Coal belt pull rope switch fault monitoring method and system

Abstract

The coal belt pull rope switch fault monitoring method and system provided by the application, since the first quality evaluation is obtained according to the comparison between the first feature vector and the second feature vector, the comparison thread is obtained based on the first quality evaluation, the difference between the performance of the comparison thread and the performance of the first sub-thread can be reduced, and thus the performance of the comparison thread is improved. The fault data processing device further processes the first to-be-compared fault data and the second to-be-compared fault data using the comparison thread to obtain the commonality coefficient of the second to-be-compared matter and the third to-be-compared matter, and the accuracy of the commonality coefficient can be improved; and thus the position of the fault can be determined more accurately.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of fault monitoring, in particular to a coal belt pull rope switch fault monitoring method and system. BACKGROUND

[0002] The pull rope switch is also called "emergency stop switch", which mainly acts on the belt conveying site. When emergency stop is needed, the switch can be actuated to stop the conveyor.

[0003] Since the conveying system is generally used for long-distance transmission, a plurality of pull rope switches need to be arranged for a belt conveyor. In order to prevent the pull rope switch from refusing to act, a series working mode is often used, which causes the pull rope switch control loop to be open at any point and cannot be found, causing great difficulty for maintenance personnel and seriously affecting the defect elimination time. The shutdown of the entire belt conveying system also causes great loss to the enterprise. Therefore, there is an urgent need for a monitoring method to solve the above technical problems. SUMMARY

[0004] To improve the technical problems in the related art, the present application provides a coal belt pull rope switch fault monitoring method and system.

[0005] In a first aspect, a coal belt pull rope switch fault monitoring method is provided, which comprises: obtaining a comparison thread, first comparison fault data and second comparison fault data; the comparison thread is obtained according to the coefficient of the first quality evaluation optimization thread; the first quality evaluation is obtained only according to the comparison of the first feature vector and the second feature vector; the first feature vector is obtained by extracting the features of the first comparison matter in the configuration data by the first sub-thread, and the second feature vector is obtained by extracting the features of the first comparison matter by the second sub-thread; the first comparison fault data and the second comparison fault data are processed using the comparison thread to obtain the commonality coefficient of the second comparison matter in the first comparison fault data and the third comparison matter in the second comparison fault data, and the switch fault positioning is determined according to the commonality coefficient of the second comparison matter and the third comparison matter.

[0006] Further, the coal belt pull rope switch fault monitoring method is applied to a fault data processing device, which comprises a monitoring device. The first comparison fault data and the second comparison fault data are obtained by using the monitoring device to collect local fault data of the analyzed whole belt running information to determine the first comparison fault data, and obtaining local fault data from the local fault database of the fault data processing device to determine the second comparison fault data. The method further comprises: outputting the information derived from the analyzed whole belt running information on the premise that the commonality coefficient exceeds the local commonality coefficient target value.

[0007] Further, the obtaining the comparison thread comprises: obtaining the configuration data, the first sub-thread and the second sub-thread; performing feature extraction processing on the configuration data using the first sub-thread to obtain the first feature vector; performing feature extraction processing on the configuration data using the second sub-thread to obtain the second feature vector; determining a comparison between the first feature vector and the second feature vector to obtain a first comparison; obtaining the first quality evaluation in combination with the first comparison; the first comparison is related to the first quality evaluation; and optimizing coefficients of the second sub-thread based on the first quality evaluation to obtain the comparison thread.

[0008] Further, before the optimizing the coefficients of the second sub-thread based on the first quality evaluation to obtain the comparison thread, the method further comprises: determining a second comparison between the second feature vector and a third feature vector; the third feature vector is a feature vector output by a decision layer corresponding to a category of the configuration data in the second sub-thread; obtaining a second quality evaluation of the second sub-thread under monitoring of the configuration data in combination with the second comparison; the second comparison is related to the second quality evaluation; obtaining an overall quality evaluation of the second sub-thread in combination with the first quality evaluation and the second quality evaluation; the overall quality evaluation is related to the first quality evaluation, and the overall quality evaluation is related to the second quality evaluation; and the optimizing the coefficients of the second sub-thread based on the first quality evaluation to obtain the comparison thread comprises: optimizing the coefficients of the second sub-thread based on the overall quality evaluation to obtain a configured second sub-thread.

