Online monitoring method and system for health state of carbon brush of motor

Abstract

The invention relates to the technical field of motor testing, in particular to an online monitoring method and system for the health state of a carbon brush of a motor, and the method comprises the steps: determining a target operation load environment cluster of a target carbon brush at the current moment according to the pre-obtained load condition and operation environment condition of the target carbon brush in the current operation time period; screening out a reference operation time period; determining an abnormal loss factor and a potential abnormal wear feature vector of the target carbon brush at the current moment; according to the Lyapunov index of the vibration signal of the target carbon brush in the current operation time period and the temperature change and the current change of the target carbon brush in the current operation time period, determining a wear fault display possible feature vector; therefore, the health state of the target carbon brush at the current moment is determined. According to the invention, the potential abnormal wear condition of the target carbon brush at the current moment is considered, the health state of the carbon brush is monitored, and the timeliness of abnormal health state monitoring of the carbon brush is improved.

Description

Technical Field

[0001] This invention relates to the field of motor testing technology, specifically to an online monitoring method and system for the carbon brush health status of a motor. Background Technology

[0002] Brushed motors, with their advantages of simple structure, low cost, and convenient control, are widely used in industrial equipment, transportation, and household appliances. As a core component, the carbon brush plays a crucial role in current conduction and commutation. During continuous operation, the carbon brush and commutator are prone to wear and aging due to mechanical friction, which can lead to poor contact, abnormal sparking, reduced efficiency, or even shutdown. Therefore, real-time monitoring of the carbon brush's health status is of great significance for ensuring stable motor operation, extending equipment life, and reducing maintenance costs. Currently, the common method for monitoring the health status of carbon brushes is to determine whether the carbon brush's operating state is abnormal by checking for abnormal sparking. If abnormal sparking is present, especially circular sparking, the carbon brush's health status is considered abnormal.

[0003] However, when determining whether the carbon brush is in abnormal operation by checking for abnormal sparks, the following technical problems often arise: In practice, abnormal sparks are a clear signal of a fault in the carbon brush or motor system. They can rapidly accelerate the burning of the carbon brush and commutator, often requiring immediate attention. The detection of abnormal sparks usually indicates that the carbon brush has already experienced abnormal wear, which may have already caused localized damage to the motor. Therefore, when monitoring the health status of carbon brushes based on the presence of abnormal sparks, it is often difficult to detect potential wear risks, resulting in poor timeliness of monitoring abnormal carbon brush health status. Summary of the Invention

[0004] To address the technical problem of poor timeliness in monitoring abnormal carbon brush health status, this invention proposes an online monitoring method and system for the health status of motor carbon brushes.

[0005] In a first aspect, the present invention provides an online monitoring method for the health status of carbon brushes in an electric motor, the method comprising: Based on the pre-obtained load and operating environment of the target carbon brush during its current running period, the target operating load environment cluster of the target carbon brush at the current moment is determined, where the current moment is the end moment of the current running period; Select the historical runtime segments belonging to the target operating load environment cluster from all historical runtime segments of all historical carbon brushes obtained in advance, and use them as reference runtime segments; Based on the total wear of the target carbon brush in the current operating period and the total wear of historical carbon brushes in their reference operating periods, the abnormal wear factor of the target carbon brush at the current moment is determined. Based on the pre-acquired current ripple time-domain signal of the target carbon brush during the current operating period, as well as its clamping force and abnormal wear factor at the current moment, the potential abnormal wear feature vector of the target carbon brush at the current moment is determined. Based on the Lyapunov exponent of the vibration signal of the target carbon brush during the current operating period, and the temperature and current changes of the target carbon brush during the current operating period, the possible feature vector of wear failure manifestation of the target carbon brush at the current moment is determined. Based on the potential abnormal wear feature vector and the wear failure manifestation feature vector, the health status of the target carbon brush at the current moment is determined.

