corresponding models for training based on the stationarity test results includes:

[0032] For the sequences extracted as trend terms and the sequences whose stationarity tests for the seasonal term, short-period term and noise term are all stationary sequences, input them into the ARIMA model for prediction.

[0033] As a preferred technical solution of the conduction interference test curve interval prediction method for mass production equipment, the process of inputting the sequences after the sequence feature decomposition into the corresponding models for training based on the stationarity test results also includes: For the sequences whose number of times of being determined as stationary sequences in the stationarity tests for the seasonal term, short-period term and noise term is 2, input them into the LSTM-Bootstrap model for prediction.

[0034] As a preferred technical solution of the conduction interference test curve interval prediction method for mass production equipment, for the sequences whose number of times of being determined as stationary sequences in the stationarity tests for the seasonal term, short-period term and noise term is less than or equal to 1, input them into the CN-LSTM-Bootstrap model for prediction.

[0035] As a preferred technical solution of the method for predicting the conduction interference test curve interval of mass production equipment, the evaluation of the interval prediction model based on the prediction result, the interval coverage rate of the prediction result, and the interval average width includes:

[0036] The evaluation index f1 for the interval coverage rate is obtained by the following formula:

[0037]

[0038] The evaluation index f2 for the interval average width is obtained by the following formula:

[0039] f2 = u i - s i ,

[0040] The final evaluation index f is obtained by the following formula i As the evaluation result,

[0041] f i = k1·f1 + k2·f2,

[0042] where β is the penalty weight factor, u i and s i represent the upper and lower bounds of the prediction range respectively, y i is the predicted value, k1 and k2 are the consideration weights of the interval coverage rate and the interval average width respectively, and k1 gradually decreases with the running process of the interval prediction model, and k2 gradually increases with the running process of the interval prediction model.

[0043] As a preferred technical solution of the method for predicting the conduction interference test curve interval of mass production equipment, the correction of the division standard of the predicted frequency band and the test frequency band based on the evaluation result specifically includes:

[0044] Compare the evaluation index f i with the standard index f0. If f i is greater than f0, increase the preset ratio based on the difference between the evaluation index f i and the standard index f0, and the increase amount is positively correlated with the difference.

[0045] The beneficial effects of the present invention are:

[0046] The present invention is a method for automatically extracting meaningful features or attributes from a prediction dataset, and these features can effectively describe the core dynamic characteristics of the data. By introducing machine learning and deep learning models, appropriate methods are used for separate prediction according to the characteristics of different components. During the separate prediction process, through joint learning and training, the potential correlation between components is utilized to further improve the accuracy and efficiency of data analysis and model prediction. Through the frequency band division of prediction and testing, the division of prediction models, the evaluation of interval prediction models, and the correction of prediction weights based on the evaluation results, the full-process adaptive control of the conducted interference characteristics of mass-produced equipment is achieved. Each step in the solution is closely related, realizing the closed-loop analysis control of prediction and testing, and having good reliability in use.

[0047] Furthermore, the present invention can predict the complete detection curve and its change trend in the conducted interference test item, thereby determining the specific range of test points under different confidence levels, and evaluating whether the product performance and test results are within the allowable thresholds.

[0048] Furthermore, the present invention can significantly reduce the test volume required for electromagnetic compatibility test items in product R & D and production. By using partial test data to replace the complex electromagnetic model simulation process, the complete test curve of the product is estimated in advance, helping engineers quickly identify the frequency bands or key parameters that need to be focused on, contributing to the early discovery of potential problems, improving the efficiency of design verification, reducing the trial-and-error cost, and shortening the development cycle.

[0049] Furthermore, through the iterative feedback of test data, the model is continuously optimized. As this method is improved and verified in static test scenarios, it can also be migrated to more complex application scenarios to achieve real-time monitoring of equipment electromagnetic interference, capture the electromagnetic laws during the operation of the equipment, and thus provide stronger guarantees for the safety and performance stability of the equipment.