[0009] Further, the obtaining the overall quality evaluation of the second sub-thread in combination with the first quality evaluation and the second quality evaluation comprises: obtaining a third quality evaluation in combination with a term in the first quality evaluation that covers the first comparison and a term in the second quality evaluation that covers the second comparison; the third quality evaluation is not related to the first comparison, and the third quality evaluation is not related to the second comparison; and obtaining the overall quality evaluation in combination with the third quality evaluation; the third quality evaluation is not related to the overall quality evaluation.

[0010] Further, the combining the item involving the first comparison in the first quality evaluation and the item involving the second comparison in the second quality evaluation to obtain the third quality evaluation comprises: obtaining a first coefficient; the first coefficient is a confidence degree of the item involving the first comparison in a process of combining the item involving the first comparison and the item involving the second comparison to obtain the third quality evaluation; determining a function calculation between the first coefficient and the item involving the first comparison to obtain a first value; combining a sum between the item involving the second comparison and the first value to obtain the third quality evaluation; and the third quality evaluation is not related to the first value.

[0011] Further, the combining the second comparison to obtain the second quality evaluation of the second sub-thread under the monitoring of the configuration data comprises: obtaining a second coefficient; the second coefficient is related to a convergence difficulty of the second quality evaluation; determining a sum between the second comparison and the second coefficient to obtain a second value; and determining a function value of the second value to determine the second quality evaluation.

[0012] Further, the combining the third quality evaluation to obtain the total quality evaluation comprises: obtaining a third coefficient and a fourth coefficient; the third coefficient is not related to a convergence difficulty of the total quality evaluation; the fourth coefficient is greater than 1; determining a function calculation between the third coefficient and the third quality evaluation to obtain a third value; taking the fourth coefficient as a base number and the third value as an index to obtain a fourth value; combining the fourth value to obtain the total quality evaluation; and the total quality evaluation is not related to the fourth value.

[0013] Further, the configuration data comprises local fault data; and the first feature vector and the second feature vector are both local feature vectors of the first to-be-compared item.

[0014] In a second aspect, a coal belt pull rope switch fault monitoring system is provided, which comprises a processor and a memory in communication with each other, and the processor is configured to read a computer program from the memory and execute the computer program to implement the method described above.

[0015] The coal belt pull rope switch fault monitoring method and system provided by the embodiments of the present application can obtain the first quality evaluation based on the comparison between the first feature vector and the second feature vector, reduce the difference between the performance of the comparison thread and the performance of the first sub-thread, and improve the performance of the comparison thread. The fault data processing device further processes the first to-be-compared fault data and the second to-be-compared fault data by using the comparison thread to obtain the commonality coefficient of the second to-be-compared item and the third to-be-compared item, and improve the accuracy of the commonality coefficient, so that the position of the fault can be determined more accurately. BRIEF DESCRIPTION OF DRAWINGS

[0016] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed to be used in the embodiments will be briefly introduced as follows. It should be understood that the following drawings only show some of the embodiments of the present application, and therefore should not be considered as limiting the scope. For those skilled in the art, other related drawings can also be obtained without creative labor on the basis of these drawings.

[0017] Figure 1 A flow chart of a coal belt pull rope switch fault monitoring method provided by the embodiments of the present application. DETAILED DESCRIPTION

[0018] In order to better understand the above technical solutions, the technical solutions of the present application will be described in detail below through the drawings and specific embodiments. It should be understood that the embodiments of the present application and the specific features in the embodiments are detailed descriptions of the technical solutions of the present application, and are not limitations of the technical solutions of the present application. In the case of no conflict, the technical features in the embodiments of the present application and the embodiments can be combined with each other.

[0019] Please refer to Figure 1 , which shows a coal belt pull rope switch fault monitoring method. The method can include the technical solutions described in steps 101 and 102.

[0020] 101. Obtain a comparison thread, first comparison fault data and second comparison fault data, wherein the comparison thread is obtained according to the coefficient of the first quality evaluation optimization thread, the first quality evaluation is obtained according to the comparison of the first feature vector and the second feature vector, the first feature vector is obtained by extracting the features of the first comparison matter in the first comparison data through the first sub-thread, and the second feature vector is obtained by extracting the features of the first comparison matter through the second sub-thread.

