Fault monitoring method and system for generator carbon brush
By establishing a carbon brush operation data monitoring system and fault analysis model in the generator, the operation data of carbon brushes is monitored and analyzed in real time, the problem of difficulty in time detection of faults in the existing technology is solved, and timely warning of carbon brush failures is achieved. Analysis has improved the operation stability of the generator.
Patent Information
- Application Number
- CN202410917715.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-10
- Publication Date
- 2024-11-08
- Estimated Expiration
- 2044-07-10
AI Technical Summary
The existing technology is difficult to detect and analyze generator carbon brush failures in a timely manner, resulting in performance degradation or failures that cannot be predicted in time.
Establish a generator carbon brush operation data monitoring system, monitor the carbon brush operation data through temperature sensors, current intensity sensors, displacement sensors and ring sensors in real time, extract fault characteristic variables, and build an initial fault analysis model, conduct accuracy training, and finally based on The optimal fault analysis model performs real-time fault monitoring and analysis.
It effectively solves the problem of timely discovering and analyzing and warning of carbon brush failures, improves the practicality of the generator's architectural functions, and ensures the normal operation of the carbon brush and the stability of the generator's performance.
Smart Images

Figure CN118503652B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of carbon brush fault monitoring, and in particular to a method and system for monitoring the fault of a generator carbon brush. Background Art
[0002] At present, the carbon brushes of the generator are also called brushes, which can add external excitation current to the rotor to achieve stable operation inside the generator. However, during the operation of the generator, the carbon brushes will change due to factors such as friction and wear, which will cause the performance of the generator to decline or fail. And because the carbon brushes are located inside the motor, the status of the carbon brushes cannot be checked in real time, so carbon brush failures cannot be discovered or predicted in time.
[0003] There is an urgent need for a method that can timely detect and analyze carbon brush failures and warn of them. Summary of the invention
[0004] To this end, the present invention provides a fault monitoring method and system for a generator carbon brush to solve the technical problem in the prior art that it is difficult to timely discover and analyze the early warning of carbon brush faults.
[0005] In order to achieve the above object, the present invention provides the following technical solutions:
[0006] A method for monitoring a generator carbon brush fault, the method specifically comprising:
[0007] Establish a generator carbon brush operation data monitoring system;
[0008] Based on the established generator carbon brush operation data monitoring system, the historical carbon brush operation data of the faulty generator is obtained, and the obtained historical carbon brush operation data is pre-screened to obtain the training carbon brush fault characteristic variables;
[0009] Establishing an initial carbon brush fault analysis model, using the training carbon brush fault characteristic variables as input elements of the initial carbon brush fault analysis model, training the initial carbon brush fault analysis model for accuracy, and obtaining the optimal carbon brush fault analysis model after the training is completed;
[0010] Continue to obtain the current carbon brush operation data in real time based on the generator carbon brush operation data monitoring system, input the original characteristic variables corresponding to the current carbon brush operation data into the optimal carbon brush fault analysis model, and analyze and judge based on the optimal carbon brush fault analysis model and output the fault monitoring analysis results of the current carbon brush operation data and the carbon brush group replacement analysis results.
[0011] On the basis of the above technical solution, the present invention is further described as follows:
[0012] As a further solution of the present invention, the establishment of a generator carbon brush operation data monitoring system specifically includes:
[0013] At least one set of temperature sensors is fixedly arranged based on the positioning brush holder inside the generator housing, and the temperature sensing probes of the at least one set of temperature sensors are respectively kept at a predetermined gap corresponding to the friction position between the carbon brush and the rotor, so as to monitor the operating temperature W of the two sets of carbon brushes respectively through the temperature sensors. m and W n ;
[0014] Based on the two groups of carbon brushes being connected in series with a current intensity sensor and / or an ammeter, the current intensity I flowing through the two groups of carbon brushes is respectively collected by the current intensity sensor and / or the ammeter. m and I n ;
[0015] Displacement sensors are respectively fixedly arranged at the middle positions of the positioning brush holders corresponding to the two groups of carbon brushes, so as to form a first brush holder monitoring point a1 and a second brush holder monitoring point a2 corresponding to the positioning brush holders of the two groups of carbon brushes;
[0016] Continue to fix the displacement sensors on both ends of the sliding base of the two groups of carbon brushes, so as to form a first carbon brush monitoring point b1 and a second carbon brush monitoring point c1 corresponding to both ends of the sliding base of the first group of carbon brushes, and form a third carbon brush monitoring point b2 and a fourth carbon brush monitoring point c2 corresponding to both ends of the sliding base of the second group of carbon brushes;
[0017] Based on the generator rotor fixed set ring position sensor, the ring displacement monitoring of the ring position sensor is fitted to the rotation center displacement monitoring point o of the rotor center, and the signal transmission connection is established between the slip ring and the electronic control component through the ring position sensor;
[0018] The rotor is controlled to be in a rotating state, and the distance change between the rotation center displacement monitoring point o and the first brush holder monitoring point a1 and the second brush holder monitoring point a2, as well as the angle change relative to the line connecting the initial points, are obtained in real time;
[0019] Alternatively, the angle change value of the first group of carbon brushes is obtained by comparing the real-time point connection line between the first carbon brush monitoring point b1 and the second carbon brush monitoring point c1 with the initial vertical connection line, and the angle change value of the second group of carbon brushes is obtained by comparing the real-time point connection line between the third carbon brush monitoring point b2 and the fourth carbon brush monitoring point c2 with the initial vertical connection line;
[0020] Then, the offset θ of the positioning brush holder and its assembled carbon brush corresponding to the first brush holder monitoring point a1 is calculated by combining the above-mentioned spacing change value and angle change value: m and / or the offset θ of the positioning brush holder and the carbon brush assembled thereon corresponding to the second brush holder monitoring point a2 n .
[0021] As a further solution of the present invention, the establishment of a generator carbon brush operation data monitoring system specifically includes:
[0022] The rotor is controlled to be in a rotating state, and the distance change between the midpoint of the real-time point connection line between the first carbon brush monitoring point b1 and the second carbon brush monitoring point c1 and the rotation center displacement monitoring point o compared with the initial distance is obtained in real time to obtain the wear amount L of the first group of carbon brushes m At the same time, the change value of the distance between the midpoint of the real-time point connection line between the third carbon brush monitoring point b2 and the fourth carbon brush monitoring point c2 and the rotation center displacement monitoring point o compared with the initial distance is obtained to obtain the wear amount L of the second group of carbon brushes n ;
[0023] At least two virtual wear monitoring lines d are divided based on the overall length of the carbon brush, and the wear areas of the carbon brushes corresponding to the positions between the at least two virtual wear monitoring lines d are set to be equal; when the monitoring positioning brush holder and the assembled carbon brushes do not produce an offset, and the wear position of the carbon brush is between the at least two virtual wear monitoring lines d, the contact areas between the carbon brushes and the rotor are obtained to be determined to be equal values;
[0024] When the monitoring positioning brush holder and its assembled carbon brushes produce an offset, the contact area is determined based on the set contact area and the angle change value of the first group of carbon brushes and / or the angle change value of the second group of carbon brushes are used as a function to calculate the contact area S between the first group of carbon brushes and the rotor. m And the contact area S between the second set of carbon brushes and the rotor n .