[0006] In conjunction with the first aspect above, in one possible implementation, determining the target operating load environment cluster of the target carbon brush at the current moment based on the pre-acquired load and operating environment conditions during the current operating period includes: Based on the load rate and dimensional data of different preset environmental dimensions at all times in each historical runtime segment of each historical carbon brush, a load environment feature vector for each historical carbon brush under each historical runtime segment is constructed. Based on the load environment feature vectors of all historical carbon brushes under all historical running periods, clusters are formed for all historical running periods of all historical carbon brushes to obtain initial clusters. Based on the load rate and dimensional data of the target carbon brush at all times during the current running period and under different preset environmental dimensions, a load environment feature vector of the target carbon brush under the current running period is constructed as the current load environment feature vector. The Euclidean distance between the current load environment feature vector and the cluster center of each initial cluster is determined as the reference distance for each initial cluster. The initial cluster with the smallest reference distance is selected from all initial clusters and used as the target operating load environment cluster for the target carbon brush at the current moment.

[0007] In conjunction with the first aspect above, in one possible implementation, the step of constructing a load environment feature vector for each historical brush in each historical runtime segment based on the pre-acquired load rate at all times within each historical runtime segment of each historical brush and dimensional data under different preset environmental dimensions includes: Any historical carbon brush is identified as a marked historical carbon brush, and any historical running segment of the marked historical carbon brush is identified as a marked historical running segment. The time series of load rates of the marked historical carbon brushes collected at all times within the marked historical running period is denoted as the load rate sequence corresponding to the marked historical running period. The time series of dimensional data of the marked historical carbon brushes collected at all times within the marked historical runtime period under the same preset environmental dimension is denoted as the dimensional data sequence of the marked historical runtime period under the preset environmental dimension. The mean, standard deviation, and number of maxima of the load rate sequence corresponding to the marked historical running segment, as well as the mean of the dimensional data sequence of the marked historical running segment under the same preset environmental dimension, constitute the load environment feature vector of the marked historical brush under the marked historical running segment.

[0008] In conjunction with the first aspect above, in one possible implementation, determining the abnormal wear factor of the target carbon brush at the current moment based on the pre-acquired total carbon brush wear of the target carbon brush during the current operating period and the total carbon brush wear of historical carbon brushes during their reference operating periods includes: The difference between the length of the target carbon brush at the start time and its length at the end time of the current running segment is determined as the total wear of the target carbon brush in the current running segment, and is used as the current total wear of the carbon brush. Similarly, the difference between the length of the historical carbon brush at the beginning of its reference operating period and its length at the end of its reference operating period is determined as the total wear of the historical carbon brush during its reference operating period, and is used as the reference total wear of the carbon brush. Determine the probability density function of the overall wear of all reference carbon brushes, and take the overall wear of the reference carbon brush corresponding to the maximum function value of the probability density function as the overall wear of the standard carbon brush. The difference between the current total wear of the carbon brush and the total wear of the standard carbon brush is normalized to obtain the initial anomaly factor of the target carbon brush at the current moment; The sum of constant 1 and the initial anomaly factor is determined as the anomaly loss factor of the target carbon brush at the current moment.

[0009] In conjunction with the first aspect above, in one possible implementation, determining the potential abnormal wear feature vector of the target carbon brush at the current moment based on the pre-acquired current ripple time-domain signal of the target carbon brush during the current operating period and its clamping force and abnormal wear factor at the current moment includes: The absolute value of the difference between the clamping force of the target carbon brush at the current moment and the pre-acquired standard clamping force is normalized to obtain the force deviation factor; The sum of constant 1 and the force deviation factor is determined as the target force deviation index; The product of the target force deviation index and the abnormal loss factor is determined as the abnormal amplification factor; Obtain the current ripple time-domain signal of the target carbon brush during the current operating period; The current ripple time-domain signal is decomposed into wavelet packet decomposition, which divides the signal into 8 characteristic frequency bands. The energy proportion of each characteristic frequency band is determined. The characteristic frequency bands are ordered, and the earlier the characteristic frequency band is, the higher the frequency. The sum of the preset positive factor and the energy proportion of each characteristic frequency band is used to determine the energy representative value corresponding to each characteristic frequency band; The target performance value of micro-discharge intensity is determined based on the abnormal amplification factor and the energy representative values ​​corresponding to the first three characteristic frequency bands. The abnormal index of carbon brush elastic deformation recovery is determined based on the sum of the energy representative values ​​corresponding to the fourth and fifth characteristic frequency bands, and the abnormal amplification factor. The material carbonization anomaly index is determined based on the ratio between the energy representative value corresponding to the 6th characteristic frequency band and the energy representative value corresponding to the 7th characteristic frequency band, as well as the anomaly amplification factor. The target micro-discharge intensity performance value, the abnormal index of carbon brush elastic deformation recovery, and the abnormal index of material carbonization are used to construct the potential abnormal wear feature vector of the target carbon brush at the current moment.