[0021] In the present embodiment, the comparison thread is used to process the comparison fault data to determine whether the comparison matter in the fault data is the same matter. For example, under the condition that the comparison matter is a local premise, the comparison thread can be a local analysis thread.

[0022] In the embodiment, the comparison thread is obtained based on the first quality evaluation, and the comparison thread is obtained based on the first quality evaluation. The comparison thread is obtained based on the first quality evaluation. The comparison between the first feature vector and the second feature vector is determined as the first quality evaluation.

[0023] Since the first quality evaluation is obtained based on the comparison between the first feature vector and the second feature vector, the comparison thread is obtained based on the first quality evaluation, and the difference between the performance of the comparison thread and the performance of the first sub-thread is reduced, thereby improving the performance of the comparison thread.

[0024] In the embodiment, the first comparison fault data covers the second comparison matter, and the second comparison fault data covers the third comparison matter.

[0025] In one implementation manner of obtaining the first comparison fault data, the fault data processing apparatus obtains the first comparison fault data by receiving the first comparison fault data input by the user through the input component.

[0026] In another implementation manner of obtaining the first comparison fault data, the fault data processing apparatus obtains the first comparison fault data by receiving the first comparison fault data sent by the terminal.

[0027] In one implementation manner of obtaining the second comparison fault data, the fault data processing apparatus obtains the second comparison fault data by receiving the second comparison fault data input by the user through the input component.

[0028] In another implementation manner of obtaining the second comparison fault data, the fault data processing apparatus obtains the second comparison fault data by receiving the second comparison fault data sent by the terminal.

[0029] 102、using the above comparison thread to process the above first comparison fault data and the above second comparison fault data, obtaining the commonality coefficient of the second comparison matter in the first comparison fault data and the third comparison matter in the second comparison fault data, and determining the switch fault positioning according to the commonality coefficient of the third comparison matter.

[0030] The fault data processing device uses the comparison thread to process the first to-be-compared fault data, and extracts a feature vector of the second to-be-compared matter. The fault data processing device uses the comparison thread to process the second to-be-compared fault data, and extracts a feature vector of the third to-be-compared matter. The fault data processing device obtains the commonality coefficient of the second to-be-compared matter and the third to-be-compared matter by calculating the commonality coefficient between the feature vector of the second to-be-compared matter and the feature vector of the third to-be-compared matter.

[0031] In an alternative embodiment, the fault data processing device obtains the commonality coefficient of the second to-be-compared matter and the third to-be-compared matter by calculating the cosine commonality coefficient between the feature vector of the second to-be-compared matter and the feature vector of the third to-be-compared matter.

[0032] In this embodiment, since the first quality evaluation is obtained according to the comparison between the first feature vector and the second feature vector, the performance of the comparison thread can be reduced based on the first quality evaluation, thereby improving the performance of the comparison thread. The fault data processing device further uses the comparison thread to process the first to-be-compared fault data and the second to-be-compared fault data, and obtains the commonality coefficient of the second to-be-compared matter and the third to-be-compared matter, thereby improving the accuracy of the commonality coefficient, and accurately determining the location of the fault.

[0033] In an alternative embodiment, the fault data processing device comprises a monitoring device. The fault data processing device obtains the first to-be-compared fault data and the second to-be-compared fault data by performing the following steps.

[0034] 1. The first to-be-compared fault data is determined by using the monitoring device to collect local fault data of the running information of the entire belt to be analyzed.

[0035] 2. The second to-be-compared fault data is determined by obtaining local fault data from the local fault database of the fault data processing device.

[0036] In this embodiment, the local fault data in the local fault database is all authenticated local fault data.

[0037] In this embodiment, the second to-be-compared matter and the third to-be-compared matter are both local. In this embodiment, the fault data processing device further performs the following steps:

[0038] 3. On the premise that the commonality coefficient exceeds the target value of the local commonality coefficient, the information derived from the running information of the entire belt to be analyzed is output.

