Abnormity monitoring and management system for dry-type bushing capacitor core production line

Abstract

The invention relates to the field of production line abnormity monitoring and evaluation, in particular to a dry-type bushing capacitor core production line abnormity monitoring and management system which comprises a data acquisition module, a fault rate evaluation module and a management module. The data acquisition module is used for acquiring working information of the winding device in real time; the fault rate evaluation module is used for predicting whether the winding device breaks down or not according to the working information; the management module is used for adjusting the maintenance time of the winding device according to the fault prediction result of the winding device; the working information of the winding device is analyzed by using the neural network model, the complex non-linear relation and time sequence dependence between parameters can be automatically learned, time sequence data is processed through a sliding window, the model can dynamically reflect the evolution trend of the equipment state, the method is suitable for predicting sudden faults, and the fault prediction efficiency is improved. Compared with a common fault prediction method, higher fault prediction accuracy can be achieved.

Description

Technical Field

[0001] The present invention relates to the field of production line abnormality monitoring and evaluation, and in particular to an abnormality monitoring and management system for a dry-type bushing capacitor core production line. Background Art

[0002] The winding device is a critical component in capacitor core production lines. Common maintenance methods rely on fixed-cycle inspections, which fail to detect changes in the device's status in real time and make it difficult to respond to sudden failures of the winding device. Furthermore, existing fault prediction methods only analyze individual winding device systems (such as support rollers, pressure rollers, and air circuits) independently, without considering the interactions between multiple systems. For example:

[0003] 1. Fluctuations in air line pressure may cause cylinder compensation lag, thus affecting the pressure stability of the roller;

[0004] 2. The combined effect of support roller vibration and speed difference may aggravate the unevenness of aluminum foil winding;

[0005] Therefore, single parameter analysis cannot accurately predict the occurrence of faults, leading to misjudgment or missed detection, and still cannot accurately predict possible sudden faults. Summary of the Invention

[0006] In view of the deficiencies in the prior art, the present invention provides an abnormality monitoring and management system for a dry-type bushing capacitor core production line, which solves the technical problems in the above-mentioned background technology.

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] An abnormal monitoring and management system for a dry-type bushing capacitor core production line includes a data acquisition module, a failure rate evaluation module, and a management module;

[0009] The data acquisition module is used to obtain the working information of the winding device in real time;

[0010] In the data acquisition module, the working information includes the support roller temperature T, the support roller speed difference Δω, the support roller vibration amplitude A v , roller pressure F p0 and cylinder compensation displacement D c ;

[0011] The failure rate assessment module is used to analyze the failure rate of the winding device caused by each type of working information and the coupling relationship between different working information. Then, based on the failure rate of the winding device caused by each type of working information and the coupling relationship between different working information, it predicts whether the winding device will fail.

[0012] The management module is used to adjust the maintenance time of the winding device according to the prediction result of the winding device failure.

[0013] Furthermore, in the data acquisition module, the working information includes the support roller temperature, the support roller speed difference, the support roller vibration amplitude, the pressure roller pressure and the cylinder compensation displacement.

[0014] Furthermore, in the failure rate assessment module, the steps for predicting whether the winding device has failed are as follows:

[0015] S11, obtaining historical working data of the winding device, and using the historical working data as training samples and sample labels;

[0016] S12. Use the training samples and sample labels to train the long short-term memory network to obtain the target model;

[0017] S13. Sort the real-time acquired work information by time to obtain a work information sequence, and shift L time points forward from the current time point in the work information sequence to obtain an input sequence;

[0018] S14. Input the input sequence into the target model to obtain a prediction result of a winding device failure.

[0019] Furthermore, in step S11, the following steps are specifically included:

[0020] S111. Arrange the historical work data in chronological order to obtain a historical work information sequence, set a sliding window of length L in the historical work information sequence, and slide the sliding window once each time, obtaining a training sample each time the sliding window slides;

[0021] S112, shifting the historical work information corresponding to k moments from the last moment of the training sample backward;

[0022] S113. The quality of the capacitor cores produced at k moments is tested. If the quality of the capacitor cores is unqualified, the sample label of the training sample is that the winding device has failed; if the quality of the capacitor cores is qualified, the sample label of the training sample is that the winding device has not failed.