[0025] As a further solution of the present invention, the generator carbon brush operation data monitoring system established is used to obtain historical carbon brush operation data of the faulty generator, and pre-screen the acquired historical carbon brush operation data to obtain training carbon brush fault feature variables, which specifically include:
[0026] Acquire historical carbon brush operation data of the faulty generator, and extract a plurality of original characteristic variables based on the historical carbon brush operation data;
[0027] Calculating the correlation coefficients of the several original characteristic variables corresponding to the carbon brush failure; specifically, respectively calculating the standard deviations of the several original characteristic variables, and using the standard deviations as the correlation coefficients of the several original characteristic variables and the carbon brush failure based on the one-to-one correspondence between the several standard deviations;
[0028] Preset standard range values corresponding to several original feature variables, compare the correlation coefficients of several original feature variables with their corresponding standard range values, and extract the original feature variables whose correlation coefficients exceed the preset standard range values as training carbon brush fault feature variables;
[0029] The training carbon brush fault characteristic variable is one of a carbon brush offset variable, a carbon brush wear variable, a carbon brush contact area variable, a carbon brush temperature variable and a carbon brush current intensity variable.
[0030] As a further solution of the present invention, the initial carbon brush fault analysis model is established, the training carbon brush fault characteristic variables are used as input elements of the initial carbon brush fault analysis model, the initial carbon brush fault analysis model is trained for accuracy, and the optimal carbon brush fault analysis model is obtained after the training is completed, specifically including:
[0031] Establishing several groups of initial carbon brush fault analysis models, wherein the initial carbon brush fault analysis models include a CNN (convolutional neural network), a time block, and an output layer;
[0032] Extracting training carbon brush fault feature variables through CNN (convolutional neural network), extracting time feature T through the time block, and analyzing and predicting future carbon brush faults based on the time feature T and the training carbon brush fault feature variables through the output layer;
[0033] Acquire several groups of historical carbon brush operation data of faulty generators, and finally obtain several groups of training carbon brush fault characteristic variables corresponding to the several groups of faulty generators, each group of the training carbon brush fault characteristic variables includes several training carbon brush fault characteristic variables, that is, each group of the training carbon brush fault characteristic variables includes at least one of a carbon brush offset variable, a carbon brush wear variable, a carbon brush contact area variable, a carbon brush temperature variable and a carbon brush current intensity variable;
[0034] Continue to sort the time-varying value sets corresponding to the several training carbon brush fault characteristic variables in each group in descending order, and form several groups of training data sets based on the lists constructed after sorting the several groups of training carbon brush fault characteristic variables in a one-to-one correspondence;
[0035] Inputting several groups of the training data sets into several groups of the initial carbon brush fault analysis models one by one for accuracy training, and inputting at least two groups of normal data and at least two groups of proposed fault data again based on the several groups of initial carbon brush fault analysis models that have completed the training to obtain a normal score and a fault risk score Y, and comparing and verifying the fault risk score Y with the normal score and taking the average to obtain the average accuracy of the trained initial carbon brush fault analysis model;
[0036] The calculated initial carbon brush fault analysis model with the highest average accuracy is used as the optimal carbon brush fault analysis model, and the input elements of the optimal carbon brush fault analysis model include carbon brush offset variable, carbon brush wear variable, carbon brush contact area variable, carbon brush temperature variable and carbon brush current intensity variable.
[0037] As a further solution of the present invention, the method of analyzing and judging the fault monitoring analysis results of the current carbon brush operation data and the carbon brush set replacement analysis results based on the optimal carbon brush fault analysis model specifically includes:
[0038] The fault risk score Y is calculated based on the optimal carbon brush fault analysis model. The calculation formula is as follows:
[0039] (1)
[0040] Where Y is the fault risk score, T is the time characteristic, L=(L m +L n ) / 2 is the carbon brush wear, θ=(θ m +θ n ) / 2 is the carbon brush offset, W is the carbon brush operating temperature, I is the carbon brush current intensity, S= (S m +S n ) / 2 is the contact area between the carbon brush and the rotor;
[0041] Since the carbon brush offset does not necessarily occur during the operation of the carbon brush, when the carbon brush offset θ is within the preset standard range, the carbon brush offset θ is taken as 0;
[0042] At the same time, the carbon brush wear amount L, carbon brush operating temperature W, carbon brush current intensity I and carbon brush and rotor contact area S parameters must exist during the operation of the carbon brush. Therefore, when the carbon brush wear amount L and / or the carbon brush operating temperature W and / or the carbon brush current intensity I and / or the carbon brush and rotor contact area S are within the preset standard range value, the above-mentioned necessarily existing parameter values are taken as 1;
[0043] Continue to determine whether the calculated fault risk score Y exceeds the preset fault risk threshold, and if so, output the fault monitoring analysis result based on the fault risk score.
[0044] As a further solution of the present invention, the fault monitoring analysis result and the carbon brush set replacement analysis result based on the optimal carbon brush fault analysis model analysis and output of the current carbon brush operation data specifically also include:
[0045] Determine whether the calculated fault risk score Y exceeds the preset fault risk threshold, and if so, continue to calculate the carbon brush group replacement score Z based on the optimal carbon brush fault analysis model;
[0046] Calculate the wear imbalance of the carbon brush set L B , the calculation formula is as follows:
[0047] L B =|L m -L n | (2)
[0048] Where, L B is the wear imbalance of the carbon brush set, L m and L n They correspond to the wear amounts of two groups of carbon brushes respectively;
[0049] Calculate the brush set offset imbalance θ B , the calculation formula is as follows:
[0050] θ B =|θ m -θ n | (3)
[0051] In the formula, θ B is the offset imbalance of the carbon brush group, θ m and θ n They correspond to the offset of two groups of carbon brushes respectively;
[0052] Calculate the temperature imbalance W of the carbon brush group B , the calculation formula is as follows:
[0053] W B =|W m -W n | (4)
[0054] Where W B is the temperature imbalance of the carbon brush group, W m and W n They correspond to the detected temperature values of two groups of carbon brushes respectively;
[0055] Calculate the unbalanced degree of current intensity I of the carbon brush group B , the calculation formula is as follows:
[0056] I B =|I m -I n | (5)
[0057] In the formula, I B is the unbalanced degree of current intensity of the carbon brush group, I m and I n They correspond to the detection current intensity values of two groups of carbon brushes respectively;
[0058] Calculate the unbalance S of the contact area of the carbon brush group B , the calculation formula is as follows:
[0059] S B =|S m -S n | (6)
[0060] In the formula, S B S is the contact area imbalance of the carbon brush group. m and S nThey correspond to the contact area values of the two groups of carbon brushes respectively;
[0061] The formula for further calculating the carbon brush set replacement score Z is as follows:
[0062] (7)
[0063] Where Z is the replacement score of the carbon brush set, T is the time characteristic, L B is the wear imbalance of the carbon brush group, θ B is the offset imbalance of the carbon brush group, W B is the temperature imbalance of the carbon brush group, I B is the current imbalance of the carbon brush group, S B is the contact area imbalance of the carbon brush group;
[0064] Continue to determine whether the calculated carbon brush set replacement score Z exceeds the preset carbon brush set replacement threshold. If so, output a carbon brush set replacement analysis result based on the carbon brush set replacement score Z.