[0050] Furthermore, the present invention divides the prediction frequency band and the test frequency band in the frequency band to be measured through partial tests. Considering that the accuracy of prediction is lower than that of actual testing, prediction is more applied to the frequency bands to be measured with relatively less influence. And the present invention designs an evaluation step for the model, and adjusts the division criteria of the prediction frequency band and the test frequency band according to the evaluation results, making the prediction more in line with the actual scenario and avoiding the impact of prediction inaccuracy on equipment quality control.

[0051] Furthermore, in the evaluation of the model of the present invention, an evaluation method designed for the characteristics of electromagnetic compatibility is adopted. Based on the concern about the upper limit value exceeding the conduction interference test threshold in the conduction interference test curve of electromagnetic compatibility, a phased evaluation index is designed. Considering that in electromagnetic compatibility testing, the standard mainly defines the qualified range through a threshold curve, during the test, more attention is paid to whether the test value exceeds the upper threshold. The obtained evaluation index is more affected by the prediction result of the upper limit part, and the obtained evaluation result is more applicable. Accordingly, the correction of the division standard based on the evaluation result is more applicable. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 It is a flowchart of the method for predicting the conduction interference test curve interval of the mass production equipment in the embodiment of the present invention;

[0053] Figure 2 It is a flowchart of the work in the model training stage in the embodiment of the present invention;

[0054] Figure 3 It is a flowchart of the construction of the CN-LSTM-Bootstrap prediction model in the embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0055] Next, the technical solutions of the present invention will be described clearly and completely with reference to the accompanying drawings. Apparently, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0056] In the description of the present invention, it should be noted that the terms "first", "second" are only used for descriptive purposes and cannot be understood as indicating or implying relative importance.

[0057] In the description of the present invention, it should be noted that unless otherwise clearly defined and limited, the terms "installed", "connected", "connected" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be directly connected or indirectly connected through an intermediate medium, and it can be the communication inside two components. For those of ordinary skill in the art, the specific meanings of the above terms in the present invention can be understood according to specific situations.

[0058] The embodiments of the present invention will be described in detail below. The examples of the embodiments are shown in the accompanying drawings, where the same or similar reference numerals represent the same or similar elements or elements with the same or similar functions from beginning to end. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention and cannot be understood as a limitation of the present invention.

[0059] Refer to Figure 1 and Figure 2 As shown, this embodiment provides a method for predicting the conduction interference test curve interval of mass production equipment, including the following steps:

[0060] Step S1, select some devices to be tested in a single batch for conduction interference testing in the frequency band to be tested to obtain a test curve;

[0061] Step S2, divide the predicted frequency band and the test frequency band in the frequency band to be tested based on the test curves of the devices to be tested that have undergone conduction interference testing;

[0062] Step S3, establish a prediction data set according to the test data of the devices to be tested that have undergone conduction interference testing and industrial big data;

[0063] Step S4, perform sequence feature decomposition on the prediction data set through moving average decomposition and SSA decomposition to obtain the trend term, seasonal term, short-period term, and noise term of each sequence in the prediction data set;

[0064] Step S5, perform stationarity tests on the seasonal term, short-period term, and noise term of each sequence;

[0065] Step S6, based on the stationarity test results, input the sequences after sequence feature decomposition into the corresponding models for training to obtain an interval prediction model for predicting the conduction interference curve of the remaining devices to be tested in the predicted frequency band;

[0066] Step S7, for each of the remaining devices to be tested, input the test data in the test frequency band into the interval prediction model to predict the conduction interference curve in the predicted frequency band;

[0067] Step S8, evaluate the interval prediction model based on the prediction results, the interval coverage rate of the prediction results, and the interval average width;

[0068] Step S9, correct the division criteria for the predicted frequency band and the test frequency band based on the evaluation results.

[0069] In the above embodiments, by means of automatically extracting meaningful features or attributes from the prediction dataset, these features can effectively describe the core dynamic characteristics of the data. By introducing machine learning and deep learning models, appropriate methods are used for separate prediction according to the characteristics of different components. During the separate prediction process, through joint learning and training, the potential correlation between components is utilized to further improve the accuracy and efficiency of data analysis and model prediction. Through the frequency band division of prediction and testing, the division of prediction models, the evaluation of the interval prediction model, and the repair of the prediction model based on the evaluation results, the full-process adaptive control of the conducted interference characteristics of mass-produced devices is achieved. Each step in the solution is closely related, realizing the closed-loop analysis control of prediction and testing, and having good reliability in use.

