Capacitor life prediction method based on ARIMA model and related equipment

Abstract

The invention provides a capacitor life prediction method based on an ARIMA model and related equipment. The method comprises the following steps: acquiring historical capacitance data of a target capacitor through an experiment; performing unit root test on the historical capacitance data to obtain a stationary sequence; performing white noise test on the stationary sequence; performing parameter fitting on the constructed differential integration moving average autoregression model, and performing residual white noise test on the fitted differential integration moving average autoregression model to obtain a tested differential integration moving average autoregression model; inputting the stationary sequence after the white noise test into the differential integration moving average autoregression model after the test for prediction, and obtaining a life prediction result of the target capacitor; compared with the prior art, the method has the advantages that the trend change in the data is captured through the constructed differential integration moving average autoregression model, the residual life of the capacitor can be predicted more accurately, the reliability of the capacitor is improved, and the service life of the capacitor is prolonged.

Description

Technical Field

[0001] The present invention relates to the technical field of electronic component aging prediction, and in particular to a capacitor life prediction method based on an ARIMA model and related equipment. Background Art

[0002] Capacitors play a vital role in modern electronic devices. Their main functions include energy storage, filtering, and signal conditioning. As electronic devices develop towards high performance and high reliability, the requirements for the performance and life of capacitors are getting higher and higher. The life prediction of capacitors has a wide range of application value in the fields of industry, automotive electronics, consumer electronics, etc. For example, in industrial control systems, capacitor failure may cause the entire system to shut down, causing huge economic losses; in automotive electronic systems, capacitor failure may affect the safety performance of the vehicle.

[0003] At present, the life prediction methods of capacitors mainly include physical model-based methods and data-driven methods. The physical model-based method predicts the life by establishing a physical degradation model of the capacitor and combining environmental and working condition parameters. This method requires a deep understanding of the physical degradation mechanism of the capacitor and usually requires a large amount of experimental data to calibrate the model. The data-driven method uses historical data to establish a life prediction model through statistical and machine learning methods. With the development of data acquisition and storage technology, data-driven life prediction methods are gaining more and more attention.

[0004] Among the data-driven life prediction methods, time series analysis methods are favored because they can handle data that changes over time. The ARIMA (Auto-Regressive Integrated Moving Average) model is a widely used method in time series analysis. It combines the autoregressive model (AR) and the moving average model (MA), and makes non-stationary time series data stationary through difference operations (I). The ARIMA model can effectively capture trend changes in time series data and is suitable for time series prediction tasks in a variety of fields.

[0005] At present, the ARIMA model has been widely used in financial market forecasting, meteorological data forecasting, and industrial equipment life forecasting. For example, in financial market forecasting, the ARIMA model is used to predict stock prices and trading volumes; in meteorological data forecasting, the ARIMA model is used to predict temperature and rainfall; in industrial equipment life forecasting, the ARIMA model is used to predict equipment failure time and remaining life. However, research on capacitor life prediction is still relatively limited, and further exploration and improvement are needed to improve the accuracy and reliability of predictions. Summary of the invention

[0006] The invention provides a capacitor life prediction method based on an ARIMA model and related equipment, the purpose of which is to improve the accuracy and reliability of life prediction.

[0007] In order to achieve the above object, the present invention provides a capacitor life prediction method based on the ARIMA model, comprising:

[0008] Step 1, collecting historical capacitance data of the target capacitor through experiments;

[0009] Step 2, perform a unit root test on the historical capacitance data to obtain a stationary series;

[0010] Step 3, perform white noise test on the stationary sequence;

[0011] Step 4, performing parameter fitting on the constructed differential integrated moving average autoregressive model, and performing residual white noise test on the fitted differential integrated moving average autoregressive model to obtain a tested differential integrated moving average autoregressive model;

[0012] Step 5, input the stationary sequence after white noise test into the tested differential integrated moving average autoregressive model for prediction, and obtain the life prediction result of the target capacitor.

[0013] More specifically, step 1 includes:

[0014] Place the target capacitor in the test box;

[0015] The historical capacitance data of the target capacitor is collected by the sensor, and the historical capacitance data is used to characterize the historical capacitance value trend.

[0016] Furthermore, before step 2, it also includes:

[0017] The historical capacitance data is preprocessed to obtain preprocessed historical capacitance data.