[0023] Furthermore, the calculation formula of the i-th hidden layer of the target model includes:

[0024] The forget gate F of the i-th hidden layer of the target model i The calculation formula is as follows:

[0025] F i =σ(x i w F1 +H i-1 wF2+b F );

[0026] Where w F1、w F2 and b F Represents the forget gate F of the i-th hidden layer i the corresponding first, second, and third weight parameters;

[0027] The input gate I of the i-th hidden layer of the target model i The calculation formula is as follows:

[0028] I i =σ(x i w I1 +H i-1 w I2 +b I );

[0029] Where w I1 、w I2 and b I Represents the input gate I of the i-th hidden layer i the corresponding first, second, and third weight parameters;

[0030] The intermediate state of the i-th hidden layer of the target model The calculation formula is as follows:

[0031]

[0032] Where w C1 、w C2 and b C Represents the intermediate state of the i-th hidden layer the corresponding first, second, and third weight parameters;

[0033] The output state C of the i-th hidden layer of the target model i The calculation formula is as follows:

[0034]

[0035] Where C i-1 Represents the output state of the i-1th hidden layer;

[0036] The output gate O of the i-th hidden layer of the target model i The calculation formula is as follows:

[0037] O i =σ(x i w O1 +H i-1 w O2 +b O );

[0038] Where w O1 、w O2and b0 represents the output gate of the i-th hidden layer

[0039] O i the corresponding first, second, and third weight parameters;

[0040] The output H of the i-th hidden layer of the target model i The calculation formula is as follows:

[0041] H i =O i ⊙tanh(C i );

[0042] In the formula, the output gate O i and tanh(C i ) multiply point by point to get the output H i ;

[0043] Definition: x i represents the input of the i-th hidden layer; σ represents the sigmoid function; ⊙ represents the dot product operation.

[0044] Furthermore, in the failure rate assessment module, the steps for predicting whether the winding device has failed are as follows:

[0045] S21. Calculate the individual failure probability P(f) corresponding to each parameter based on the working information. i ), which includes the support roller temperature fault parameter P(f T ), support roller speed difference fault parameter P(f ω ), support roller vibration amplitude fault parameter P(f v ), roller pressure fault parameter P(f p ) and cylinder compensation displacement fault parameter P(f c );

[0046] S22. Calculate the coupling fault correction coefficient α based on the working information jk ;

[0047] S23, according to the coupling fault correction coefficient α jk Calculate the probability of each two individual failures P(f i ) between the mutual fault impact value λ 耦合 ;

[0048] S24, according to the single failure probability P(f i ) and mutual fault impact value λ 耦合 Calculate the comprehensive failure rate λ 综合 ;

[0049] S25. Determine the comprehensive failure rate λ 综合 Is it higher than the standard failure rate λ? 标准 ;

[0050] If so, it means that the winding device will malfunction;

[0051] If not, it means that the winding device will not malfunction.

[0052] Furthermore, in step S21, the support roller temperature fault parameter P(f T ) is calculated as:

[0053]

[0054] Where, T represents the support roller temperature; T0 represents the set temperature; ΔT represents the allowable temperature difference; σ T represents the standard deviation of temperature fluctuation; Φ represents the cumulative distribution function of the standard normal distribution;

[0055] Support roller speed difference fault parameter P(f ω ) is calculated as:

[0056]

[0057] Where, exp represents the exp function; Δω represents the speed difference of the support roller;

[0058] ω crit Indicates critical speed difference;

[0059] Support roller vibration amplitude fault parameter P(f v ) is calculated as:

[0060]

[0061] Where A v Indicates the vibration amplitude of the support roller;

[0062] Roller pressure fault parameter P(f p ) is calculated as:

[0063]

[0064] Where, F p0 Indicates the set roller pressure; F p Indicates roller pressure;

[0065] represents the characteristic function;

[0066] Cylinder compensation displacement fault parameter P(f c ) is calculated as:

[0067]

[0068] Where k represents the steepness coefficient; D cIndicates the cylinder compensation displacement; D max Indicates the maximum compensation amount of the cylinder.