[0070] Furthermore, the prediction method of this embodiment can predict the complete detection curve and its change trend in the conducted interference test item, thereby determining the specific range of test points under different confidence levels, and evaluating whether the product performance and test results are within the allowable thresholds.

[0071] Based on the above effects, the prediction method of this embodiment can significantly reduce the test volume required for electromagnetic compatibility test items in product R & D and production. By using partial test data to replace the complex electromagnetic model simulation process, the complete test curve of the product is estimated in advance, which helps engineers quickly identify the frequency bands or key parameters that need to be focused on, contributes to the early discovery of potential problems, improves the efficiency of design verification, reduces the trial-and-error cost, and shortens the development cycle. Furthermore, through the iterative feedback of test data, the model is continuously optimized. As this method is improved and verified in static test scenarios, it can also be migrated to more complex application scenarios to achieve real-time monitoring of equipment electromagnetic interference, capture the electromagnetic laws during the operation of the equipment, and thus provide stronger guarantees for the safety and performance stability of the equipment.

[0072] Moreover, in this embodiment, the prediction frequency band and the test frequency band in the frequency band to be measured are divided through partial tests. Considering that the accuracy of prediction is lower than that of actual testing, prediction is more applied to the frequency bands to be measured with relatively less influence. And for the model, an evaluation step is designed, and the division criteria for the prediction frequency band and the test frequency band are adjusted according to the evaluation results, making the prediction more in line with the actual scenario and avoiding the impact of prediction inaccuracy on equipment quality control.

[0073] Specifically, the division of the prediction frequency band and the test frequency band specifically includes:

[0074] Step S21, extracting the part where the test values (unit: dBμA or dBμV) in the frequency band to be measured are continuously higher than the first threshold as the test frequency band;

[0075] Step S22: Evenly divide the remaining frequency bands to be measured except the test frequency band into several sub - frequency bands;

[0076] Step S23: For the device under test that has undergone conducted interference testing, count the number of times each sub - frequency band exceeds the limit value. Take the sub - frequency bands whose proportion of the number of times exceeding the limit value (i.e., the number of times the sub - frequency band exceeds the limit / the number of tests) is greater than the second threshold as the test frequency band;

[0077] Step S24: Take the frequency bands to be measured except the test frequency band as the prediction frequency band;

[0078] Among them, the first threshold changes dynamically with the test curve of the device under test that has undergone conducted interference testing, so that the proportion of the test frequency band in the frequency bands to be measured is a preset ratio. Specifically, the dynamic change of the first threshold enables the controllability of the ratio between the test frequency band and the prediction frequency band. Maintaining a fixed ratio between the two can make the operation effect of the model and the prediction accuracy controllable. For the setting of the second threshold, the purpose is to test the frequency bands that are prone to problems and avoid ignoring them in the prediction. In implementation, there is a situation where the test frequency band determined by the second threshold has exceeded the preset ratio. For this situation, the control of the ratio by the first threshold is not carried out, and the test frequency band determined by the second threshold is directly tested, and the remaining frequency bands are predicted. The optional value of the second threshold is 0.05, and it can be adjusted in combination with the scenario in implementation.

[0079] Specifically, establishing a prediction data set based on the test data of the device under test that has undergone conducted interference testing and industrial big data includes:

[0080] Step S31: Obtain the historical conducted interference test data of the devices with the same power supply method and signal conduction path as the device under test that have been used cumulatively as prediction data;

[0081] Step S32: Obtain the historical conducted interference test data of the devices with the same oscillation frequency and impedance characteristics as the device under test as prediction data;

[0082] Step S33: Obtain the historical conducted interference test data of the devices of the same model as the device under test as prediction data;

[0083] Step S34: Integrate the above - mentioned prediction data into a prediction data set.