[0065] As a further solution of the present invention, it also includes:
[0066] Based on the fault monitoring analysis results and the carbon brush set replacement analysis results, the corresponding fault handling plan is obtained, and a fault prompt message is generated. If the fault prompt message is not processed within the preset time, the fault handling plan is sent to the corresponding fault handling personnel user end, and the fault handling personnel take fault handling measures based on the fault handling plan; specifically, the following measures are included:
[0067] A set of fault handling plans is configured for each fault type based on the fault monitoring and analysis results;
[0068] The fault handling plan includes user information for executing the fault handling;
[0069] When the fault prompt information is not processed within a preset time, a corresponding fault processing plan is determined based on the fault type, and user information for executing the fault processing is determined based on the fault processing plan;
[0070] Obtain fault analysis data corresponding to carbon brush fault data;
[0071] Construct a carbon brush fault analysis map based on fault monitoring analysis data and carbon brush set replacement analysis data, and integrate the optimal carbon brush fault analysis model and carbon brush fault analysis map into the carbon brush fault monitoring platform;
[0072] The carbon brush fault monitoring platform can respond to a fault search request, obtain a fault search term, obtain a corresponding search result based on the fault search term, and display the search result on the carbon brush fault monitoring page; at the same time, the carbon brush fault monitoring platform can respond to a knowledge card viewing request initiated by a user, and display a knowledge card corresponding to the target search result;
[0073] The knowledge card includes the fault type and the corresponding fault handling plan;
[0074] The fault analysis data and the carbon brush fault data are fused to obtain a data fusion result, and the carbon brush fault analysis graph is further adjusted in a targeted manner based on the data fusion result.
[0075] A fault monitoring system for executing the generator carbon brush fault monitoring method, comprising:
[0076] Carbon brush data monitoring module, used to establish generator carbon brush operation data monitoring system;
[0077] A pre-screening processing module is used to obtain historical carbon brush operation data of a faulty generator based on an established generator carbon brush operation data monitoring system, and to perform pre-screening processing on the acquired historical carbon brush operation data to obtain training carbon brush fault characteristic variables;
[0078] A model training module is used to establish an initial carbon brush fault analysis model, use the training carbon brush fault characteristic variables as input elements of the initial carbon brush fault analysis model, perform accuracy training on the initial carbon brush fault analysis model, and obtain the optimal carbon brush fault analysis model after the training is completed;
[0079] The model prediction module is used to continue to obtain the current carbon brush operation data in real time based on the generator carbon brush operation data monitoring system, input the original characteristic variables corresponding to the current carbon brush operation data into the optimal carbon brush fault analysis model, and analyze and judge based on the optimal carbon brush fault analysis model and output the fault monitoring analysis results of the current carbon brush operation data and the carbon brush group replacement analysis results.
[0080] An electronic device comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the method when executing the computer program.
[0081] The present invention has the following beneficial effects:
[0082] The fault monitoring method of the generator carbon brush in the present invention establishes a generator carbon brush operation data monitoring system; based on the established generator carbon brush operation data monitoring system, the historical carbon brush operation data of the faulty generator is obtained, and the obtained historical carbon brush operation data is pre-screened to obtain training carbon brush fault characteristic variables; an initial carbon brush fault analysis model is established, and the training carbon brush fault characteristic variables are used as input elements of the initial carbon brush fault analysis model, the initial carbon brush fault analysis model is trained for accuracy, and the optimal carbon brush fault analysis model is obtained after the training is completed; the current carbon brush operation data is continuously obtained in real time based on the generator carbon brush operation data monitoring system, the original characteristic variables corresponding to the current carbon brush operation data are input into the optimal carbon brush fault analysis model, and the fault monitoring analysis results of the current carbon brush operation data and the carbon brush group replacement analysis results are analyzed and judged based on the optimal carbon brush fault analysis model; the problem that it is difficult to timely discover and analyze the early warning of carbon brush faults in the prior art is effectively solved, and the practicality of the architecture function is improved. BRIEF DESCRIPTION OF THE DRAWINGS
[0083] In order to more clearly illustrate the implementation mode of the present invention or the technical solution in the prior art, the drawings required for the implementation mode or the description of the prior art will be briefly introduced below. The structures, proportions, sizes, etc. illustrated in this specification are only used to match the contents disclosed in the specification for people familiar with this technology to understand and read. Any structural modification, change in proportional relationship or adjustment of size should still fall within the scope of the technical contents disclosed in the present invention without affecting the effects and purposes that can be achieved by the present invention.
[0084] Figure 1 A schematic flow chart of a method for monitoring a generator carbon brush fault provided by an embodiment of the present invention.
[0085] Figure 2 One of the principle schematic diagrams of the carbon brush operation data monitoring system in the generator carbon brush fault monitoring method provided in an embodiment of the present invention.
[0086] Figure 3 The second schematic diagram of the principle of the carbon brush operation data monitoring system in the generator carbon brush fault monitoring method provided in the embodiment of the present invention.
[0087] Figure 4 The third principle schematic diagram of the carbon brush operation data monitoring system in the generator carbon brush fault monitoring method provided in the embodiment of the present invention.
[0088] Figure 5 A schematic diagram of the architecture of a generator carbon brush fault monitoring system provided in an embodiment of the present invention.
[0089] Figure 6 A schematic diagram of the physical structure of an electronic device provided by an embodiment of the present invention.
[0090] In the accompanying drawings, the components represented by the reference numerals are listed as follows:
[0091] Carbon brush data monitoring module 10; pre-screening processing module 20; model training module 30; model prediction module 40; fault processing module 50;
[0092] Electronic device 60: processor 601, memory 602, internal bus 603;
[0093] Generator housing 1; rotor 2; positioning brush holder 3; carbon brush 4; ring position sensor 5;
[0094] Rotational displacement monitoring point o;
[0095] The first brush holder monitoring point a1, the second brush holder monitoring point a2;
[0096] The first carbon brush monitoring point b1, the second carbon brush monitoring point c1, the third carbon brush monitoring point b2, the fourth carbon brush monitoring point c2, and the wear monitoring line d. DETAILED DESCRIPTION
[0097] The following is a description of the implementation of the present invention by specific embodiments. People familiar with the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. Obviously, the described embodiments are part of the embodiments of the present invention, not 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.
[0098] The terms such as "upper", "lower", "left", "right", and "middle" used in this specification are only for the convenience of description and are not intended to limit the scope of the invention. Changes or adjustments to their relative relationships should be regarded as within the scope of the invention without substantially changing the technical content.