[0069] Furthermore, in step S22, the coupling fault correction coefficient α jk The calculation formula is:

[0070]

[0071] Where η represents the system coupling strength coefficient; β represents the time attenuation coefficient; t represents the cumulative operating time of the winding device; C jk represents the covariance of the j-th and k-th job information; C j and C k denote the variance of the j-th and k-th job information respectively.

[0072] Further, in step S23, the mutual fault impact value λ 耦合 The calculation formula is:

[0073]

[0074] Where, P(f j ) and P(f k ) represent the jth and kth individual failure probabilities P(f i ).

[0075] Furthermore, in step S24, the comprehensive failure rate λ 综合 The calculation formula is:

[0076]

[0077] Where λ 基线 Represents the background failure rate of the winding device.

[0078] Compared with the existing technology, the present invention provides a dry-type bushing capacitor core production line abnormality monitoring and management system, which has the following beneficial effects:

[0079] 1. The present invention uses a neural network model to analyze the working information of the winding device, which can automatically learn the complex nonlinear relationship and time series dependency between various parameters, and processes time series data through a sliding window, so that the model can dynamically reflect the evolution trend of the equipment status, which is suitable for predicting sudden failures. Compared with common fault prediction methods, it can achieve a higher fault prediction accuracy.

[0080] 2. When predicting the probability of failure of the winding device, the present invention first calculates the individual failure probability of each parameter, then introduces a coupled fault correction coefficient to quantify the mutual influence between the parameters, and finally obtains the weighted comprehensive failure rate. This can ensure the prediction accuracy of the failure probability of the winding device, and has low requirements on hardware computing power, does not require a large amount of training data, and can adapt to more old systems and edge devices. BRIEF DESCRIPTION OF THE DRAWINGS

[0081] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0082] Figure 1 This is a module block diagram of the abnormal monitoring and management system of the present invention. DETAILED DESCRIPTION

[0083] To make the above-mentioned objectives, features, and advantages of the present invention more clearly understood, the present invention is further described below in detail with reference to the accompanying drawings and specific embodiments. This will enable a full understanding of how this application uses technical means to solve technical problems and achieve technical effects, and to implement the invention accordingly.

[0084] Those skilled in the art will appreciate that all or part of the steps in the following embodiments can be accomplished by instructing related hardware through a program. Therefore, the present application may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present application may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0085] The winding device is a key piece of equipment in the dry-type bushing capacitor core production line. It is used to layer crepe paper and aluminum foil onto the coiling tube according to process requirements and is a crucial component in core winding. The winding device consists of a support roller system, a pressure roller system, a coiling tube fixing system, and an air circuit system. If one or more of these systems suddenly malfunction, the core winding will be substandard, requiring downtime for maintenance and impacting delivery cycles. To reduce unexpected winding device failures, please refer to Figure 1 As shown, the first embodiment of the present invention provides a dry-type bushing capacitor core production line abnormality monitoring and management system, including a data acquisition module, a failure rate evaluation module and a management module;

[0086] The data acquisition module is used to obtain the working information of the winding device in real time; specifically, in the data acquisition module, the working information includes the support roller temperature, support roller speed difference, support roller vibration amplitude, pressure roller pressure and cylinder compensation displacement; specifically, the support roller temperature is obtained by infrared temperature sensing; the support roller speed difference is obtained by encoders installed in the drive motors of the two support rollers; the support roller vibration amplitude is obtained by vibration acceleration sensors installed in the support roller bearing seats; the pressure roller pressure is obtained by pressure sensors in the pressure roller cylinders; and the cylinder compensation position is obtained by displacement sensors installed on the Festo DSBC cylinders of the pressure rollers.