[0039] The fault data processing apparatus determines that the second to-be-compared matter and the third to-be-compared matter are the same running device on the premise that the commonality coefficient (i.e., the commonality coefficient of the second to-be-compared matter and the third to-be-compared matter) exceeds the local commonality coefficient target value. The fault data processing apparatus determines that the second to-be-compared matter and the third to-be-compared matter are not the same running device on the premise that the commonality coefficient does not exceed the local commonality coefficient target value.

[0040] In an optional implementation, the fault data processing apparatus obtains the comparison thread by performing the following steps:

[0041] 4. Obtain the configuration data, the first sub-thread, and the second sub-thread.

[0042] In this embodiment, the configuration data can be one of the following: fault data, voice, and sentence. The configuration data all contain annotation information.

[0043] In this embodiment, the number of coefficients of the first sub-thread is greater than the number of coefficients of the second sub-thread. The first sub-thread and the second sub-thread can both be deep learning threads of an arbitrary structure. For example, the first sub-thread can include a convolution layer, a down-sampling layer, an up-sampling layer, a pooling layer, a normalization layer, and a decision layer. The second sub-thread can include a convolution layer, a pooling layer, a normalization layer, and a decision layer.

[0044] In this embodiment, the first sub-thread is a configured deep learning thread, and the first sub-thread has the ability to perform the task required to be performed by the second sub-thread. For example, if the task required to be performed by the second sub-thread is local analysis, the first sub-thread has the ability to perform the local analysis task.

[0045] In one implementation of obtaining the first sub-thread, the fault data processing apparatus obtains the first sub-thread by receiving the first sub-thread input by a user through an input component.

[0046] In another implementation of obtaining the first sub-thread, the fault data processing apparatus obtains the first sub-thread by receiving the first sub-thread sent by a terminal.

[0047] In one implementation of obtaining the second sub-thread, the fault data processing apparatus obtains the second sub-thread by receiving the second sub-thread input by a user through an input component.

[0048] In another implementation of obtaining the second sub-thread, the fault data processing apparatus obtains the second sub-thread by receiving the second sub-thread sent by a terminal.

[0049] 5. Perform feature extraction processing on the configuration data using the first sub-thread to obtain the first feature vector.

[0050] The feature information extracted in the feature extraction processing in this step is related to the task to be executed by the first sub-thread. For example, the first sub-thread is used to execute a local analysis task, and at this time, the first sub-thread can extract local feature information of the configuration data by performing feature extraction processing on the configuration data, to obtain a first feature vector.

[0051] 6. The feature extraction processing on the configuration data is performed using the second sub-thread, to obtain the second feature vector.

[0052] The feature information extracted in the feature extraction processing in this step is the same as the feature information extracted in the feature extraction processing in step 5.

[0053] 7. A first comparison between the first feature vector and the second feature vector is determined.

[0054] 8. A first quality evaluation is obtained according to the first comparison.

[0055] In this embodiment, the first quality evaluation is a quality evaluation of the second sub-thread under the monitoring of the first sub-thread, i.e., a soft quality evaluation. The first comparison is related to the first quality evaluation.

[0056] 9. The coefficient of the second sub-thread is optimized based on the first quality evaluation, to obtain the comparison thread.

[0057] As described above, since the difference between the length of the feature vector output by the first sub-thread and the length of the feature vector output by the second sub-thread is constrained in the process of obtaining the soft quality evaluation, the difference between the performance of the second sub-thread and the performance of the first sub-thread is increased.

[0058] In this embodiment, the fault data processing apparatus obtains a soft quality evaluation according to the comparison between the first feature vector and the second feature vector, which can reduce the difference between the performance of the second sub-thread and the performance of the first sub-thread, thereby improving the configuration effect of the second sub-thread.

[0059] As an optional implementation, the fault data processing apparatus further performs the following step before step 9.

[0060] 10. A second comparison between the second feature vector and a third feature vector is determined.

[0061] 11. A second quality evaluation of the second sub-thread under the monitoring of the configuration data is obtained according to the second comparison.