[0099] Figure 1 A flow chart of a fault monitoring method for a generator carbon brush provided by an embodiment of the present invention is shown in FIG. Figure 1 As shown, the generator carbon brush fault monitoring method specifically includes the following steps:
[0100] S101: Establish a generator carbon brush operation data monitoring system;
[0101] The specific process is: please refer to Figure 2 At least one set of temperature sensors is fixedly arranged based on the positioning brush holder 3 inside the generator housing 1, and the temperature sensing probes of the at least one set of temperature sensors are respectively kept at a predetermined gap corresponding to the friction position between the carbon brush 4 and the rotor 2;
[0102] Alternatively, the operating temperature W of the two groups of carbon brushes 4 can be monitored separately on the outside of the generator housing 1 by using an infrared thermal imager. m and W n ;
[0103] Based on the two groups of carbon brushes 4 being connected in series with a current intensity sensor and / or an ammeter, the current intensity I flowing through the two groups of carbon brushes 4 is respectively collected by the current intensity sensor and / or the ammeter. m and I n ;
[0104] Please continue to refer to Figure 2 , based on the middle positions of the positioning brush holders 3 corresponding to the two groups of carbon brushes 4, the displacement sensors are respectively fixedly arranged to form the first brush holder monitoring point a1 and the second brush holder monitoring point a2 corresponding to the positioning brush holders 3 of the two groups of carbon brushes 4;
[0105] Continue to fix the displacement sensors on both ends of the sliding base of the two groups of carbon brushes 4, so as to form a first carbon brush monitoring point b1 and a second carbon brush monitoring point c1 corresponding to both ends of the sliding base of the first group of carbon brushes 4, and form a third carbon brush monitoring point b2 and a fourth carbon brush monitoring point c2 corresponding to both ends of the sliding base of the second group of carbon brushes 4;
[0106] Based on the fixed set ring position sensor 5 of the generator rotor 2, the ring displacement monitoring of the ring position sensor 5 is fitted to the rotation center displacement monitoring point o of the center of the rotor 2, and the signal transmission connection is established between the slip ring and the electronic control component through the ring position sensor 5;
[0107] Please refer to Figure 2 and Figure 3 , control the rotor 2 to be in a rotating state, and monitor in real time to obtain the distance change values between the rotation center displacement monitoring point o and the first brush holder monitoring point a1 and the second brush holder monitoring point a2, as well as the angle change values relative to the initial point connection line, and then calculate the offset θ of the positioning brush holder 3 and its assembled carbon brush 4 corresponding to the first brush holder monitoring point a1 by combining the above distance change values and angle change values m and / or the offset θ of the positioning brush holder 3 and the carbon brush 4 mounted thereon corresponding to the second brush holder monitoring point a2 n , or, the angle change value of the first group of carbon brushes 4 is obtained by comparing the real-time point connection line between the first carbon brush monitoring point b1 and the second carbon brush monitoring point c1 with the initial vertical connection line, and the angle change value of the second group of carbon brushes 4 is obtained by comparing the real-time point connection line between the third carbon brush monitoring point b2 and the fourth carbon brush monitoring point c2 with the initial vertical connection line;
[0108] Please refer to Figure 2 and Figure 4, control the rotor 2 to be in a rotating state, and monitor in real time to obtain the distance change value between the midpoint of the real-time point connection line between the first carbon brush monitoring point b1 and the second carbon brush monitoring point c1 and the rotation center displacement monitoring point o compared with the initial distance, and obtain the wear amount L of the first group of carbon brushes 4 m At the same time, the change value of the distance between the midpoint of the real-time point connection line between the third carbon brush monitoring point b2 and the fourth carbon brush monitoring point c2 and the rotation center displacement monitoring point o compared with the initial distance is obtained to obtain the wear amount L of the second group of carbon brushes 4 n ;
[0109] At least two virtual wear monitoring lines d are divided based on the overall length of the carbon brush 4, and the wear areas of the carbon brush 4 corresponding to the positions between the at least two virtual wear monitoring lines d are set to be equal; when the monitoring positioning brush holder 3 and the assembled carbon brush 4 do not produce an offset, and the wear position of the carbon brush 4 is between the at least two virtual wear monitoring lines d, the contact areas between the carbon brush 4 and the rotor 2 are obtained to be determined to be equal values;
[0110] When the monitoring positioning brush holder 3 and its assembled carbon brush 4 produce an offset, the contact area S between the first group of carbon brushes 4 and the rotor 2 is obtained by performing a function calculation based on the set contact area determination value and the angle change value of the first group of carbon brushes 4 and / or the angle change value of the second group of carbon brushes 4. m And the contact area S between the second group of carbon brushes 4 and the rotor 2 n ;
[0111] S102: based on the established generator carbon brush operation data monitoring system, obtaining historical carbon brush operation data of the faulty generator, and performing pre-screening processing on the obtained historical carbon brush operation data to obtain training carbon brush fault feature variables;
[0112] The specific process is: obtaining historical carbon brush operation data of the faulty generator, and extracting a number of original characteristic variables based on the historical carbon brush operation data;
[0113] Calculating the correlation coefficients of the several original characteristic variables corresponding to the carbon brush failure; specifically, respectively calculating the standard deviations of the several original characteristic variables, and using the standard deviations as the correlation coefficients of the several original characteristic variables and the carbon brush failure based on the one-to-one correspondence between the several standard deviations;
[0114] Preset standard range values corresponding to several original feature variables, compare the correlation coefficients of several original feature variables with their corresponding standard range values, and extract the original feature variables whose correlation coefficients exceed the preset standard range values as training carbon brush fault feature variables;
[0115] The training carbon brush fault characteristic variable is one of a carbon brush offset variable, a carbon brush wear variable, a carbon brush contact area variable, a carbon brush temperature variable and a carbon brush current intensity variable;
[0116] For example, among several original feature variables, the correlation coefficients of carbon brush wear feature, carbon brush offset feature, carbon brush temperature feature and carbon brush current intensity feature are 0.218, 0.122, 0.111 and 0.185 respectively, while the maximum value of the preset standard range is 0.1. At this time, the correlation coefficients of the above original feature variables all exceed the preset standard range value, so they are used as training carbon brush fault feature variables, that is, carbon brush wear variable, carbon brush offset variable, carbon brush temperature variable and carbon brush current intensity variable;
[0117] S103: establishing an initial carbon brush fault analysis model, using the training carbon brush fault characteristic variables as input elements of the initial carbon brush fault analysis model, performing accuracy training on the initial carbon brush fault analysis model, and obtaining an optimal carbon brush fault analysis model after the training is completed;
[0118] The specific process is: establishing several groups of initial carbon brush fault analysis models, wherein the initial carbon brush fault analysis models include CNN (convolutional neural network), time block and output layer;