[0087] The failure rate assessment module is used to analyze the failure rate of the winding device caused by each type of work information, and analyze the coupling relationship between different work information. Then, based on the failure rate of the winding device caused by each type of work information and the coupling relationship between different work information, it predicts whether the winding device will fail. Specifically, the failure probability of common winding devices only calculates the failure rate of each system separately based on the acquired parameters, without considering the coupling relationship between different systems. For example, the cylinder failure rate is normal, but the compensation lag caused by the fluctuation of the gas line pressure leads to low accuracy in calculating the overall failure rate. Therefore, in the failure rate assessment module, the steps for predicting whether the winding device will fail are as follows:

[0088] S11, obtaining historical working data of the winding device, and using the historical working data as training samples and sample labels; specifically, in step S11, the following steps are specifically included:

[0089] S111. Arrange the historical work data in chronological order to obtain a historical work information sequence, set a sliding window of length L in the historical work information sequence, and slide the sliding window once each time, obtaining a training sample each time the sliding window slides; specifically, L is a custom parameter;

[0090] S112. Shift the historical work information corresponding to k moments backward from the last moment of the training sample; specifically, k is a custom parameter;

[0091] S113. The quality of the capacitor cores produced at k moments is tested. If the quality of the capacitor cores is unqualified, the sample label of the training sample is that the winding device has failed; if the quality of the capacitor cores is qualified, the sample label of the training sample is that the winding device has not failed.

[0092] S12. Train the long short-term memory network using the training samples and sample labels to obtain a target model. Specifically, the calculation formula for the i-th hidden layer of the target model includes:

[0093] The forget gate F of the i-th hidden layer of the target model i The calculation formula is as follows:

[0094] F i =σ(x i w F1 +H i-1 w F2 +b F );

[0095] Where w F1 、w F2 and b F Represents the forget gate F of the i-th hidden layer i the corresponding first, second, and third weight parameters;

[0096] The input gate I of the i-th hidden layer of the target model i The calculation formula is as follows:

[0097] I i =σ(x i w I1 +H i-1 w I2 +b I );

[0098] Where w I1 、w I2 and b I Represents the input gate I of the i-th hidden layer i the corresponding first, second, and third weight parameters;

[0099] The intermediate state of the i-th hidden layer of the target model The calculation formula is as follows:

[0100]

[0101] Where w C1 、w C2 and b C Represents the intermediate state of the i-th hidden layer the corresponding first, second, and third weight parameters;

[0102] The output state C of the i-th hidden layer of the target model i The calculation formula is as follows:

[0103]

[0104] Where C i-1 Represents the output state of the i-1th hidden layer;

[0105] The output gate O of the i-th hidden layer of the target model iThe calculation formula is as follows:

[0106] O i =σ(x i w O1 +H i-1 W O2 +b O );

[0107] Where w O1 、w O2 and b O represents the output gate O of the i-th hidden layer i the corresponding first, second, and third weight parameters;

[0108] The output H of the i-th hidden layer of the target model i The calculation formula is as follows:

[0109] H i =O i ⊙tanh(C i );

[0110] In the formula, the output gate O i and tanh(C i ) multiply point by point to get the output H i ;

[0111] Definition: x i represents the input of the i-th hidden layer; σ represents the sigmoid function; ⊙ represents the dot product operation.

[0112] S13. Sort the real-time acquired work information by time to obtain a work information sequence, and shift L time points forward from the current time point in the work information sequence to obtain an input sequence;

[0113] S14. Input the input sequence into the target model to obtain a prediction result of a winding device failure.

[0114] The management module is used to adjust the maintenance time of the winding device according to the prediction result of the winding device failure. Specifically, since the winding equipment is generally inspected and repaired periodically, it cannot cope with sudden failures, which may easily lead to a batch of capacitor cores with unqualified quality, thereby reducing production efficiency and yield rate. Therefore, after obtaining the prediction result of the winding device failure, if the prediction result is that a failure is about to occur, then maintenance and repair are carried out in advance; otherwise, maintenance is carried out according to the normal cycle.