[0010] In conjunction with the first aspect above, in one possible implementation, determining the target performance value of the micro-discharge intensity based on the abnormal amplification factor and the energy representative values ​​corresponding to the first three characteristic frequency bands includes: The sum of the energy representative value corresponding to the first characteristic frequency band and the energy representative value corresponding to the second characteristic frequency band is determined as the normal contact friction current ripple performance value. The ratio of the energy representative value corresponding to the third characteristic frequency band to the normal contact friction current ripple performance value is determined as the initial performance value of the micro-discharge intensity. The product of the abnormal amplification factor and the initial performance value of the micro-discharge intensity is determined as the target performance value of the micro-discharge intensity.

[0011] In conjunction with the first aspect above, in one possible implementation, determining the wear fault manifestation potential feature vector of the target carbon brush at the current moment based on the pre-acquired Lyapunov exponent of the vibration signal of the target carbon brush during the current operating period, and the pre-acquired temperature and current changes of the target carbon brush during the current operating period, includes: The vibration signal of the target carbon brush during the current operating period is acquired, and the surface temperature, internal temperature and current value of the target carbon brush at each moment during the current operating period are acquired. The difference between the surface temperature and the internal temperature of the target carbon brush at each moment during the current operating period is determined as the initial temperature difference factor of the target carbon brush at each moment during the current operating period. The difference between the initial temperature difference factors of the target carbon brush at each adjacent time point in the current operating period is determined as the temperature difference change between each adjacent time point, thus obtaining a temperature difference change sequence. The absolute value of the difference between the current values ​​of the target carbon brush at each adjacent time point in the current operating period is determined as the current change between each adjacent time point, thus obtaining a current change sequence. Based on the temperature difference change and current change between adjacent moments in the current operating period of the target carbon brush, the target change index between adjacent moments is determined, and the target change index sequence is obtained. The temperature rise-current coupling effect value is determined based on the temperature difference change sequence, the current change sequence, and the target change index sequence. The temperature rise current coupling effect value and the maximum Lyapunov exponent of the vibration signal of the target carbon brush during the current operating period are used to construct the wear fault manifestation feature vector of the target carbon brush at the current moment.

[0012] In conjunction with the first aspect above, in one possible implementation, determining the temperature rise-current coupling effect value based on the temperature difference change sequence, the current change sequence, and the target change index sequence includes: By performing linear fitting on the target change index sequence, a target change line is obtained; The temperature rise-current coupling effect value is determined based on the mean of the temperature difference change sequence, the mean of the current change sequence, and the slope of the target change line.

[0013] In conjunction with the first aspect above, in one possible implementation, determining the health status of the target carbon brush at the current moment based on the potential abnormal wear feature vector and the wear fault manifestation potential feature vector includes: Normalize the cumulative product of all elements in the potential abnormal wear feature vector and the wear fault manifestation possible feature vector to obtain the target abnormal state index of the target carbon brush at the current moment. If the target abnormal state index is greater than the preset abnormal threshold, it is determined that the target carbon brush has a potential abnormality or a real abnormality at the current moment. If the target abnormal state index is less than or equal to the preset abnormal threshold, the target carbon brush is determined to be in good health at the current moment.