[0062] The second comparison characterizes a commonality coefficient between the feature vector output by the second sub-thread and the feature vector output by the category decision layer, as described in step 10. Thus, the fault data processing apparatus can determine, based on the second comparison, a difference between the analysis result based on the feature vector output by the second sub-thread and the labeled information of the configuration data, i.e., the second quality evaluation of the second sub-thread under the monitoring of the configuration data.

[0063] 12. Based on the first quality evaluation and the second quality evaluation, the overall quality evaluation of the second sub-thread is obtained.

[0064] In this embodiment, the overall quality evaluation is related to the first quality evaluation, and the overall quality evaluation is related to the second quality evaluation.

[0065] After the overall quality evaluation is obtained, the fault data processing apparatus performs the following steps in the process of executing step 9.

[0066] 13. Based on the overall quality evaluation, the coefficients of the second sub-thread are optimized to obtain a configured second sub-thread.

[0067] Based on the overall quality evaluation, the fault data processing apparatus optimizes the coefficients of the thread, so that the second sub-thread learns the ability of the first sub-thread through configuration, and the second sub-thread learns the ability to perform tasks under the monitoring of the labeled information of the configuration data through configuration.

[0068] As an optional implementation, the fault data processing apparatus performs the following steps in the process of executing step 12.

[0069] 14. Based on the item in the first quality evaluation that encompasses the first comparison and the item in the second quality evaluation that encompasses the second comparison, a third quality evaluation is obtained.

[0070] As described above, the first quality evaluation is obtained based on the first comparison, and the second quality evaluation is obtained based on the second comparison. Therefore, there is an item in the first quality evaluation that encompasses the first comparison, and there is an item in the second quality evaluation that encompasses the second comparison.

[0071] 15. Based on the third quality evaluation, the overall quality evaluation is obtained.

[0072] In this embodiment, by executing steps 14 and 15, the fault data processing apparatus is conducive to adjusting the size of the overall quality evaluation by adjusting the ratio between the first comparison and the second comparison. That is, it is conducive to adjusting the size of the overall quality evaluation by adjusting the ratio between the monitoring of the first sub-thread and the monitoring of the labeled information of the configuration data.

[0073] For example, if the overall quality evaluation is obtained according to the first quality evaluation and the second quality evaluation, the first quality evaluation is obtained based on the first quality evaluation function and the first comparison, the second quality evaluation is obtained according to the second quality evaluation function and the second comparison, and there is a difference in order of magnitude between the first quality evaluation function value and the second quality evaluation function value on the premise that the first comparison and the second comparison are the same. At this time, if the influence degree of the first comparison on the overall quality evaluation and the influence degree of the second comparison on the overall quality evaluation are adjusted by adjusting the proportion between the first quality evaluation and the second quality evaluation, neither of them can bring good results.

[0074] 16、obtain the first coefficient.

[0075] In this embodiment, the first coefficient is the confidence of the item covering the first comparison in the process of obtaining the third quality evaluation according to the item covering the first comparison and the item covering the second comparison.

[0076] 17、determine the function calculation between the first coefficient and the first quality evaluation to obtain the first value.

[0077] 18、obtain the third quality evaluation according to the second quality evaluation and the first value.

[0078] The determination is an optional implementation, so the fault data processing device performs the following steps in the process of performing step 11.

[0079] 19、obtain the second coefficient.

[0080] Since the second quality evaluation converges on the premise that the second quality evaluation is not greater than the first convergence target value, the smaller the first convergence target value of the second quality evaluation, the greater the convergence difficulty of the second quality evaluation; the greater the convergence difficulty of the second quality evaluation, the greater the first convergence target value of the second quality evaluation.

[0081] Because the size of the second quality evaluation is related to the second comparison, on the premise that the second comparison is not greater than the second convergence target value, the second quality evaluation is not greater than the first convergence target value, and at this time the second quality evaluation converges. Therefore, the greater the convergence difficulty of the second quality evaluation, the smaller the convergence target value of the second comparison; the smaller the convergence difficulty of the second quality evaluation, the greater the convergence target value of the second comparison.

[0082] In this embodiment, the second coefficient is related to the convergence difficulty of the second quality evaluation. Therefore, the second coefficient is not related to the convergence target value of the second comparison.

[0083] 20、determine the sum of the second comparison and the second coefficient to obtain the second value.