[0115] In the first embodiment of the present invention, a neural network model is used to analyze the working information of the winding device, which can automatically learn the complex nonlinear relationship and timing dependency between various parameters, and process the timing data through a sliding window, so that the model can dynamically reflect the evolution trend of the equipment status, which is suitable for predicting sudden failures. Compared with common fault prediction methods, it can achieve a higher fault prediction accuracy.

[0116] The purpose of the second embodiment of the present invention is that since using a neural network to calculate the failure rate of a winding device requires a large number of training samples and the training process consumes a lot of time, the steps for predicting whether a winding device has failed in the failure rate assessment module are as follows:

[0117] S21. Calculate the individual failure probability P(f) corresponding to each parameter based on the working information. i ), which includes the support roller temperature fault parameter P(f T ), support roller speed difference fault parameter P(f ω ), support roller vibration amplitude fault parameter P(f v ), roller pressure fault parameter P(f p ) and cylinder compensation displacement fault parameter P(f c ); Specifically, in step S21, the support roller temperature fault parameter P (f T ) is calculated as:

[0118]

[0119] Where, T represents the support roller temperature; T0 represents the set temperature; ΔT represents the allowable temperature difference; σ T represents the standard deviation of temperature fluctuation; Φ represents the cumulative distribution function of the standard normal distribution; in the present invention, T0 is 100; ΔT is 2.5; σ T is 1.2;

[0120] Support roller speed difference fault parameter P(f ω ) is calculated as:

[0121]

[0122] Where, exp represents the exp function; Δω represents the speed difference of the support roller; ω crit Indicates the critical speed difference; in the present invention, ω crit is 0.5;

[0123] Support roller vibration amplitude fault parameter P(f v ) is calculated as:

[0124]

[0125] Where A v Indicates the vibration amplitude of the support roller;

[0126] Roller pressure fault parameter P(f p ) is calculated as:

[0127]

[0128] Where, F p0 Indicates the set roller pressure; F p Indicates roller pressure;

[0129] represents the characteristic function; in the present invention, F p0 is 30-120; when |ΔF p |>2% when established, When |ΔF p |>2% does not hold true,

[0130] Cylinder compensation displacement fault parameter P(f c ) is calculated as:

[0131]

[0132] Where k represents the steepness coefficient; D c Indicates the cylinder compensation displacement; D max Indicates the maximum compensation amount of the cylinder; in the present invention, k is 0.5; D max is 20;

[0133] S22. Calculate the coupling fault correction coefficient α based on the working information jk Specifically, the common method for calculating the failure rate of a winding device generally only considers the influence of a single parameter on the failure rate, without considering the mutual influence between multiple factors, for example: To this end, in step S22, the coupling fault correction coefficient α jk The calculation formula is:

[0134]

[0135] Where η represents the system coupling strength coefficient; β represents the time attenuation coefficient; t represents the cumulative operating time of the winding device; C jk represents the covariance of the j-th and k-th job information; C j and C k Respectively represent the variance of the j-th and k-th work information; in the present invention, the value range of η is 0.35; β is 0.0015; C jk =Cov(P(f j ), P(f k ));Cj =Var(P(f j ));C k =Var(P(f k ));

[0136] S23, according to the coupling fault correction coefficient α jk Calculate the probability of each two individual failures P(f i ) between the mutual fault impact value λ 耦合 Specifically, in step S23, the mutual fault impact value λ 耦合 The calculation formula is:

[0137]

[0138] Where, P(f i ) and P(f k ) represent the jth and kth individual failure probabilities P(f i );

[0139] S24, according to the probability of single failure P(f i ) and mutual fault impact value λ 耦合 Calculate the comprehensive failure rate λ 综合 Specifically, in step S24, the comprehensive failure rate λ 综合 The calculation formula is:

[0140]

[0141] Where λ 基线 Represents the background failure rate of the winding device; In the present invention, λ 基线 5×10 -6 ;

[0142] S25. Determine the comprehensive failure rate λ 综合 Is it higher than the standard failure rate λ? 标准 ; Specifically, the standard failure rate λ 标准 65%;

[0143] If so, it means that the winding device will malfunction;

[0144] If not, it means that the winding device will not malfunction.