[0014] Secondly, the present invention provides an online monitoring system for the carbon brush health status of an electric motor, the system comprising: The target operating load environment cluster acquisition module is used to determine the target operating load environment cluster of the target carbon brush at the current moment based on the load and operating environment of the target carbon brush in the current running segment, which are obtained in advance. The runtime segment filtering module is used to filter out the historical runtime segments belonging to the target runtime environment cluster from all the historical runtime segments of all historical carbon brushes obtained in advance, and use them as reference runtime segments; The abnormal wear factor determination module is used to determine the abnormal wear factor of the target carbon brush at the current moment based on the total wear amount of the target carbon brush in the current running segment and the total wear amount of the historical carbon brush in its reference running segment. The potential abnormal wear feature vector acquisition module is used to determine the potential abnormal wear feature vector of the target carbon brush at the current moment based on the pre-acquired current ripple time domain signal of the target carbon brush in the current operating period and its clamping force and abnormal wear factor at the current moment. The wear fault manifestation potential feature vector acquisition module is used to determine the wear fault manifestation potential feature vector of the target carbon brush at the current moment based on the Lyapunov exponent of the vibration signal of the target carbon brush in the current operating period, as well as the temperature change and current change of the target carbon brush in the current operating period. The health status monitoring module is used to determine the health status of the target carbon brush at the current moment based on the potential abnormal wear feature vector and the wear failure manifestation feature vector.

[0015] Thirdly, a server is provided, including a memory and a processor. The memory is used to store executable program code, and the processor is used to call and run the executable program code from the memory, causing the device to perform the methods of the first aspect or any possible implementation thereof.

[0016] Fourthly, a computer program product is provided, comprising: computer program code, which, when run on a computer, causes the computer to perform the methods described in the first aspect or any possible implementation thereof.

[0017] Fifthly, a computer-readable storage medium is provided that stores computer program code, which, when executed on a computer, causes the computer to perform the methods described in the first aspect or any possible implementation thereof.

[0018] The present invention has the following beneficial effects: This invention provides an online monitoring method for the health status of carbon brushes in motors. It considers the potential abnormal wear of the target carbon brush at the current moment, achieving carbon brush health status monitoring and solving the technical problem of poor timeliness in monitoring abnormal carbon brush health status, thus improving the timeliness of such monitoring. Specifically, this invention analyzes the current ripple time-domain signal, load conditions, operating environment conditions, vibration signals, temperature changes, and current changes of the target carbon brush during the current operating period. It quantifies multiple indicators related to potential carbon brush anomalies, such as abnormal loss factors, potential abnormal wear feature vectors, and wear fault manifestation possibility feature vectors. This allows for relatively timely identification of potential carbon brush anomalies and improves the timeliness of monitoring abnormal carbon brush health status. Attached Figure Description

[0019] To more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 A flowchart illustrating an online monitoring method for the carbon brush health status of an electric motor according to the present invention; Figure 2 This is a schematic diagram of the composition structure of an online monitoring system for the carbon brush health status of a motor according to the present invention. Figure 3 This is a schematic diagram of the structure of a computer device according to the present invention. Detailed Implementation

[0021] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the specific implementation methods, structures, features, and effects of the technical solution proposed according to the present invention are described in detail below with reference to the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0022] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0023] refer to Figure 1 The flowchart illustrates some embodiments of an online monitoring method for the carbon brush health status of a motor according to the present invention. The online monitoring method for the carbon brush health status of the motor includes the following steps: Step S1: Based on the load and operating environment of the target carbon brush during its current operating period, determine the target operating load environment cluster of the target carbon brush at the current moment.

[0024] The target carbon brush can be a brushed motor carbon brush used for online health status monitoring. The current running segment can be the time period during which the target carbon brush is running, and the current time can be the end time of the current running segment. The duration of the current running segment can be 1 hour.

[0025] As an example, this step may include the following steps: The first step is to construct the load environment feature vector for each historical brush under each historical running segment based on the load rate and dimensional data under different preset environmental dimensions for each historical brush under each historical running segment.

[0026] The historical carbon brushes can be those that have operated normally in the past and are of the same specifications and model as the target carbon brush. The historical operating period can be the time during which the historical carbon brush operated normally in the past. The duration of the historical operating period can be equal to the duration of the current operating period. The carbon brush load rate typically refers to the current intensity passing through a unit contact area. The preset environmental dimensions can be pre-set environmental-related dimensions. The number of preset environmental dimensions can be pre-set. For example, preset environmental dimensions can be, but are not limited to, surface temperature, internal temperature, and humidity. The dimension data under the preset environmental dimensions can be normalized values ​​of the data collected under the preset environmental dimensions. For example, the dimension data under the surface temperature dimension can be the normalized value of the surface temperature of the collected carbon brush. The surface temperature can be the carbon brush surface temperature monitored in real time using an infrared thermal imager. The internal temperature can be the internal temperature of the carbon brush obtained through an RFID (Radio Frequency Identification) temperature tag embedded inside the carbon brush. Humidity can be collected through a humidity sensor.