[0018] Specifically, the historical capacitance data is preprocessed to obtain the preprocessed historical capacitance data, including:

[0019] Processing missing values ​​according to the number of missing values ​​in historical capacitance data;

[0020] When the number of missing values ​​is less than the preset threshold, all missing values ​​are deleted;

[0021] When the number of missing values ​​is greater than or equal to a preset threshold, the missing values ​​are filled using the Lagrange interpolation method to obtain the filled historical capacitance data;

[0022] Performing high-order nonlinear fitting on the historical capacitance data after filling to obtain the fitted historical capacitance data;

[0023] Use the Z-score method to detect the fitted historical capacitance data, obtain the outliers and delete them;

[0024] The moving sliding window method is used to set the window size and threshold to fill the historical capacitance data with abnormal values ​​deleted, so as to obtain the preprocessed historical capacitance data.

[0025] More specifically, step 2 includes:

[0026] The unit root test is used to test the stability of the preprocessed historical capacitance data;

[0027] When the stability value is less than 0.05, the preprocessed historical capacitance data is a stable sequence;

[0028] When the stationary value is greater than 0.05, the preprocessed historical capacitance data is a non-stationary sequence, and a differential operation is performed on the preprocessed historical capacitance data to obtain a stationary sequence.

[0029] More specifically, step 3 includes:

[0030] Calculate the Q statistic. The calculation expression of the Q statistic Q(m) is:

[0031]

[0032] Where T represents the amount of data in the historical capacitance data, represents the correlation coefficient of the historical capacitance data at the first-order lag, and m represents the degree of freedom;

[0033] The Q statistic is used to test the white noise of the stationary series.

[0034] Furthermore, the constructed differential integrated moving average autoregressive model is fitted with parameters, including:

[0035] Draw autocorrelation and partial correlation graphs of non-white noise sequences;

[0036] Observe the lag term at the cutoff of the non-white noise sequence in the autocorrelation curve, and set the order of the autoregressive model to the order of the lag term at the cutoff;

[0037] Observe the lag term at the truncation of the non-white noise series in the partial correlation curve plot, and set the moving average model order to the order of the lag term at the truncation.

[0038] The present invention also provides a capacitor life prediction device based on the ARIMA model, comprising:

[0039] An acquisition module, used for acquiring historical capacitance data of a target capacitor through experiments;

[0040] The first test module is used to perform a unit root test on the historical capacitance data to obtain a stable sequence;

[0041] The second test module is used to perform white noise test on the stationary sequence;

[0042] A fitting module is used to perform parameter fitting on the constructed differential integrated moving average autoregressive model, and to perform a residual white noise test on the fitted differential integrated moving average autoregressive model to obtain a tested differential integrated moving average autoregressive model;

[0043] The prediction module is used to input the stationary sequence after white noise test into the differential integrated moving average autoregressive model after test to predict and obtain the life prediction result of the target capacitor.

[0044] The present invention also provides a terminal device, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, a capacitor life prediction method based on the ARIMA model is implemented.

[0045] The present invention also provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, a capacitor life prediction method based on an ARIMA model is implemented.

[0046] The above scheme of the present invention has the following beneficial effects:

[0047] The present invention collects historical capacitance data of a target capacitor through an experiment; performs a unit root test on the historical capacitance data to obtain a stationary sequence; performs a white noise test on the stationary sequence; performs parameter fitting on a constructed differential integrated moving average autoregressive model, and performs a residual white noise test on the fitted differential integrated moving average autoregressive model to obtain a tested differential integrated moving average autoregressive model; inputs the stationary sequence that has passed the white noise test into the tested differential integrated moving average autoregressive model for prediction, and obtains a life prediction result of the target capacitor; compared with the prior art, the present invention can more accurately predict the remaining life of the capacitor by capturing trend changes in the data through the constructed differential integrated moving average autoregressive model, thereby improving the reliability and service life of the capacitor.

[0048] Other beneficial effects of the present invention will be described in detail in the subsequent specific implementation section. BRIEF DESCRIPTION OF THE DRAWINGS

[0049] Figure 1 A schematic diagram of a flow chart of an embodiment of the present invention;

[0050] Figure 2 Schematic diagram of the structure of a capacitor life prediction device in an embodiment of the present invention;

[0051] Figure 3 Schematic diagram of the structure of a terminal device in an embodiment of the present invention. DETAILED DESCRIPTION

[0052] In order to make the technical problems, technical solutions and advantages to be solved by the present invention clearer, the following will be described in detail with reference to the accompanying drawings and specific embodiments. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0053] In the description of the present invention, it should be noted that the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc., indicating the orientation or positional relationship, are based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, the terms "first", "second", and "third" are used for descriptive purposes only, and cannot be understood as indicating or implying relative importance.