[0145] In the second embodiment of the present invention, when predicting the probability of failure of the winding device, the individual failure probability of each parameter is first calculated, and then a coupling fault correction coefficient is introduced to quantify the mutual influence between the parameters. Finally, the comprehensive failure rate is weighted to ensure the prediction accuracy of the failure probability of the winding device, and has low requirements on hardware computing power, does not require a large amount of training data, and can adapt to more old systems and edge devices.

[0146] The above embodiments provide a detailed introduction to the present invention. Specific examples are used herein to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea. At the same time, for those skilled in the art, according to the ideas of the present invention, there may be changes in the specific implementation methods and application scopes. In summary, the contents of this specification should not be understood as limiting the present invention.

Claims

1. A dry-type bushing capacitor core production line abnormality monitoring and management system, characterized in that: Including data acquisition module, failure rate assessment module and management module; The data acquisition module is used to obtain the working information of the winding device in real time; In the data acquisition module, the working information includes the support roller temperature T, the support roller speed difference Δω, the support roller vibration amplitude A v , roller pressure F p0 and cylinder compensation displacement D c ; The failure rate assessment module is used to analyze the failure rate of the winding device caused by each type of working information and the coupling relationship between different working information. Then, based on the failure rate of the winding device caused by each type of working information and the coupling relationship between different working information, it predicts whether the winding device will fail. The management module is used to adjust the maintenance time of the winding device according to the prediction result of the winding device failure.

2. The dry-type bushing capacitor core production line abnormality monitoring and management system according to claim 1, characterized in that: In the failure rate assessment module, the steps for predicting whether the winding device has failed are as follows: S11, obtaining historical working data of the winding device, and using the historical working data as training samples and sample labels; S12. Use the training samples and sample labels to train the long short-term memory network to obtain the target model; S13. Sort the real-time acquired work information by time to obtain a work information sequence, and shift L time points forward from the current time point in the work information sequence to obtain an input sequence; S14. Input the input sequence into the target model to obtain a prediction result of a winding device failure.

3. The dry-type bushing capacitor core production line abnormality monitoring and management system according to claim 2, characterized in that: In step S11, the following steps are specifically included: S111. Arrange the historical work data in chronological order to obtain a historical work information sequence, set a sliding window of length L in the historical work information sequence, and slide the sliding window once each time, obtaining a training sample each time the sliding window slides; S112, shifting the historical work information corresponding to k moments from the last moment of the training sample backward; S113. The quality of the capacitor cores produced at k moments is tested. If the quality of the capacitor cores is unqualified, the sample label of the training sample is that the winding device has failed; if the quality of the capacitor cores is qualified, the sample label of the training sample is that the winding device has not failed.