[0027] For example, constructing the load environment feature vector for each historical runtime segment of each historical carbon brush may include the following sub-steps: The first sub-step involves identifying any historical carbon brush as a marked historical carbon brush and identifying any historical running segment of the marked historical carbon brush as a marked historical running segment.

[0028] The second sub-step is to record the time series of the load rates of the marked historical carbon brushes collected at all times within the marked historical running period as the load rate sequence corresponding to the marked historical running period.

[0029] The third sub-step is to record the time series of dimensional data of the marked historical brushes collected at all times during the marked historical runtime period under the same preset environmental dimension as the dimensional data sequence of the marked historical runtime period under the preset environmental dimension.

[0030] The fourth sub-step involves taking the mean, standard deviation, and number of maxima of the load rate sequence corresponding to the marked historical running period, and the mean of the dimensional data sequence of the marked historical running period under the same preset environmental dimension, to form the load environment feature vector of the marked historical carbon brush under the marked historical running period.

[0031] For example, the load environment feature vector of the marked historical brush during the marked historical running period can be represented as: ( , N, , ,..., , ).in, It is the average of all load rates in the load rate sequence corresponding to the historical runtime segment. is the standard deviation of all load rates in the load rate sequence corresponding to the marked historical runtime periods. N is the number of maxima in the load rate sequence corresponding to the marked historical runtime periods. It is the mean of all dimensional data in the dimensional data sequence of the historical runtime segment under the first preset environment dimension. It is the mean of all dimensional data in the dimensional data sequence of the marked historical runtime segment under the second preset environment dimension. It is the mean of all dimensional data in the dimensional data sequence of the (n-1)th preset environment dimension of the marked historical runtime segment. It is the mean of all dimension data in the dimension data sequence of the marked historical runtime segment under the nth preset environment dimension. n is the number of preset environment dimensions.

[0032] The second step is to cluster all historical running periods of all historical carbon brushes based on the load environment feature vectors of all historical running periods of all historical carbon brushes to obtain the initial cluster clusters.

[0033] For example, based on the load environment feature vectors of all historical carbon brushes under all historical running periods, the K-means++ clustering algorithm can be used to cluster all historical running periods of all historical carbon brushes, and the resulting clusters are denoted as the initial clusters. The number of clusters can be determined by the elbow method.

[0034] The third step is to construct the load environment feature vector of the target carbon brush under the current running period based on the load rate and dimensional data of different preset environmental dimensions at all times during the current running period of the target carbon brush, and use it as the current load environment feature vector.

[0035] It should be noted that the method for constructing the load environment feature vector of the target brush under the current running segment can be the same as the method for constructing the load environment feature vector of the marked historical running segment of the marked brush, and will not be repeated here.

[0036] The fourth step is to determine the Euclidean distance between the current load environment feature vector and the cluster center of each initial cluster as the reference distance for each initial cluster.

[0037] The fifth step is to select the initial cluster with the smallest reference distance from all the initial clusters, and use it as the target operating load environment cluster for the target carbon brush at the current moment.

[0038] It should be noted that the more historical carbon brushes there are and the more historical running periods they have, the more reasonable the determination of the target operating load environment cluster tends to be.

[0039] Step S2: Select the historical runtime segments belonging to the target operating load environment cluster from all historical runtime segments of all historical carbon brushes obtained in advance, and use them as reference runtime segments.

[0040] As an example, any historical brush can be identified as a marked historical brush, and the historical runtime segments belonging to the target runtime environment cluster can be selected from all the historical runtime segments of the marked historical brush as the reference runtime segments of the marked historical brush.

[0041] It should be noted that the marked historical carbon brushes refer to the period when the load and operating environment of the marked historical carbon brushes were similar to the current operating period during normal operation.