[0054] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "installation", "connection" and "connection" should be understood in a broad sense, for example, it can be a locking connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0055] In addition, the technical features involved in the different embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0056] In view of the existing problems, the present invention provides a capacitor life prediction method based on an ARIMA model and related equipment.

[0057] like Figure 1 As shown, an embodiment of the present invention provides a capacitor life prediction method based on an ARIMA model, comprising:

[0058] Step 1, collecting historical capacitance data of the target capacitor through experiments;

[0059] Step 2, perform a unit root test on the historical capacitance data to obtain a stationary series;

[0060] Step 3, perform white noise test on the stationary sequence;

[0061] Step 4, performing parameter fitting on the constructed differential integrated moving average autoregressive model, and performing residual white noise test on the fitted differential integrated moving average autoregressive model to obtain a tested differential integrated moving average autoregressive model;

[0062] Step 5, input the stationary sequence after white noise test into the tested differential integrated moving average autoregressive model for prediction, and obtain the life prediction result of the target capacitor.

[0063] It should be noted that the differential integrated moving average autoregressive model is the ARIMA (Autoregressive Integrated Moving Average model) model.

[0064] Specifically, step 1 includes:

[0065] Place the target capacitor in the test box;

[0066] The historical capacitance data of the target capacitor is collected by the sensor, and the historical capacitance data is used to characterize the historical capacitance value trend.

[0067] Specifically, before step 2, it also includes:

[0068] The historical capacitance data is preprocessed to obtain preprocessed historical capacitance data.

[0069] Specifically, the historical capacitance data is preprocessed to obtain the preprocessed historical capacitance data, including:

[0070] Processing missing values ​​according to the number of missing values ​​in historical capacitance data;

[0071] When the number of missing values ​​is less than the preset threshold, all missing values ​​are deleted;

[0072] When the number of missing values ​​is greater than or equal to a preset threshold, the missing values ​​are filled using the Lagrange interpolation method to obtain the filled historical capacitance data;

[0073] Performing high-order nonlinear fitting on the historical capacitance data after filling to obtain the fitted historical capacitance data;

[0074] Since the capacitance data is a non-stationary, non-normally distributed time series, and the capacitance data has an obvious trend, the Z-score method is used to detect the fitted historical capacitance data, obtain the outliers and delete them;

[0075] The moving sliding window method is used to set the window size and threshold to fill the historical capacitance data with abnormal values ​​deleted, so as to obtain the preprocessed historical capacitance data.

[0076] In the embodiment of the present invention, the Z-score method is a standardization method, also known as a normalization method, which standardizes data based on the mean and standard deviation of the original data.

[0077] Specifically, step 2 includes:

[0078] The unit root test is used to test the stability of the preprocessed historical capacitance data;

[0079] When the stability value is less than 0.05, the preprocessed historical capacitance data is a stable sequence;

[0080] When the stationary value is greater than 0.05, the preprocessed historical capacitance data is a non-stationary sequence, and a differential operation is performed on the preprocessed historical capacitance data to obtain a stationary sequence.

[0081] In the embodiment of the present invention, the unit root test refers to testing whether there is a unit root in the historical capacitance data, because the existence of a unit root is a non-stationary time series, and the unit root can usually be expressed as a random walk process, and the expression is:

[0082] y t =y t-1 +∈ t

[0083] Among them, ∈ t It is white noise. Since the value at each time point is the value at the previous time point plus a random error, the historical capacitance data will drift randomly over time and may form an upward or downward trend. This trend causes the mean of the historical capacitance data to change over time and cannot remain stable.

[0084] In an embodiment of the present invention, the Dickey-Fuller test ADF method is used to detect whether the historical capacitance data has a unit, and the specific steps are as follows:

[0085] Set hypothesis: The null hypothesis (H0) is that the historical capacitance data has a unit root, that is, the historical capacitance data is non-stationary;

[0086] Construct a difference integrated moving average autoregressive model;

[0087] Estimate parameters: Estimate the parameter γ in the difference integrated moving average autoregressive model by the least squares method;

[0088] Test statistic calculation: Calculate the autoregressive model order of the ADF test, i.e., the p-value;

[0089] If the p-value corresponding to the unit root test statistic is significantly less than 0.05, the null hypothesis is rejected and the sequence is considered to be a stationary sequence (the mean, variance, and autocorrelation structure are constant over time);