[0084] 21、determine the function value of the second value as the second quality evaluation.

[0085] The fault data processing device determines the function value of the second value as the second quality evaluation, when the second quality evaluation is not greater than the first convergence target value, on the premise that the second value is not greater than the second convergence target value, and the second quality evaluation converges.

[0086] Since the second value is the sum of the second contrast and the second coefficient, on the premise that the second convergence target value is constant, the fault data processing device can change the convergence target value of the second contrast, and further change the convergence difficulty of the second quality evaluation, by adjusting the size of the second coefficient. Specifically, the fault data processing device can increase the convergence target value of the second contrast, and further reduce the convergence difficulty of the second quality evaluation, by adjusting the second coefficient to be smaller. The fault data processing device can reduce the convergence target value of the second contrast, and further increase the convergence difficulty of the second quality evaluation, by adjusting the second coefficient to be larger.

[0087] As a kind of optional implementation, the fault data processing device executes the following steps in the process of executing step 15.

[0088] 22, obtain the third coefficient and the fourth coefficient.

[0089] In the embodiment, the fourth coefficient is a real number greater than 1, and the third coefficient is not related to the convergence difficulty of all quality evaluations. Since the first quality evaluation converges on the premise that all quality evaluations are not greater than the second convergence target value, and the third quality evaluation is not related to all quality evaluations, the first quality evaluation converges on the premise that the third quality evaluation is not less than the third convergence target value, that is, the greater the convergence difficulty of all quality evaluations, the smaller the second convergence target value, and the greater the third convergence target value; the smaller the convergence difficulty of all quality evaluations, the greater the second convergence target value, and the smaller the third convergence target value. Therefore, the convergence difficulty of all quality evaluations is related to the third convergence target value. Since the third coefficient is not related to the convergence difficulty of all quality evaluations, the third coefficient is not related to the third convergence target value.

[0090] 23, determine the function calculation between the third coefficient and the third quality evaluation to obtain the third value.

[0091] 24, take the fourth coefficient as the base number and the third value as the exponent to obtain the fourth value.

[0092] 25, obtain all quality evaluations according to the fourth value.

[0093] On the premise that the fault data processing device obtains all quality evaluations by executing steps 23-25, the user can adjust the size of the third coefficient input to the fault data processing device to adjust the convergence difficulty of all quality evaluations, that is, the configuration difficulty of the second sub-thread.

[0094] It is determined that the configuration data mentioned above includes local fault data. The first feature vector and the second feature vector mentioned above are both local feature vectors in the first to-be-compared matter. That is, in this embodiment, the configured second sub-thread obtained by configuring the second sub-thread based on the coal conveying belt pull rope switch fault monitoring method provided above can be used for local analysis.

[0095] In summary, based on the above scheme, since the first quality evaluation is obtained based on the comparison between the first feature vector and the second feature vector, the comparison thread obtained based on the first quality evaluation can reduce the difference between the performance of the comparison thread and the performance of the first sub-thread, thereby improving the performance of the comparison thread. The fault data processing device further processes the first to-be-compared fault data and the second to-be-compared fault data using the comparison thread to obtain the commonality coefficient of the second to-be-compared matter and the third to-be-compared matter, which can improve the accuracy of the commonality coefficient; thereby the position of the fault can be determined more accurately.

Claims

1. A coal belt pull cord switch failure monitoring method, characterized in that, The method comprises: obtaining a comparison thread, first comparison fault data and second comparison fault data; the comparison thread is obtained according to the first quality evaluation optimization thread coefficient; the first quality evaluation is only obtained according to the comparison of the first feature vector and the second feature vector; the first feature vector is obtained by extracting the features of the first comparison matter in the first comparison fault data through the first sub-thread; and the second feature vector is obtained by extracting the features of the first comparison matter through the second sub-thread; processing the first comparison fault data and the second comparison fault data using the comparison thread to obtain the commonality coefficient of the second comparison matter in the first comparison fault data and the third comparison matter in the second comparison fault data, and determining the switch fault positioning according to the commonality coefficient of the second comparison matter and the third comparison matter.