[0042] Step S3: Based on the pre-obtained total wear of the target carbon brush during the current operating period and the total wear of historical carbon brushes during their reference operating periods, determine the abnormal wear factor of the target carbon brush at the current moment.

[0043] As an example, this step may include the following steps: The first step is to determine the difference between the length of the target carbon brush at the beginning of the current running period and its length at the end of the current running period as the total wear of the target carbon brush during the current running period, and use this as the total wear of the current carbon brush.

[0044] For example, the formula for determining the total wear of the target carbon brush during the current operating period can be: ; Where A is the total wear of the target carbon brush during the current operating period. It is the length of the target carbon brush at the start of the current running segment. It is the length of the target carbon brush at the end of the current running segment.

[0045] The second step, similarly, is to determine the difference between the length of the historical carbon brush at the beginning of its reference operating period and its length at the end of its reference operating period as the total wear amount of the historical carbon brush during its reference operating period, and use it as the reference total wear amount of the carbon brush.

[0046] The third step is to use the Gaussian kernel function to perform density estimation, determine the probability density function of the overall wear of all reference carbon brushes, and take the overall wear of the reference carbon brush corresponding to the maximum function value of the above probability density function as the overall wear of the standard carbon brush.

[0047] It should be noted that the standard carbon brush overall wear can, to some extent, characterize the maximum overall carbon brush wear that would occur during normal operation under the current load environment. It can represent the general situation under the current load environment. Therefore, when the overall carbon brush wear under the current load environment is higher than the standard carbon brush wear, it often indicates that the carbon brush wear may be relatively more abnormal. Optionally, the standard carbon brush overall wear can also be set manually based on experience.

[0048] For example, the formula for determining the total wear of a historical carbon brush during its reference operating period can be: ; in, This represents the total wear of the a-th historical carbon brush during its b-th reference operating period. 'a' is the serial number of the historical carbon brush, and 'b' is the serial number of the reference operating period for the a-th historical carbon brush. It is the length of the a-th historical brush at the start of its b-th reference running period. It is the length of the a-th historical brush at the end of its b-th reference running period.

[0049] The fourth step is to normalize the difference between the current overall wear of the carbon brush and the overall wear of the standard carbon brush to obtain the initial anomaly factor of the target carbon brush at the current moment.

[0050] It should be noted that when the current overall wear of the carbon brush is higher than that of the standard carbon brush, it often indicates that the wear of the target carbon brush may be more abnormal at the current moment.

[0051] The fifth step is to determine the sum of constant 1 and the above-mentioned initial anomaly factor as the anomaly loss factor of the target carbon brush at the above-mentioned current moment.

[0052] Step S4: Based on the pre-acquired current ripple time-domain signal of the target carbon brush during the current operating period, as well as its clamping force and abnormal wear factor at the current moment, determine the potential abnormal wear feature vector of the target carbon brush at the current moment.

[0053] As an example, this step may include the following steps: The first step is to normalize the absolute value of the difference between the clamping force of the target carbon brush at the current moment and the pre-acquired standard clamping force to obtain the force deviation factor.

[0054] One method is to use a pressure sensor to measure the clamping force of the spring on the carbon brush. The standard clamping force characterizes the clamping force of the spring on the carbon brush during normal operation, and it can be the ideal clamping force set at the factory.

[0055] It should be noted that when the force deviation factor is larger, it often means that the friction force generated by the spring on the carbon brush deviates more from the ideal friction force, which often means that the subtle changes should be amplified.

[0056] The second step is to determine the sum of constant 1 and the above-mentioned force deviation factor as the target force deviation index.

[0057] The third step is to determine the abnormal amplification factor by multiplying the target force deviation index and the abnormal loss factor mentioned above.

[0058] The fourth step is to use a coupled current probe to obtain the current ripple time-domain signal of the target carbon brush during the current operating period.

[0059] Among them, the current ripple time-domain signal is also known as the current ripple time-domain signal.

[0060] The fifth step is to perform wavelet packet decomposition on the current ripple time-domain signal to decompose the signal into 8 characteristic frequency bands and determine the energy proportion of each characteristic frequency band.

[0061] Among them, the characteristic frequency bands can be ordered, and the earlier the characteristic frequency band is, the higher the frequency tends to be.