[0084] Furthermore, the overall trend of the test curve is mainly affected by factors such as the filter circuit, parasitic inductance and capacitance, and cables of the product. Before predicting the harmonic terms and coupling terms, the most crucial thing is to accurately extract and grasp the change of the overall trend. Other spike and fluctuation components are more superposed in the trend term. Therefore, sequence feature decomposition extraction is performed on the prediction data set through moving average decomposition and SSA decomposition, including:

[0085] Step S41: Extract the trend terms of each sequence in the prediction dataset through moving average decomposition;

[0086] Step S42: Based on the remaining terms of each sequence in the prediction dataset excluding the trend terms, extract them through SSA decomposition to obtain seasonal terms, short-cycle terms, and noise terms. Specifically, moving average decomposition, as a smoothing technique, can effectively extract the overall trend from time series data. By calculating the average value of the data within a moving window, it can reduce the influence of short-term fluctuations, thereby highlighting the long-term trend more prominently. The STL decomposition is used to obtain and improve the stability of the trend terms, which reflects the long-term change trend of the sequence and is the core influencing factor for the future trend of the sequence. It will be directly input into the ARIMA model for training and prediction.

[0087] Specifically, the stationarity test specifically includes:

[0088] Step S51: Conduct ADF tests and KPSS tests on the seasonal terms, short-cycle terms, and noise terms of the sequence respectively;

[0089] Step S52: If both the ADF test and the KPSS test determine that the sequence is stationary, determine that the sequence is a stationary sequence; otherwise, it is a non-stationary sequence.

[0090] Specifically, based on the stationarity test results, inputting the sequences after sequence feature decomposition into the corresponding models for training specifically includes:

[0091] For the sequences extracted as trend terms and the sequences whose stationarity tests for seasonal terms, short-cycle terms, and noise terms are all stationary sequences, input them into the ARIMA model for prediction.

[0092] Specifically, based on the stationarity test results, inputting the sequences after sequence feature decomposition into the corresponding models for training also includes: For the sequences whose number of times determined to be stationary sequences in the stationarity tests for seasonal terms, short-cycle terms, and noise terms is 2, input them into the LSTM-Bootstrap model for prediction.

[0093] Specifically, the LSTM (Long Short-Term Memory) model is a variant of the Recurrent Neural Network (RNN), specifically designed to address the dependency issues of long sequence data. The structure of LSTM is based on the cell state and multiple gating mechanisms to manage the information flow, thus maintaining long-term dependencies in sequence modeling. In this embodiment, the input_features of the LSTM model used is 60, i.e., the number of input features at each time step, the number of model layers is 4, and finally a fully connected layer is added to transform the output of LSTM into the target shape 1, i.e., single-step prediction. The complete conducted interference test curves in the historical tests of the same type of products are used as training data. After including quadratic decomposition and stationarity tests, they are input into the corresponding model to complete the model training.

[0094] Specifically, for sequences whose number of times judged as stationary sequences in the stationarity tests of seasonal terms, short-period terms, and noise terms is less than or equal to 1, they are input into the CN-LSTM-Bootstrap model for prediction. Specifically, the CN-LSTM-Bootstrap prediction model is similar to the LSTM-Bootstrap model in interval prediction, but introduces the similarity influence matrix of Cross-stitch Networks to jointly interact among multiple models. The construction flow chart is as Figure 3 shown:

[0095] Taking two sequences as an example, two LSTM models with the same basic structure as in the previous part are used to process sequences X and Y respectively. Although their variation rules and characteristics are different, they jointly affect the final performance of the sequences. Especially in the electromagnetic compatibility test scenario, the same interference source will show various forms in the test results through different coupling methods and propagation paths. Ignoring the potential associations between them will lead to the distortion of the prediction of non-stationary partial sequences. To solve this problem, the CN-LSTM model, through the joint training mechanism, can not only fully capture the independent features of each sequence, but also reveal the dependencies and interactions between sequences. Through this collaborative learning, the model is more coordinated when processing the prediction of seasonal terms, short-period terms, and noise terms, thus significantly improving the prediction accuracy of the overall time series. Specifically in implementation, the CN-LSTM introduces an interaction mechanism into the weight matrix of the output gate of each LSTM model. The degree of interaction is expressed by the similarity metric formula. The output gate weights of the two LSTM models are W1 and W2 respectively. According to the Similarity Influence Matrix (SIM), they interact with each other, then there is