[0119] Extracting training carbon brush fault feature variables through CNN (convolutional neural network), extracting time feature T through the time block, and analyzing and predicting future carbon brush faults based on the time feature T and the training carbon brush fault feature variables through the output layer;
[0120] Acquire several groups of historical carbon brush operation data of faulty generators, and finally obtain several groups of training carbon brush fault characteristic variables corresponding to the several groups of faulty generators, each group of the training carbon brush fault characteristic variables includes several training carbon brush fault characteristic variables, that is, each group of the training carbon brush fault characteristic variables includes at least one of a carbon brush offset variable, a carbon brush wear variable, a carbon brush contact area variable, a carbon brush temperature variable and a carbon brush current intensity variable;
[0121] Continue to sort the time-varying value sets corresponding to the several training carbon brush fault characteristic variables in each group in descending order, and form several groups of training data sets based on the lists constructed after sorting the several groups of training carbon brush fault characteristic variables in a one-to-one correspondence;
[0122] Inputting several groups of the training data sets into several groups of the initial carbon brush fault analysis models one by one for accuracy training, and inputting at least two groups of normal data and at least two groups of proposed fault data again based on the several groups of initial carbon brush fault analysis models that have completed the training to obtain a normal score and a fault risk score Y, and comparing and verifying the fault risk score Y with the normal score and taking the average to obtain the average accuracy of the trained initial carbon brush fault analysis model;
[0123] The calculated initial carbon brush fault analysis model with the highest average accuracy is used as the optimal carbon brush fault analysis model, wherein the input elements of the optimal carbon brush fault analysis model include a carbon brush offset variable, a carbon brush wear variable, a carbon brush contact area variable, a carbon brush temperature variable, and a carbon brush current intensity variable;
[0124] S104: continue to obtain the current carbon brush operation data in real time based on the generator carbon brush operation data monitoring system, input the original characteristic variables corresponding to the current carbon brush operation data into the optimal carbon brush fault analysis model, and analyze and judge based on the optimal carbon brush fault analysis model and output the fault monitoring analysis results of the current carbon brush operation data and the carbon brush group replacement analysis results;
[0125] At present, the magnitude of the current intensity I is closely related to the carbon brush operating temperature W and the carbon brush wear amount L; the magnitude of the current intensity I will have a direct impact on the carbon brush 4. A high current intensity I will cause the carbon brush 4 to wear faster. In the process of transmitting current, the carbon brush 4 will generate heat energy due to the friction on the rotor 2, which will cause the temperature of the carbon brush 4 to rise. At the same time, when the current passes through the carbon brush 4, energy loss will also occur, part of which is converted into heat energy, further causing the temperature of the carbon brush 4 to rise. The degree of wear of the carbon brush 4 and the contact area between the carbon brush 4 and the rotor 2 also have a great influence on the current intensity effect. When the degree of wear of the carbon brush 4 is high and the contact with the rotor 2 is poor, it will lead to increased energy loss, which will in turn lead to a decrease in the output power of the generator, affecting the overall effect of the motor. Experiments show that the smaller the contact area S between the carbon brush 4 and the rotor 2, the greater the contact resistance, and the resistance of the carbon brush 4 itself and the commutator copper block and winding are all fixed values, so the resistance of the entire generator series circuit will increase and the current will decrease.
[0126] Generally speaking, the normal wear of carbon brush 4 is 0.5 mm every 50 hours. Carbon brush 4 must be replaced when it is worn to one-third to one-quarter of its normal size. Carbon brush 4 must be replaced when the contact area between the rotor is less than 70% of the initial size.
[0127] Therefore, the fault monitoring analysis result and the carbon brush replacement analysis result based on the optimal carbon brush fault analysis model analysis and output of the current carbon brush operation data specifically include:
[0128] The fault risk score Y is calculated based on the optimal carbon brush fault analysis model. The calculation formula is as follows:
[0129] (1)
[0130] Where Y is the fault risk score, T is the time characteristic, L=(L m +L n ) / 2 is the carbon brush wear, θ=(θ m +θ n) / 2 is the carbon brush offset, W is the carbon brush operating temperature, I is the carbon brush current intensity, S= (S m +S n ) / 2 is the contact area between the carbon brush 4 and the rotor 2;
[0131] Since the carbon brush offset does not necessarily occur during the operation of the carbon brush, when the carbon brush offset θ is within the preset standard range, the carbon brush offset θ is taken as 0;
[0132] At the same time, the carbon brush wear amount L, the carbon brush operating temperature W, the carbon brush current intensity I and the contact area S between the carbon brush 4 and the rotor 2 parameters are necessarily present during the operation of the carbon brush. Therefore, when the carbon brush wear amount L and / or the carbon brush operating temperature W and / or the carbon brush current intensity I and / or the contact area S between the carbon brush 4 and the rotor 2 are within the preset standard range values, the above-mentioned necessarily existing parameter values are taken as 1;
[0133] Continue to determine whether the calculated fault risk score Y exceeds the preset fault risk threshold. If so, output the fault monitoring analysis result based on the fault risk score, and continue to calculate the carbon brush group replacement score Z based on the optimal carbon brush fault analysis model;
[0134] Specifically, calculate the wear imbalance of the carbon brush group L B , the calculation formula is as follows:
[0135] L B =|L m -L n | (2)
[0136] Where, L B is the wear imbalance of the carbon brush set, L m and L n They correspond to the wear amounts of two groups of carbon brushes 4 respectively;
[0137] Calculate the brush set offset imbalance θ B , the calculation formula is as follows:
[0138] θ B =|θ m -θ n | (3)
[0139] In the formula, θ B is the offset imbalance of the carbon brush group, θ m and θ n They correspond to the offsets of the two groups of carbon brushes 4 respectively;
[0140] Calculate the temperature imbalance W of the carbon brush group B , the calculation formula is as follows:
[0141] W B =|W m -Wn | (4)
[0142] Where W B is the temperature imbalance of the carbon brush group, W m and W n They correspond to the detected temperature values of two groups of carbon brushes respectively;
[0143] Calculate the unbalanced degree of current intensity I of the carbon brush group B , the calculation formula is as follows:
[0144] I B =|I m -I n | (5)
[0145] In the formula, I B is the unbalanced degree of current intensity of the carbon brush group, I m and I n They correspond to the detection current intensity values of two groups of carbon brushes respectively;
[0146] Calculate the unbalance S of the contact area of the carbon brush group B , the calculation formula is as follows:
[0147] S B =|S m -S n | (6)
[0148] In the formula, S B S is the contact area imbalance of the carbon brush group. m and S n They correspond to the contact area values of the two groups of carbon brushes respectively;
[0149] The formula for further calculating the carbon brush set replacement score Z is as follows:
[0150] (7)
[0151] Where Z is the replacement score of the carbon brush set, T is the time characteristic, L B is the wear imbalance of the carbon brush group, θ B is the offset imbalance of the carbon brush group, W B is the temperature imbalance of the carbon brush group, I B is the current imbalance of the carbon brush group, S B is the contact area imbalance of the carbon brush group;
[0152] Continue to determine whether the calculated carbon brush set replacement score Z exceeds the preset carbon brush set replacement threshold. If so, output a carbon brush set replacement analysis result based on the carbon brush set replacement score Z.