4. The dry-type bushing capacitor core production line abnormality monitoring and management system according to claim 2, characterized in that: The calculation formula for the i-th hidden layer of the target model includes: The forget gate F of the i-th hidden layer of the target model i The calculation formula is as follows: F i =σ(x i w F1 +H i-1 w F2 +b F ); Where w F1 、w F2 and b F Represents the forget gate F of the i-th hidden layer i the corresponding first, second, and third weight parameters; The input gate I of the i-th hidden layer of the target model i The calculation formula is as follows: I i =σ(x i w I1 +H i-1 w I2 +b I ); Where w I1 、w I2 and b I Represents the input gate I of the i-th hidden layer i the corresponding first, second, and third weight parameters; The intermediate state of the i-th hidden layer of the target model The calculation formula is as follows: Where w C1 、w C2 and b C Represents the intermediate state of the i-th hidden layer the corresponding first, second, and third weight parameters; The output state C of the i-th hidden layer of the target model i The calculation formula is as follows: Where C i-1 Represents the output state of the i-1th hidden layer; The output gate O of the i-th hidden layer of the target model i The calculation formula is as follows: O i =σ(x i w 01 +H i-1 w 02 +b0); Where w 01 、w 02 and b0 represents the output gate O of the i-th hidden layer i the corresponding first, second, and third weight parameters; The output H of the i-th hidden layer of the target model i The calculation formula is as follows: H i =O i ⊙tanh(C i ); In the formula, the output gate O i and tanh(C i ) multiply point by point to get the output H i ; Definition: x i represents the input of the i-th hidden layer; σ represents the sigmoid function; ⊙ represents the dot product operation.

5. The dry-type bushing capacitor core production line abnormality monitoring and management system according to claim 1, characterized in that: In the failure rate assessment module, the steps for predicting whether the winding device has failed are as follows: S21. Calculate the individual failure probability P(f) corresponding to each parameter based on the working information. i ), which includes the support roller temperature fault parameter P(f T ), support roller speed difference fault parameter P(f ω ), support roller vibration amplitude fault parameter P(f v ), roller pressure fault parameter P(f p ) and cylinder compensation displacement fault parameter P(f c ); S22. Calculate the coupling fault correction coefficient α based on the working information jk ; S23, according to the coupling fault correction coefficient α jk Calculate the probability of each two individual failures P(f i ) between the mutual fault impact value λ 耦合 ; S24, according to the probability of single failure P(f i ) and mutual fault impact value λ 耦合 Calculate the comprehensive failure rate λ 综合 ; S25. Determine the comprehensive failure rate λ 综合 Is it higher than the standard failure rate λ? 标准 ; If so, it means that the winding device will malfunction; If not, it means that the winding device will not malfunction.

6. The dry-type bushing capacitor core production line abnormality monitoring and management system according to claim 5, characterized in that: In step S21, the support roller temperature fault parameter P(f T ) is calculated as: Where, T represents the support roller temperature; f0 represents the set temperature; ΔT represents the allowable temperature difference; σ T represents the standard deviation of temperature fluctuation; Φ represents the cumulative distribution function of the standard normal distribution; Support roller speed difference fault parameter P(f ω ) is calculated as: Where, exp represents the exp function; Δω represents the speed difference of the support roller; ω crit Indicates critical speed difference; Support roller vibration amplitude fault parameter P(f v ) is calculated as: Where A v Indicates the vibration amplitude of the support roller; Roller pressure fault parameter P(f p ) is calculated as: Where, F p0 Indicates the set roller pressure; F p Indicates roller pressure; represents the characteristic function; Cylinder compensation displacement fault parameter P(f c ) is calculated as: Where k represents the steepness coefficient; D c Indicates the cylinder compensation displacement; D max Indicates the maximum compensation amount of the cylinder.

7. The dry-type bushing capacitor core production line abnormality monitoring and management system according to claim 5, characterized in that: In step S22, the coupling fault correction coefficient α jk The calculation formula is: Where η represents the system coupling strength coefficient; β represents the time attenuation coefficient; t represents the cumulative operating time of the winding device; C jk represents the covariance of the j-th and k-th job information; C j and C k denote the variance of the j-th and k-th job information respectively.

8. The dry-type bushing capacitor core production line abnormality monitoring and management system according to claim 5, characterized in that: In step S23, the mutual fault impact value λ 耦合 The calculation formula is: Where, P(f j ) and P(f k ) represent the jth and kth individual failure probabilities P(f i ).

9. The dry-type bushing capacitor core production line abnormality monitoring and management system according to claim 5, characterized in that: In step S24, the comprehensive failure rate λ 综合 The calculation formula is: Where λ 基线 Represents the background failure rate of the winding device.

Images