[0096]

[0097] W1′ and W2′ are the output gate weight matrices after interactive update, and a in the SIM matrix 11 and a 22 represent the degree to which each model retains its own information, while a 12 and a 21 represent the degree of mutual influence between the two models. Replacing them with similarity metrics gives:

[0098]

[0099] Then W1′ and W2′ are updated as follows:

[0100]

[0101] The calculation method of the index p(l) mentioned in the above formula is as follows:

[0102] First, the numerical sequence is symbolized. Each data point is simplified into three basic categories according to its change state: "increase", "decrease", or "remain unchanged". This process aims to simplify the complexity of the numerical sequence, focus on the formal change relationship between local sequences, and thus capture the trend and fluctuation of the sequence more effectively. The specific operation is as follows: By comparing the size relationship between each data point and its previous data point, each data point is classified into one of the three states: "increase", "decrease", or "remain unchanged". Further, based on the magnitude of the change, each state is refined into two cases: "sharp change" and "gentle change". Through the above processing, the numerical values in each sequence are transformed into the following five states, converting them into clear structured data, where the parameter δ is used to distinguish sharp changes from gentle changes.

[0103]

[0104] Among them, the threshold parameter δ is used to distinguish the severity and smoothness of the spectrum change. The value of this parameter is determined based on the difference between the amplitude of each frequency point in the complete test curve and the standard threshold. The smaller the gap, the more sensitive the product is to the conducted interference test at this frequency point, and the greater the potential risk. Therefore, a smaller threshold should be set to more sensitively capture the changes and make more stringent judgments. Considering that the frequency points with a large amplitude gap usually represent relatively safe areas, these frequency points do not participate in the determination of the parameter δ. Finally, the calculation of the parameter δ is only based on those frequency points whose amplitude difference from the standard threshold is less than one-tenth of it. The amplitudes of these frequency points are summed and averaged to obtain the final threshold parameter. A s 、A m 、U、D s 、D m represent sharp increase, gentle increase, remain unchanged, sharp decrease, and gentle decrease respectively. x i -x i-1Represents the change between two adjacent data points in a numerical sequence. Specifically, it is the change in the current data point x i With the previous data point x i-1 The difference.

[0105] After discretizing the sequences, the edit distance method in biological genetics is used to measure the similarity between sequences. The similarity of the change patterns between two sequences is studied from the perspective of macroscopic change patterns. The edit distance can quantify their similarity by calculating the minimum number of operations required between sequences (such as insertion, deletion or substitution). After completing the macroscopic pattern analysis, the covariance coefficient Cov(X,Y) of the two sequences is further calculated to quantitatively judge the similarity of the numerical change degree between the sequences. The covariance coefficient is used to measure the synchronous changes of the two sequences, that is, their numerical correlation. Combining the two similarity measures of edit distance and covariance coefficient can comprehensively judge the overall similarity between sequences. On this basis, as the depth of the model increases, the sharing ratio is exponentially decayed, and multiplied by the decay coefficient to finally obtain the sharing ratio index p(l) of the two training models:

[0106]

[0107] Where d(X,Y) is the edit distance between sequences, n is the sequence length, l is the depth of the model, and α is the similarity weight ratio. When it is larger, it pays more attention to the changing trend of the sequence, and when it is smaller, it pays more attention to the quantitative change relationship of the sequence value. λ is the decay rate, and α and λ participate in the iterative training of the model. σ X and σ Y are the standard deviations of series X and Y, respectively.

[0108] In each layer of LSTM, the weight of the output gate is interactively adjusted through the similarity influence matrix, so that the two LSTM models can not only train their own sequences independently, but also learn the information of each other's sequences through the interaction of the output gate weights.