[0062] It should be noted that wavelet packet decomposition can be performed on the current ripple time-domain signal to decompose the signal into 8 characteristic frequency bands (0-10kHz). Each characteristic frequency band has a frequency range width of 1.25kHz, and the earlier the number, the higher the frequency. The energy proportion of each frequency band (E1-E8) can be calculated.

[0063] The sixth step is to determine the energy representative value corresponding to each characteristic frequency band by summing the preset positive factor and the energy proportion of each characteristic frequency band.

[0064] It should be noted that the preset positive factor can be a pre-set positive factor, which can be 1. This is mainly to avoid the denominator being 0 when using the energy representative value corresponding to the characteristic frequency band as the denominator.

[0065] Step 7, determining the target performance value of micro-discharge intensity based on the above-mentioned abnormal amplification factor and the energy representative values ​​corresponding to the first three characteristic frequency bands may include the following sub-steps: The first sub-step is to determine the sum of the energy representative values ​​corresponding to the first characteristic frequency band and the energy representative values ​​corresponding to the second characteristic frequency band as the normal contact friction current ripple performance value.

[0066] The second sub-step is to determine the initial performance value of the micro-discharge intensity by the ratio of the energy representative value corresponding to the third characteristic frequency band to the above-mentioned normal contact friction current ripple performance value.

[0067] The third sub-step involves determining the target performance value of the micro-discharge intensity by multiplying the aforementioned abnormal amplification factor by the aforementioned initial performance value of the micro-discharge intensity.

[0068] For example, the formula for determining the target performance value of micro-discharge intensity can be: ; in, It is the target performance value of micro-discharge intensity. It is the energy representative value corresponding to the first characteristic frequency band. It is the energy representative value corresponding to the second characteristic frequency band. This is the energy representative value corresponding to the third characteristic frequency band. C is the anomalous amplification factor. This is the normal contact friction current ripple performance value. It is the initial performance value of the micro-discharge intensity.

[0069] It should be noted that the first, second, and third characteristic frequencies are often in the high-frequency range. Among them, the third characteristic frequency is often closest to the micro-discharge signal caused by poor contact between the carbon brush and the commutator. The first and second characteristic frequencies may be the triboelectric current ripple of normal contact. A higher value often indicates a more intense micro-discharge. When C is larger, it often means that the frictional force generated by the spring on the carbon brush deviates more from the ideal frictional force, and the wear of the target carbon brush at the current moment may be more abnormal, which often indicates that subtle changes should be amplified at this time.

[0070] Step 8: Based on the sum of the energy representative values ​​corresponding to the 4th and 5th characteristic frequency bands, and the aforementioned anomalous amplification factor, determine the abnormal index of carbon brush elastic deformation recovery.

[0071] For example, the formula for determining the abnormal index of carbon brush elastic deformation recovery can be: ; in, It is an abnormal indicator of carbon brush elastic deformation recovery. It is an exponential function with the natural constant as its base. It is the energy representative value corresponding to the 4th characteristic frequency band. This is the energy representative value corresponding to the 5th characteristic frequency band. C is the anomalous amplification factor.

[0072] It should be noted that the fourth and fifth characteristic frequencies are often in the mid-frequency range, which together reflect the vibration frequency of the carbon brush caused by the elastic deformation under the pressure of the spring. The decrease in the proportion of the mid-frequency component often indicates that the deformation recovery ability is more likely to decrease due to spring fatigue.

[0073] The ninth step is to determine the material carbonization anomaly index based on the ratio between the energy representative value corresponding to the 6th characteristic frequency band and the energy representative value corresponding to the 7th characteristic frequency band, as well as the above-mentioned anomalous amplification factor.

[0074] For example, the formula for determining the anomaly index of material carbonization can be: ; in, It is an indicator of abnormal carbonization in materials. It is an exponential function with the natural constant as its base. It is the energy representative value corresponding to the 6th characteristic frequency band. This is the energy representative value corresponding to the 7th characteristic frequency band. C is the anomalous amplification factor.