[0153] S105: Obtaining a corresponding fault handling plan based on the fault monitoring analysis results and the carbon brush set replacement analysis results, and generating fault prompt information. If the fault prompt information is not processed within a preset time, the fault handling plan is sent to the corresponding fault handling personnel user end, and the fault handling personnel take fault handling measures based on the fault handling plan;
[0154] The specific process is as follows: a set of fault handling plans is configured for each fault type based on the fault monitoring and analysis results, where:
[0155] The fault handling plan includes user information for executing the fault handling;
[0156] When the fault prompt information is not processed within a preset time, a corresponding fault processing plan is determined based on the fault type, and user information for executing the fault processing is determined based on the fault processing plan;
[0157] Obtain fault analysis data corresponding to carbon brush fault data;
[0158] Constructing a carbon brush fault analysis map based on the fault monitoring analysis data and the carbon brush set replacement analysis data, and integrating the optimal carbon brush fault analysis model and the carbon brush fault analysis map into a carbon brush fault monitoring platform;
[0159] The carbon brush fault monitoring platform can respond to a fault search request, obtain a fault search term, obtain a corresponding search result based on the fault search term, and display the search result on the carbon brush fault monitoring page; at the same time, the carbon brush fault monitoring platform can respond to a knowledge card viewing request initiated by a user, and display a knowledge card corresponding to the target search result;
[0160] The knowledge card includes the fault type and the corresponding fault handling plan.
[0161] The fault analysis data and the carbon brush fault data are fused to obtain a data fusion result, and the carbon brush fault analysis graph is further adjusted in a targeted manner based on the data fusion result.
[0162] Figure 5 FIG. 1 is a schematic diagram of an embodiment of a fault monitoring system for a generator carbon brush according to the present invention. Figure 5 As shown, an embodiment of the present invention provides a generator carbon brush fault monitoring system, comprising:
[0163] The carbon brush data monitoring module 10 is used to establish a generator carbon brush operation data monitoring system;
[0164] A pre-screening processing module 20 is used to obtain historical carbon brush operation data of a faulty generator based on an established generator carbon brush operation data monitoring system, and perform pre-screening processing on the obtained historical carbon brush operation data to obtain a training carbon brush fault characteristic variable;
[0165] A model training module 30 is used to establish an initial carbon brush fault analysis model, use the training carbon brush fault characteristic variables as input elements of the initial carbon brush fault analysis model, perform accuracy training on the initial carbon brush fault analysis model, and obtain an optimal carbon brush fault analysis model after the training is completed;
[0166] The model prediction module 40 is used to continue to obtain the current carbon brush operation data in real time based on the generator carbon brush operation data monitoring system, input the original characteristic variables corresponding to the current carbon brush operation data into the optimal carbon brush fault analysis model, and analyze and judge based on the optimal carbon brush fault analysis model and output the fault monitoring analysis results of the current carbon brush operation data and the carbon brush group replacement analysis results.
[0167] The fault handling module 50 is used to obtain the corresponding fault handling plan based on the fault monitoring analysis results and the carbon brush group replacement analysis results, and generate a fault prompt message. If the fault prompt message is not processed within a preset time, the fault handling plan is sent to the corresponding fault handling personnel user end, and the fault handling personnel takes fault handling measures based on the fault handling plan.
[0168] Figure 6 FIG. 1 is a schematic diagram of the physical structure of an electronic device according to an embodiment of the present invention. Figure 6 As shown, the electronic device 60 includes: a processor 601 (processor), a memory 602 (memory) and an internal bus 603;
[0169] The processor 601 and the memory 602 communicate with each other via the internal bus 603;
[0170] The processor 601 is used to call the program instructions in the memory 602 to execute the methods provided by the above-mentioned method embodiments, for example, including: establishing a generator carbon brush operation data monitoring system; based on the established generator carbon brush operation data monitoring system, obtaining the historical carbon brush operation data of the faulty generator, and pre-screening the obtained historical carbon brush operation data to obtain training carbon brush fault feature variables; establishing an initial carbon brush fault analysis model, using the training carbon brush fault feature variables as input elements of the initial carbon brush fault analysis model, training the initial carbon brush fault analysis model for accuracy, and obtaining the optimal carbon brush fault analysis model after the training is completed; continuing to obtain the current carbon brush operation data in real time based on the generator carbon brush operation data monitoring system, inputting the original feature variables corresponding to the current carbon brush operation data into the optimal carbon brush fault analysis model, and analyzing and judging based on the optimal carbon brush fault analysis model and outputting the fault monitoring analysis results of the current carbon brush operation data and the carbon brush group replacement analysis results.
[0171] The present embodiment provides a non-transitory computer-readable storage medium, which stores computer instructions. The computer instructions enable the computer to execute the methods provided by the above-mentioned method embodiments, for example, including: establishing a generator carbon brush operation data monitoring system; based on the established generator carbon brush operation data monitoring system, obtaining historical carbon brush operation data of the faulty generator, and pre-screening the obtained historical carbon brush operation data to obtain training carbon brush fault feature variables; establishing an initial carbon brush fault analysis model, using the training carbon brush fault feature variables as input elements of the initial carbon brush fault analysis model, performing accuracy training on the initial carbon brush fault analysis model, and obtaining an optimal carbon brush fault analysis model after the training is completed; continuing to obtain current carbon brush operation data in real time based on the generator carbon brush operation data monitoring system, inputting the original feature variables corresponding to the current carbon brush operation data into the optimal carbon brush fault analysis model, and analyzing and judging based on the optimal carbon brush fault analysis model and outputting the fault monitoring analysis results of the current carbon brush operation data and the carbon brush group replacement analysis results.
[0172] A person skilled in the art can understand that all or part of the steps of implementing the above method embodiment can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above method embodiment; and the aforementioned storage medium includes: ROM, RAM, magnetic disk or optical disk, etc., various storage media that can store program codes.
[0173] The device embodiments described above are merely illustrative, wherein the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, i.e., they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Those of ordinary skill in the art may understand and implement it without creative effort.
[0174] Through the description of the above implementation methods, those skilled in the art can clearly understand that each implementation method can be implemented by means of software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solution is essentially or the part that contributes to the prior art can be embodied in the form of a software product, and the computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a disk, an optical disk, etc., including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods of each embodiment or some parts of the embodiment.
[0175] Although the present invention has been described in detail above by general description and specific embodiments, it is obvious to those skilled in the art that some modifications or improvements can be made to the present invention. Therefore, these modifications or improvements made without departing from the spirit of the present invention all belong to the scope of protection claimed by the present invention.