[0109] As the number of network layers increases, the mutual influence between models gradually weakens. By designing a reasonable attenuation mechanism, the interaction between high-level layers can be gradually reduced. This design can ensure that the model fully utilizes the correlation information between sequences at a lower level, while maintaining a certain degree of independence at a high level, so that different non-stationary sequences can be predicted more accurately. The introduction of this similarity influence matrix and the layer-by-layer attenuation mechanism balances the information sharing and independence in the multi-sequence prediction task, reduces the prediction error caused by the interference of irrelevant information between sequences, and improves the stability and overall performance of model training.

[0110] In the above embodiments, the core objective is to maximize and utilize the advantages of different models by further decomposing the sequence to extract its features. This method aims to improve the accuracy and reliability of prediction. The ARIMA model (Autoregressive Integrated Moving Average model) is suitable for dealing with stationary sequences. However, if it is directly applied to non-stationary sequences, it may lead to overfitting due to the model's overly complex fitting of the data, thus reducing the prediction performance. On the other hand, the LSTM model (Long Short-Term Memory network), as a powerful non-linear model, is good at capturing complex patterns, long-term dependencies, and dynamic changes in sequences. For the electromagnetic performance testing of highly integrated electronic products, due to the presence of complex factors such as harmonics, oscillators, clock signals, differential and common-mode interferences, it often shows reciprocating fluctuations with different densities in different frequency bands. Therefore, in order to improve the purity of the features and refine the remaining signal components, the sequence needs to be decomposed twice.

[0111] In the above embodiments, a stationarity test is performed on the three sequences obtained from the previous decomposition. In this embodiment, a consistency test is used, combining the ADF test (Augmented Dickey-Fuller) and the KPSS test (Kwiatkowski-Phillips-Schmidt-Shin). When the test results of both indicate that the sequence is stationary, the sequence can be determined to be a stationary sequence. The stationary sequence is input into the ARIMA model for prediction, while the non-stationary sequence selects the LSTM-Bootstrap model to fully utilize its advantages in capturing non-linearity and long-term dependencies. If there are two or more non-stationary sequences, these sequences will be input into the CN-LSTM-Bootstrap model for joint training to enhance the model's learning ability of the correlations between sequences and further improve the accuracy and efficiency of prediction.

[0112] For the evaluation of the model, the evaluation of the model performance in interval prediction usually combines two key indicators: Prediction Interval Coverage Probability (PICP) and Mean Prediction Interval Width (MPIW). The interval coverage rate is used to ensure the reliability of the model and ensure that the prediction interval can contain the true value; while the mean prediction interval width reflects the accuracy of the model, and the narrower the interval, the higher the accuracy. A good interval prediction evaluation criterion must take both of these indicators into account, and neither can be missing. However, there is generally a competitive and adversarial relationship between these two indicators. Therefore, in this embodiment, in view of the specific requirements of electromagnetic compatibility testing, especially the concern about the upper limit value exceeding the conduction interference test threshold, a phased evaluation index is designed. The evaluation of the interval prediction model based on the prediction result, the interval coverage rate of the prediction result, and the mean prediction interval width includes:

[0113] Step S81, obtain the evaluation index f1 for the interval coverage rate through the following formula

[0114]

[0115] Step S82, obtain the evaluation index f2 for the mean prediction interval width through the following formula

[0116] f2 = u i - s i ,

[0117] Step S83, obtain the final evaluation index f through the following formula i As the evaluation result,