[0090] If the p value is greater than 0.05, it is concluded that the historical capacitance data is a non-stationary sequence. The periodicity and trend of the historical capacitance data can be eliminated by the differential operation method. The mathematical formula of the differential operation is as follows:

[0091] If the historical capacitance data is stable and no differentiation is required, then d = 0, y t =Y t ;

[0092] If the historical capacitance data is stable after a difference, then d = 1, y t =Y t -Y t-1 ;

[0093] If the historical capacitance data is stable after the second difference, d = 2, y t =(Y t -Y t-1 )(Y t-1 -Y t-2 )=Y t -2Y t-1 +Y t-2 ;

[0094] Generally speaking, within the second difference, the historical capacitance data will become a stable sequence. If it is still not stable, the historical capacitance data will be subjected to n times of differential operation until it becomes stable.

[0095] Where d is the difference order, and y is the difference of Y at time t.

[0096] Specifically, step 3 includes:

[0097] Calculate the Q statistic, which is used to determine whether the autocorrelation of the historical capacitance data within the m-order lag range is significant, or whether the historical capacitance data is white noise. The Q statistic obeys the chi-square distribution with m degrees of freedom. The calculation expression of the Q statistic Q(m) is:

[0098]

[0099] Where T represents the amount of data in the historical capacitance data, represents the correlation coefficient of the historical capacitance data at the first-order lag, and m represents the degree of freedom;

[0100] The Q statistic is used to test the white noise of the stationary series.

[0101] If the calculated moving average model order p-value drops below the confidence level of 0.05, it is determined that a significant autoregressive relationship occurs and the sequence is not white noise; if the moving average model order p-value is greater than 0.05, the sequence is white noise, indicating that this sequence is not suitable for analysis and prediction using the ARIMA method.

[0102] Specifically, the constructed differential integrated moving average autoregressive model is fitted with parameters, including:

[0103] Draw the autocorrelation curve (ACF, Auto-correlation Function) and partial correlation curve (PACF, PartialAuto-correlation Function) of the non-white noise sequence;

[0104] Observe the lag term at the truncation of the non-white noise sequence in the autocorrelation curve (i.e., the PACF drops significantly and approaches zero after the kth order), and set the order of the autoregressive model to the order of the lag term at the truncation;

[0105] Observe the lag term at the truncation of the non-white noise sequence in the partial correlation curve (i.e., the ACF drops significantly and approaches zero after the mth order), and set the order of the moving average model to the order of the lag term at the truncation.

[0106] The embodiment of the present invention adopts the information criterion (AIC, Akaike Information Criterion) to select the optimal model. The information criterion is a balance indicator based on the model's fit and complexity (number of parameters). By drawing the autocorrelation graph and partial autocorrelation graph of the stable non-white noise sequence, observing the distribution interval of the graphical distribution estimation related parameters p and q, the model with a smaller AIC value is taken as the optimal model, and the order of the model with a smaller AIC value is taken as the optimal autoregressive model order p and moving average model order q of the differential integrated moving average autoregressive model. The calculation formula of the information criterion is:

[0107] AIC=2k-2ln(L)

[0108] k=p+q+1

[0109] Where k is the number of model parameters and L is the maximum likelihood estimate.

[0110] Specifically, the residual white noise test is performed on the fitted differential integrated moving average autoregressive model to obtain the tested differential integrated moving average autoregressive model, including:

[0111] When constructing an ARIMA model, the residual sequence should be a white noise sequence, that is, it has the properties of a mean of 0, a constant variance, and a zero autocovariance. If the residual sequence is not white noise, it means that the ARIMA model fails to fully capture the characteristics of the time series, and it is necessary to reconsider the model's parameter configuration or introduce other factors for modeling. If it passes the white noise test, it means that the modeling is successful and prediction can be performed.

[0112] Calculate the residual sequence of the ARIMA model. For each time point t, the residual where y t is the actual value, is the predicted value. Referring to the stationarity test steps, use the ADF test to test the stationarity of the residual sequence to ensure that it meets the stationarity requirements;

[0113] The autocorrelation test is performed on the stationary residual sequence by using the autocorrelation function to determine whether there is autocorrelation;

[0114] The specific method is: draw the ACF graph to observe whether the autocorrelation coefficient of the lag order is significant; if the ACF values ​​of some lag orders deviate significantly from zero, it means that there is significant autocorrelation at these lag orders;

[0115] Perform a white noise test on the residual sequence. Refer to the white noise test steps to check whether the sequence satisfies the white noise property. The null hypothesis of this test is that the sequence is white noise. If the p-value is less than the significance level of 0.05, the null hypothesis is rejected and it is considered that the sequence is not white noise.