Claims
1. A method for monitoring faults of carbon brushes of a generator, characterized in that: The method specifically comprises: Establish a generator carbon brush operation data monitoring system; Based on the established generator carbon brush operation data monitoring system, the historical carbon brush operation data of the faulty generator is obtained, and the obtained historical carbon brush operation data is pre-screened to obtain the training carbon brush fault characteristic variables; Establishing an initial carbon brush fault analysis model, using the training carbon brush fault characteristic variables as input elements of the initial carbon brush fault analysis model, training the initial carbon brush fault analysis model for accuracy, and obtaining the optimal carbon brush fault analysis model after the training is completed; Continue to obtain the current carbon brush operation data in real time based on the generator carbon brush operation data monitoring system, input the original characteristic variables corresponding to the current carbon brush operation data into the optimal carbon brush fault analysis model, and analyze and judge based on the optimal carbon brush fault analysis model and output the fault monitoring analysis results of the current carbon brush operation data and the carbon brush group replacement analysis results, specifically including: The fault risk score Y is calculated based on the optimal carbon brush fault analysis model. The calculation formula is as follows: (1) Where Y is the fault risk score, T is the time characteristic, L=(L m +L n ) / 2 is the carbon brush wear, θ=(θ m +θ n ) / 2 is the carbon brush offset, W is the carbon brush operating temperature, I is the carbon brush current intensity, S= (S m +S n ) / 2 is the contact area between the carbon brush and the rotor; Since the carbon brush offset does not necessarily occur during the operation of the carbon brush, when the carbon brush offset θ is within the preset standard range, the carbon brush offset θ is taken as 0; At the same time, the carbon brush wear amount L, carbon brush operating temperature W, carbon brush current intensity I and carbon brush and rotor contact area S parameters must exist during the operation of the carbon brush. Therefore, when the carbon brush wear amount L and / or the carbon brush operating temperature W and / or the carbon brush current intensity I and / or the carbon brush and rotor contact area S are within the preset standard range value, the above-mentioned necessarily existing parameter values are taken as 1; Continue to determine whether the calculated fault risk score Y exceeds a preset fault risk threshold, and if so, output a fault monitoring analysis result based on the fault risk score; Determine whether the calculated fault risk score Y exceeds the preset fault risk threshold, and if so, continue to calculate the carbon brush group replacement score Z based on the optimal carbon brush fault analysis model; Calculate the wear imbalance of the carbon brush set L B , the calculation formula is as follows: L B =|L m -L n | (2) Where, L B is the wear imbalance of the carbon brush set, L m and L n They correspond to the wear amounts of two groups of carbon brushes respectively; Calculate the brush set offset imbalance θ B , the calculation formula is as follows: i B =|θ m -θ n | (3) In the formula, θ B is the offset imbalance of the carbon brush group, θ m and θ n They correspond to the offset of two groups of carbon brushes respectively; Calculate the temperature imbalance W of the carbon brush group B , the calculation formula is as follows: IN B =|In m -IN n | (4) Where W B is the temperature imbalance of the carbon brush group, W m and W n They correspond to the detected temperature values of two groups of carbon brushes respectively; Calculate the unbalance degree of the carbon brush group current intensity l B , the calculation formula is as follows: l B =|I m -l n | (5) In the formula, l B is the unbalanced degree of current intensity of the carbon brush group, l m and l n They correspond to the detection current intensity values of two groups of carbon brushes respectively; Calculate the unbalance S of the contact area of the carbon brush group B , the calculation formula is as follows: S B =|S m -S n | (6) In the formula, S B S is the contact area imbalance of the carbon brush group. m and S n They correspond to the contact area values of the two groups of carbon brushes respectively; The formula for further calculating the carbon brush set replacement score Z is as follows: (7) Where Z is the replacement score of the carbon brush set, T is the time characteristic, L B is the wear imbalance of the carbon brush group, θ B is the offset imbalance of the carbon brush group, W B is the temperature imbalance of the carbon brush group, I B is the current imbalance of the carbon brush group, S B is the contact area imbalance of the carbon brush group; Continue to determine whether the calculated carbon brush set replacement score Z exceeds the preset carbon brush set replacement threshold. If so, output a carbon brush set replacement analysis result based on the carbon brush set replacement score Z.
2. The fault monitoring method of generator carbon brush according to claim 1, characterized in that: The establishment of a generator carbon brush operation data monitoring system specifically includes: At least one set of temperature sensors is fixedly arranged based on the positioning brush holder inside the generator housing, and the temperature sensing probes of the at least one set of temperature sensors are respectively kept at a predetermined gap corresponding to the friction position between the carbon brush and the rotor, so as to monitor the operating temperature W of the two sets of carbon brushes respectively through the temperature sensors. m and W n ; Based on the two groups of carbon brushes being connected in series with a current intensity sensor and / or an ammeter, the current intensity I flowing through the two groups of carbon brushes is respectively collected by the current intensity sensor and / or the ammeter. m and I n ; Displacement sensors are respectively fixedly arranged at the middle positions of the positioning brush holders corresponding to the two groups of carbon brushes, so as to form a first brush holder monitoring point a1 and a second brush holder monitoring point a2 corresponding to the positioning brush holders of the two groups of carbon brushes; Continue to fix the displacement sensors on both ends of the sliding base of the two groups of carbon brushes, so as to form a first carbon brush monitoring point b1 and a second carbon brush monitoring point c1 corresponding to both ends of the sliding base of the first group of carbon brushes, and form a third carbon brush monitoring point b2 and a fourth carbon brush monitoring point c2 corresponding to both ends of the sliding base of the second group of carbon brushes; Based on the generator rotor fixed set ring position sensor, the ring displacement monitoring of the ring position sensor is fitted to the rotation center displacement monitoring point o of the rotor center, and the signal transmission connection is established between the slip ring and the electronic control component through the ring position sensor; The rotor is controlled to be in a rotating state, and the distance change between the rotation center displacement monitoring point o and the first brush holder monitoring point a1 and the second brush holder monitoring point a2, as well as the angle change relative to the line connecting the initial points, are obtained in real time; Alternatively, the angle change value of the first group of carbon brushes is obtained by comparing the real-time point connection line between the first carbon brush monitoring point b1 and the second carbon brush monitoring point c1 with the initial vertical connection line, and the angle change value of the second group of carbon brushes is obtained by comparing the real-time point connection line between the third carbon brush monitoring point b2 and the fourth carbon brush monitoring point c2 with the initial vertical connection line; Then, the offset θ of the positioning brush holder and its assembled carbon brush corresponding to the first brush holder monitoring point a1 is calculated by combining the above-mentioned spacing change value and angle change value: m and / or the offset θ of the positioning brush holder and the carbon brush assembled thereon corresponding to the second brush holder monitoring point a2 n .