2. The method of claim 1, wherein, The coal belt pull rope switch fault monitoring method is applied to a fault data processing device, and the fault data processing device comprises a monitoring device. The first comparison fault data is determined by collecting local fault data of the whole belt running information to be analyzed using the monitoring device; and the second comparison fault data is determined by obtaining local fault data from a local fault database of the fault data processing device.

3. The method of claim 2, wherein, The method further comprises: outputting the information derived from the whole belt running information to be analyzed on the premise that the commonality coefficient exceeds a local commonality coefficient target value. The method further comprises: obtaining the configuration data, the first sub-thread and the second sub-thread; performing feature extraction processing on the configuration data using the first sub-thread to obtain the first feature vector; performing feature extraction processing on the configuration data using the second sub-thread to obtain the second feature vector; determining the comparison between the first feature vector and the second feature vector to obtain a first comparison; obtaining the first quality evaluation in combination with the first comparison; the first comparison is related to the first quality evaluation; 4. The method of claim 3, wherein, optimizing the coefficient of the second sub-thread based on the first quality evaluation to obtain the comparison thread. Before the coefficient of the second sub-thread is optimized based on the first quality evaluation to obtain the comparison thread, the method further comprises: determining the comparison between the second feature vector and a third feature vector to obtain a second comparison; the third feature vector is a feature vector output by a decision layer corresponding to the type of the configuration data in the second sub-thread; obtaining the second quality evaluation of the second sub-thread under the monitoring of the configuration data in combination with the second comparison; the second comparison is related to the second quality evaluation; obtaining the overall quality evaluation of the second sub-thread in combination with the first quality evaluation and the second quality evaluation; the overall quality evaluation is related to the first quality evaluation, and the overall quality evaluation is related to the second quality evaluation; The optimizing the coefficients of the second sub-thread based on the first quality evaluation to obtain the comparison thread comprises: optimizing the coefficients of the second sub-thread based on the overall quality evaluation to obtain the configured second sub-thread.

5. The method of claim 4, wherein, The combining the first quality evaluation and the second quality evaluation to obtain the overall quality evaluation of the second sub-thread comprises: combining the item involving the first comparison in the first quality evaluation and the item involving the second comparison in the second quality evaluation to obtain a third quality evaluation; the third quality evaluation has no connection with the first comparison, and the third quality evaluation has no connection with the second comparison; combining the third quality evaluation to obtain the overall quality evaluation; the third quality evaluation has no connection with the overall quality evaluation.

6. The method of claim 5, wherein, The combining the item involving the first comparison in the first quality evaluation and the item involving the second comparison in the second quality evaluation to obtain a third quality evaluation comprises: obtaining a first coefficient; the first coefficient is a confidence degree of the item involving the first comparison in a process of combining the item involving the first comparison and the item involving the second comparison to obtain the third quality evaluation; determining a function calculation between the first coefficient and the item involving the first comparison to obtain a first value; combining a sum between the item involving the second comparison and the first value to obtain the third quality evaluation; the third quality evaluation has no connection with the first value.

7. The method of claim 4, wherein, The combining the second comparison to obtain the second quality evaluation of the second sub-thread under the monitoring of the configuration data comprises: obtaining a second coefficient; the second coefficient has a connection with a convergence difficulty of the second quality evaluation; determining a sum of the second comparison and the second coefficient to obtain a second value; determining a function value of the second value as the second quality evaluation.

8. The method of claim 5, wherein, The combining the third quality evaluation to obtain the overall quality evaluation comprises: obtaining a third coefficient and a fourth coefficient; the third coefficient has no connection with a convergence difficulty of the overall quality evaluation; the fourth coefficient is greater than 1; determining a function calculation between the third coefficient and the third quality evaluation to obtain a third value; taking the fourth coefficient as a base number and the third value as an index to obtain a fourth value; combining the fourth value to obtain the overall quality evaluation; the overall quality evaluation has no connection with the fourth value.

9. The method of claim 3, wherein, The configuration data comprises local fault data; the first feature vector and the second feature vector are both local feature vectors of the first to-be-compared matter.

10. A coal belt pull cord switch failure monitoring system, characterized by, The processor and the memory are used to communicate with each other, the processor is used to read a computer program from the memory and execute the computer program to implement the method in any one of claims 1-9.

Images