[0116] If the residual sequence is found to be a white noise sequence through the above steps, a suitable ARIMA model is obtained; if the residual sequence is not white noise, return to the previous step and reselect appropriate p and q values.

[0117] Specifically, step 5 includes:

[0118] The prediction time parameters are input into the optimal differential integrated moving average autoregressive model to predict the historical capacitance data and obtain the life prediction result of the target capacitor.

[0119] The calculation formula of the loss function RMSE of the differential integrated moving average autoregressive model is as follows:

[0120]

[0121] Where n is the number of historical capacitance data used for training, y i is the true value, is the predicted value.

[0122] The embodiment of the present invention collects historical capacitance data of the target capacitor through experiments; performs a unit root test on the historical capacitance data to obtain a stationary sequence; performs a white noise test on the stationary sequence; performs parameter fitting on the constructed differential integrated moving average autoregression model, and performs a residual white noise test on the fitted differential integrated moving average autoregression model to obtain a tested differential integrated moving average autoregression model; inputs the stationary sequence after the white noise test into the tested differential integrated moving average autoregression model for prediction to obtain a life prediction result of the target capacitor; compared with the prior art, the embodiment of the present invention can more accurately predict the remaining life of the capacitor by constructing a differential integrated moving average autoregression model to capture trend changes in the data, thereby improving the reliability and service life of the capacitor.

[0123] like Figure 2 As shown, the embodiment of the present invention further provides a capacitor life prediction device 100 based on the ARIMA model, and the capacitor life prediction device 100 includes:

[0124] The acquisition module 101 is used to acquire historical capacitance data of the target capacitor through experiments;

[0125] A first testing module 102 is used to perform a unit root test on historical capacitance data to obtain a stationary sequence;

[0126] The second testing module 103 is used to perform a white noise test on the stationary sequence;

[0127] The fitting module 104 is used to perform parameter fitting on the constructed differential integrated moving average autoregressive model, and perform residual white noise test on the fitted differential integrated moving average autoregressive model to obtain a tested differential integrated moving average autoregressive model;

[0128] The prediction module 105 is used to input the stationary sequence after the white noise test into the tested differential integrated moving average autoregressive model for prediction, so as to obtain the life prediction result of the target capacitor.

[0129] It should be noted that the information interaction, execution process, etc. between the above-mentioned devices / units are based on the same concept as the method embodiment of the present application. Their specific functions and technical effects can be found in the method embodiment part and will not be repeated here.

[0130] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0131] The embodiment of the present invention also provides a terminal device, such as Figure 3 As shown, the terminal device D10 of this embodiment includes: at least one processor D100 ( Figure 3 Only one processor is shown in the figure), a memory D101, and a computer program D102 stored in the memory D101 and executable on the at least one processor D100, wherein the processor D100 implements the above-mentioned capacitor life prediction method based on the ARIMA model when executing the computer program D102.

[0132] The terminal device D10 may be a computing device such as a desktop computer, a notebook, a PDA, a server, a server cluster, a cloud server, etc. The terminal device may include, but is not limited to, a processor D100 and a memory D101. Those skilled in the art will appreciate that Figure 3 This is only an example of the terminal device D10 and does not constitute a limitation on the terminal device D10. The terminal device D10 may include more or fewer components than shown in the figure, or a combination of certain components, or different components. For example, it may also include input and output devices, network access devices, etc.

[0133] The processor D100 may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor, etc.

[0134] In some embodiments, the memory D101 may be an internal storage unit of the terminal device D10, such as a hard disk or memory of the terminal device D10. In other embodiments, the memory D101 may also be an external storage device of the terminal device D10, such as a plug-in hard disk, a smart memory card (SMC, SmartMedia Card), a secure digital (SD, Secure Digital) card, a flash card (Flash Card), etc. equipped on the terminal device D10. Further, the memory D101 may also include both an internal storage unit of the terminal device D10 and an external storage device. The memory D101 is used to store an operating system, an application program, a boot loader (BootLoader), data and other programs, such as the program code of the computer program, etc. The memory D101 may also be used to temporarily store data that has been output or is to be output.

[0135] It should be noted that the information interaction, execution process, etc. between the above-mentioned devices / units are based on the same concept as the method embodiment of the present application. Their specific functions and technical effects can be found in the method embodiment part and will not be repeated here.

[0136] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0137] An embodiment of the present invention provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, a capacitor life prediction method based on an ARIMA model is implemented.