3. The fault monitoring method of generator carbon brush according to claim 2, characterized in that: The establishment of a generator carbon brush operation data monitoring system specifically includes: The rotor is controlled to be in a rotating state, and the distance change between the midpoint of the real-time point connection line between the first carbon brush monitoring point b1 and the second carbon brush monitoring point c1 and the rotation center displacement monitoring point o compared with the initial distance is obtained in real time to obtain the wear amount L of the first group of carbon brushes m At the same time, the change value of the distance between the midpoint of the real-time point connection line between the third carbon brush monitoring point b2 and the fourth carbon brush monitoring point c2 and the rotation center displacement monitoring point o compared with the initial distance is obtained to obtain the wear amount L of the second group of carbon brushes n ; At least two virtual wear monitoring lines d are divided based on the overall length of the carbon brush, and the wear areas of the carbon brushes corresponding to the positions between the at least two virtual wear monitoring lines d are set to be equal; when the monitoring positioning brush holder and the assembled carbon brushes do not produce an offset, and the wear position of the carbon brush is between the at least two virtual wear monitoring lines d, the contact areas between the carbon brushes and the rotor are obtained to be determined to be equal values; When the monitoring positioning brush holder and its assembled carbon brushes produce an offset, the contact area is determined based on the set contact area and the angle change value of the first group of carbon brushes and / or the angle change value of the second group of carbon brushes are used as a function to calculate the contact area S between the first group of carbon brushes and the rotor. m And the contact area S between the second set of carbon brushes and the rotor n .
4. The fault monitoring method of generator carbon brush according to claim 3, characterized in that: The generator carbon brush operation data monitoring system established above obtains historical carbon brush operation data of the faulty generator, and pre-screens the acquired historical carbon brush operation data to obtain training carbon brush fault feature variables, which specifically include: Acquire historical carbon brush operation data of the faulty generator, and extract a plurality of original characteristic variables based on the historical carbon brush operation data; Calculating the correlation coefficients of the several original characteristic variables corresponding to the carbon brush failure; specifically, respectively calculating the standard deviations of the several original characteristic variables, and using the standard deviations as the correlation coefficients of the several original characteristic variables and the carbon brush failure based on the one-to-one correspondence between the several standard deviations; Preset standard range values corresponding to several original feature variables, compare the correlation coefficients of several original feature variables with their corresponding standard range values, and extract the original feature variables whose correlation coefficients exceed the preset standard range values as training carbon brush fault feature variables; The training carbon brush fault characteristic variable is one of a carbon brush offset variable, a carbon brush wear variable, a carbon brush contact area variable, a carbon brush temperature variable and a carbon brush current intensity variable.
5. The fault monitoring method of generator carbon brush according to claim 4, characterized in that: The initial carbon brush fault analysis model is established, the training carbon brush fault characteristic variables are used as input elements of the initial carbon brush fault analysis model, the initial carbon brush fault analysis model is trained for accuracy, and the optimal carbon brush fault analysis model is obtained after the training is completed, specifically including: Establishing several groups of initial carbon brush fault analysis models, wherein the initial carbon brush fault analysis models include a convolutional neural network, a time block, and an output layer; Extracting training carbon brush fault feature variables through a convolutional neural network, extracting time features T through the time block, and analyzing and predicting future carbon brush faults based on the time features T and the training carbon brush fault feature variables through the output layer; Acquire several groups of historical carbon brush operation data of faulty generators, and finally obtain several groups of training carbon brush fault characteristic variables corresponding to the several groups of faulty generators, each group of the training carbon brush fault characteristic variables includes several training carbon brush fault characteristic variables, that is, each group of the training carbon brush fault characteristic variables includes at least one of a carbon brush offset variable, a carbon brush wear variable, a carbon brush contact area variable, a carbon brush temperature variable and a carbon brush current intensity variable; Continue to sort the time-varying value sets corresponding to the several training carbon brush fault characteristic variables in each group in descending order, and form several groups of training data sets based on the lists constructed after sorting the several groups of training carbon brush fault characteristic variables in a one-to-one correspondence; Inputting several groups of the training data sets into several groups of the initial carbon brush fault analysis models one by one for accuracy training, and inputting at least two groups of normal data and at least two groups of proposed fault data again based on the several groups of initial carbon brush fault analysis models that have completed the training to obtain a normal score and a fault risk score Y, and comparing and verifying the fault risk score Y with the normal score and taking the average to obtain the average accuracy of the trained initial carbon brush fault analysis model; The calculated initial carbon brush fault analysis model with the highest average accuracy is used as the optimal carbon brush fault analysis model, and the input elements of the optimal carbon brush fault analysis model include carbon brush offset variable, carbon brush wear variable, carbon brush contact area variable, carbon brush temperature variable and carbon brush current intensity variable.
6. The fault monitoring method of generator carbon brush according to claim 5, characterized in that: Also includes: Obtain the corresponding fault handling plan based on the fault monitoring analysis results and the carbon brush set replacement analysis results, and generate fault prompt information. If the fault prompt information is not processed within the preset time, the fault handling plan will be sent to the corresponding fault handling personnel user end, and the fault handling personnel will take fault handling measures based on the fault handling plan; Specifically include: A set of fault handling plans is configured for each fault type based on the fault monitoring and analysis results; The fault handling plan includes user information for executing the fault handling; When the fault prompt information is not processed within a preset time, a corresponding fault processing plan is determined based on the fault type, and user information for executing the fault processing is determined based on the fault processing plan; Obtain fault analysis data corresponding to carbon brush fault data; Construct a carbon brush fault analysis map based on fault monitoring analysis data and carbon brush set replacement analysis data, and integrate the optimal carbon brush fault analysis model and carbon brush fault analysis map into the carbon brush fault monitoring platform; The carbon brush fault monitoring platform can respond to a fault search request, obtain a fault search term, obtain a corresponding search result based on the fault search term, and display the search result on the carbon brush fault monitoring page; at the same time, the carbon brush fault monitoring platform can respond to a knowledge card viewing request initiated by a user, and display a knowledge card corresponding to the target search result; The knowledge card includes the fault type and the corresponding fault handling plan; The fault analysis data and the carbon brush fault data are fused to obtain a data fusion result, and the carbon brush fault analysis graph is further adjusted in a targeted manner based on the data fusion result.
7. A fault monitoring system for executing the fault monitoring method for a generator carbon brush as claimed in claim 6, characterized in that: The fault monitoring system comprises: Carbon brush data monitoring module, used to establish generator carbon brush operation data monitoring system; A pre-screening processing module is used to obtain historical carbon brush operation data of a faulty generator based on an established generator carbon brush operation data monitoring system, and to perform pre-screening processing on the acquired historical carbon brush operation data to obtain training carbon brush fault characteristic variables; A model training module is used to establish an initial carbon brush fault analysis model, use the training carbon brush fault characteristic variables as input elements of the initial carbon brush fault analysis model, perform accuracy training on the initial carbon brush fault analysis model, and obtain the optimal carbon brush fault analysis model after the training is completed; The model prediction module is used to continue to obtain the current carbon brush operation data in real time based on the generator carbon brush operation data monitoring system, input the original characteristic variables corresponding to the current carbon brush operation data into the optimal carbon brush fault analysis model, and analyze and judge based on the optimal carbon brush fault analysis model and output the fault monitoring analysis results of the current carbon brush operation data and the carbon brush group replacement analysis results.
8. An electronic 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 method according to any one of claims 1 to 6 is implemented.