[0118] f i = k1·f1 + k2·f2,

[0119] where β is the penalty weight factor, u i and s i represent the upper and lower bounds of the prediction range respectively, and y iis the predicted value, k1 and k2 are the consideration weights for the interval coverage rate and the average interval width respectively (the specific values and the increase and decrease ranges are set in combination with the actual scenario), and k1 gradually decreases with the running process of the interval prediction model, while k2 gradually increases with the running process of the interval prediction model. In the early stage, emphasis is placed on ensuring the interval coverage rate to ensure the reliability of the model; in the later stage, the consideration of the average interval width is gradually increased to optimize the accuracy of the model, so as to achieve higher accuracy in the upper limit prediction in electromagnetic compatibility testing. Specifically, in the formula, f1 and f2 consider the interval coverage rate and the average interval width respectively. f1 first calculates the difference between the predicted value and the midpoint of the interval. If the predicted point does not fall within the interval, an additional penalty will be imposed in the second part. β is the penalty weight factor. Considering that the national and industrial standards in electromagnetic compatibility testing mainly define the qualified range through a threshold curve, during the test, more attention is paid to whether the test value exceeds the upper limit threshold. In this embodiment, an exponential function is used to distinguish the situations of exceeding the upper and lower limits of the interval. Through this phased consideration evaluation mechanism, the contradiction between the model reliability and accuracy can be effectively balanced, meeting the dual requirements of high reliability and precision in electromagnetic compatibility testing.

[0120] Specifically, the correction of the division standard for the predicted frequency band and the test frequency band based on the evaluation results specifically includes:

[0121] Compare the evaluation index f i with the standard index f0,

[0122] If f i is greater than f0, increase the preset ratio based on the difference between the evaluation index f i and the standard index f0, and the increase amount is positively correlated with the difference. If f i is less than or equal to f0, maintain the original preset ratio or restore it to the initial value of the preset ratio.

[0123] Optionally, the correction process specifically includes:

[0124] If f i is greater than f0, increase the preset ratio R0 to the corrected preset ratio R through the following formula, and the preset ratio R shall not exceed the upper limit R min and the lower limit R max ,

[0125]

[0126] R = min(max(R, R min ), R max ),

[0127] Specifically, the larger the evaluation index is, the less accurate the prediction result is. At this time, reducing the prediction frequency band can reduce the impact on the conducted interference analysis. As the training data gradually increases, the interval prediction model can also restore the accuracy of the prediction result. After the restoration, the preset ratio can be restored, that is, restored to the initial value. Optionally, the initial value of the preset ratio is 0.4.

[0128] Obviously, the above embodiments of the present invention are merely examples for clearly explaining the present invention, rather than limiting the implementation manners of the present invention. For those of ordinary skill in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to enumerate all implementation manners here. Any modifications, equivalent replacements, and improvements made within the spirit and principle of the present invention shall be included in the protection scope of the claims of the present invention.

Claims

1. A method for predicting the interval of a conducted interference test curve of a mass-produced device, characterized in that: include: Select some devices under test from a single batch to conduct conducted interference test in the frequency band under test to obtain the test curve; Divide the frequency band to be tested into a predicted frequency band and a test frequency band based on the test curve of the device to be tested that has been tested for conducted interference; Establish a prediction data set based on the test data of the equipment under test that has been tested for conducted interference and industrial big data; Decomposing the predicted data set into sequence features by sliding average decomposition and SSA decomposition to obtain trend items, seasonal items, short-term items and noise items of each sequence in the predicted data set; Perform stationarity tests on the seasonal term, short-cycle term and noise term of each series; Based on the stationarity test result, the sequences decomposed by the sequence features are respectively input into the corresponding models for training, so as to obtain an interval prediction model for predicting the conducted interference curve of the predicted frequency band of the remaining equipment to be tested; For each of the remaining devices to be tested, the test data of the test frequency band is input into the interval prediction model to predict the conducted interference curve of the predicted frequency band; Evaluating the interval prediction model based on the prediction results, the interval coverage of the prediction results, and the average width of the intervals; The division criteria of the predicted frequency band and the test frequency band are modified based on the evaluation result.

2. The method for predicting the interval of the conducted interference test curve of mass-produced equipment according to claim 1, characterized in that: The division of the predicted frequency band and the test frequency band specifically includes: Extracting a portion of the frequency band to be tested whose test value is continuously higher than the first threshold as the test frequency band; Dividing the remaining frequency bands to be tested except the test frequency band into a plurality of sub-frequency bands; For the device under test that has been subjected to the conducted interference test, the number of times each sub-frequency band exceeds the limit is counted, and the sub-frequency band whose proportion of times exceeding the limit is greater than the second threshold is used as the test frequency band; The frequency band to be tested except the test frequency band is used as the predicted frequency band; The first threshold value changes dynamically with the test curve of the device under test that has been subjected to the conducted interference test, so that the proportion of the test frequency band in the frequency band to be tested is a preset ratio.