[0138] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present application implements all or part of the process in the above-mentioned embodiment method, which can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may at least include: any entity or device that can carry the computer program code to the construction device / terminal device, recording medium, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium. For example, a USB flash drive, a mobile hard disk, a disk or an optical disk.

[0139] The above is a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A capacitor life prediction method based on ARIMA model, characterized in that: include: Step 1, collecting historical capacitance data of the target capacitor through experiments; Step 2, performing a unit root test on the historical capacitance data to obtain a stationary sequence; Step 3, performing a white noise test on the stationary sequence; Step 4, performing parameter fitting on the constructed differential integrated moving average autoregressive model, and performing residual white noise test on the fitted differential integrated moving average autoregressive model to obtain a tested differential integrated moving average autoregressive model; Step 5, input the stationary sequence after white noise test into the tested differential integrated moving average autoregressive model for prediction, and obtain the life prediction result of the target capacitor.

2. The capacitor life prediction method based on the ARIMA model according to claim 1 is characterized in that: The step 1 comprises: placing the target capacitor in a test box; The historical capacitance data of the target capacitor is collected by a sensor, and the historical capacitance data is used to characterize the historical capacitance value trend.

3. The capacitor life prediction method based on the ARIMA model according to claim 2 is characterized in that: Before step 2, also include: The historical capacitance data is preprocessed to obtain preprocessed historical capacitance data.

4. The capacitor life prediction method based on the ARIMA model according to claim 3 is characterized in that: Preprocessing the historical capacitance data to obtain preprocessed historical capacitance data includes: Processing the missing values ​​according to the number of missing values ​​in the historical capacitance data; When the number of missing values ​​is less than a preset threshold, all missing values ​​are deleted; When the number of missing values ​​is greater than or equal to a preset threshold, the missing values ​​are filled using a Lagrange interpolation method to obtain filled historical capacitance data; Performing high-order nonlinear fitting on the filled historical capacitance data to obtain fitted historical capacitance data; Using a Z-score method to detect the fitted historical capacitance data, obtain abnormal values ​​and delete the abnormal values; The moving sliding window method is used to set the window size and threshold to fill the historical capacitance data with abnormal values ​​deleted, so as to obtain the preprocessed historical capacitance data.

5. The capacitor life prediction method based on the ARIMA model according to claim 4 is characterized in that: The step 2 comprises: The unit root test is used to test the stability of the preprocessed historical capacitance data; When the stable value is less than 0.05, the preprocessed historical capacitance data is a stable sequence; When the stationary value is greater than 0.05, the preprocessed historical capacitance data is a non-stationary sequence, and a differential operation is performed on the preprocessed historical capacitance data to obtain a stationary sequence.

6. The capacitor life prediction method based on the ARIMA model according to claim 5, characterized in that: The step 3 comprises: Calculate the Q statistic. The calculation expression of the Q statistic Q(m) is: Where T represents the amount of data in the historical capacitance data, represents the correlation coefficient of the historical capacitance data at the first-order lag, and m represents the degree of freedom; The Q statistic is used to perform a white noise test on the stationary sequence.

7. The capacitor life prediction method based on the ARIMA model according to claim 6 is characterized in that: Parameter fitting of the constructed differential integrated moving average autoregressive model includes: Draw autocorrelation and partial correlation graphs of non-white noise sequences; Observe the lag term at the cutoff point of the non-white noise sequence in the autocorrelation curve graph, and set the order of the autoregressive model to the order of the lag term at the cutoff point; The lag term at the truncation point of the non-white noise sequence is observed in the partial correlation curve diagram, and the order of the moving average model is set to the order of the lag term at the truncation point.

8. A capacitor life prediction device based on ARIMA model, characterized in that: include: An acquisition module, used for acquiring historical capacitance data of a target capacitor through experiments; A first testing module is used to perform a unit root test on the historical capacitance data to obtain a stable sequence; A second testing module is used to perform a white noise test on the stationary sequence; A fitting module is used to perform parameter fitting on the constructed differential integrated moving average autoregressive model, and to perform a residual white noise test on the fitted differential integrated moving average autoregressive model to obtain a tested differential integrated moving average autoregressive model; The prediction module is used to input the stationary sequence after white noise test into the differential integrated moving average autoregressive model after test to predict and obtain the life prediction result of the target capacitor.

9. A terminal device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the capacitor life prediction method based on the ARIMA model is implemented as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the capacitor life prediction method based on the ARIMA model is implemented.

Images