3. The method for predicting the interval of the conducted interference test curve of mass production equipment according to claim 1, characterized in that: The establishment of a prediction data set based on the test data of the device under test that has been tested for conducted interference and the industrial big data includes: Obtain historical conducted interference test data of devices that use the same power supply mode and signal conduction path as the device under test as prediction data; Obtain historical conducted interference test data of a device having the same oscillation frequency and impedance characteristics as the device under test as prediction data; Obtain historical conducted interference test data of the same model of equipment as the equipment to be tested as prediction data; The above prediction data are integrated into the prediction data set.

4. The method for predicting the interval of the conducted interference test curve of mass production equipment according to claim 1, characterized in that: The extracting of sequence features from the prediction data set by sliding average decomposition and SSA decomposition includes: Extracting trend items of each sequence in the forecast data set by sliding average decomposition; For each sequence in the forecast data set, the remaining terms except the trend term are extracted based on SSA decomposition to obtain seasonal terms, short-term terms and noise terms.

5. The method for predicting the interval of the conducted interference test curve of mass production equipment according to claim 4, characterized in that: The stability test specifically includes: Perform ADF test and KPSS test on the seasonal term, short cycle term and noise term of the sequence respectively; If both the ADF test and the KPSS test determine that the sequence is stationary, the sequence is considered to be a stationary sequence, otherwise it is a non-stationary sequence.

6. The method for predicting the interval of the conducted interference test curve of mass production equipment according to claim 5, characterized in that: The step of inputting the sequences decomposed by the sequence features into corresponding models for training based on the stationarity test results specifically includes: For the sequences extracted as trend items, and the sequences whose stationarity tests of seasonal items, short-cycle items and noise items are all stationary sequences, they are input into the ARIMA model for prediction.

7. The method for predicting the interval of the conducted interference test curve of mass-produced equipment according to claim 6, characterized in that: The step of inputting the sequences decomposed by the sequence features into corresponding models for training based on the stationarity test results also includes: for the sequences that are determined to be stationary sequences with a frequency of 2 in the stationarity tests of seasonal terms, short-cycle terms and noise terms, inputting them into the LSTM-Bootstrap model for prediction.

8. The method for predicting the interval of the conducted interference test curve of mass-produced equipment according to claim 7, characterized in that: For the sequences that are judged to be stationary sequences with a frequency less than or equal to 1 in the stationarity test of seasonal terms, short-cycle terms, and noise terms, they are input into the CN-LSTM-Bootstrap model for prediction.

9. The method for predicting the interval of the conducted interference test curve of mass production equipment according to claim 2, characterized in that: The evaluating the interval prediction model based on the prediction results, the interval coverage rate of the prediction results and the average width of the intervals includes: The evaluation index f1 for the interval coverage is obtained by the following formula: The evaluation index f2 for the average width of the interval is obtained by the following formula: f2=u i -s i , The final evaluation index f is obtained by the following formula: i As a result of the evaluation, <h2 style=";text-align:left;direction:ltr">f<h2 style=";text-align:left;direction:ltr"> i <h2 style=";text-align:left;direction:ltr"> (k1 f1 + k2 f2) Among them, β is the penalty weight factor, u i and i Represent the upper and lower bounds of the prediction range, y i is the predicted value, k1 and k2 are the consideration weights of interval coverage and interval average width respectively, and k1 decreases gradually with the running progress of the interval prediction model, and k2 increases gradually with the running progress of the interval prediction model.

10. The method for predicting the interval of the conducted interference test curve of mass-produced equipment according to claim 9, characterized in that: The modifying of the division standard of the predicted frequency band and the test frequency band based on the evaluation result specifically includes: The evaluation index f i Compared with the standard index f0, If f i is greater than f0, the preset ratio is based on the evaluation index f i The difference with the standard index f0 increases, and the increase is positively correlated with the difference.

Images