[0075] It should be noted that the 6th and 7th characteristic frequencies are often low-frequency components. Carbon brushes typically contain both carbonized and uncarbonized portions. The carbonization process is often a material degradation process; therefore, the friction frequency of the uncarbonized material is relatively higher. It can be used to characterize the frictional features of the uncarbonized portion of carbon brush materials. It can then be used to address the friction characteristics of the corresponding carbide layer. The smaller the size, the more severe the carbonization.

[0076] Step 10: Combine the above-mentioned micro-discharge intensity target performance value, the above-mentioned carbon brush elastic deformation recovery abnormal index, and the material carbonization abnormal index to form the potential abnormal wear feature vector of the target carbon brush at the current moment.

[0077] It should be noted that the potential abnormal wear feature vector can often dynamically capture subtle changes in the contact state between the carbon brush and the commutator. Even when there are no significant fluctuations in electrical parameters, potential wear risks can often be identified through the abnormal trends of each component.

[0078] Step S5: Based on the Lyapunov exponent of the vibration signal of the target carbon brush during the current operating period, and the temperature and current changes of the target carbon brush during the current operating period, determine the possible feature vector of wear failure manifestation of the target carbon brush at the current moment.

[0079] As an example, this step may include the following steps: The first step is to acquire the vibration signal of the target carbon brush during the current operating period, and to acquire the surface temperature, internal temperature and current value of the target carbon brush at each moment during the current operating period.

[0080] The second step is to determine the difference between the surface temperature and the internal temperature of the target carbon brush at each moment during the current operating period as the initial temperature difference factor of the target carbon brush at each moment during the current operating period.

[0081] It should be noted that the greater the difference between the surface temperature and the internal temperature of the target carbon brush, the more severe the heat conduction problems caused by wear inside the carbon brush.

[0082] The third step is to determine the difference between the initial temperature difference factors of the target carbon brush at each adjacent time point within the current operating period as the temperature difference change between each adjacent time point, thus obtaining the temperature difference change sequence.

[0083] For example, the formula for determining the temperature difference between adjacent time points can be: ; in, It represents the temperature difference change of the target carbon brush between the i-th time and the (i+1)-th time within the current operating segment. i is the sequence number of the different times within the current operating segment. It is the initial temperature difference factor of the target carbon brush at the (i+1)th time in the current running segment. It is the initial temperature difference factor of the target carbon brush at the i-th time in the current running segment.

[0084] The fourth step is to determine the absolute value of the difference between the current values ​​of the target carbon brush at each adjacent time point within the current operating period as the current change between adjacent time points, thus obtaining the current change sequence.

[0085] The fifth step is to determine the target change index between each adjacent moment based on the temperature difference and current change of the target carbon brush during the current operating period, and obtain the target change index sequence.

[0086] For example, the formula for determining the target change index between adjacent time points can be: ; in, It is the target change index of the target carbon brush between the i-th time and the (i+1)-th time within the current running segment. i is the sequence number of different times within the current running segment. It is the change in temperature difference between the target carbon brush at time i and time i+1 within the current operating period. It is the change in current of the target carbon brush between the i-th time and the (i+1)-th time during the current operating period. It is a pre-set factor greater than 0, mainly used to prevent the denominator from being 0, and it can be 0.0001.

[0087] It should be noted that when A larger value usually indicates a greater temperature difference caused by a change in unit current.

[0088] Step 6, determining the temperature rise-current coupling effect value based on the above-mentioned temperature difference change sequence, current change sequence, and target change index sequence may include the following sub-steps: The first sub-step involves performing linear fitting on the above target change index sequence to obtain the target change line.

[0089] In this context, the horizontal axis of the target change line can be time, and the vertical axis can be the target change index in the target change index sequence.

[0090] The second sub-step involves determining the temperature rise-current coupling effect value based on the mean of the temperature difference change sequence, the mean of the current change sequence, and the slope of the target change line.

[0091] For example, the formula for determining the value of the temperature rise current coupling effect can be: ; Where Q is the value of the temperature rise current coupling effect. It is the normalization function. k is the slope of the target changing line. It is the mean of all temperature difference changes in the temperature difference change sequence. It is the mean of all current changes in the current change sequence. It is a pre-set factor greater than 0, mainly used to prevent the denominator from being 0, and it can be 0.0001.

[0092] It should be noted that a larger k value often indicates a greater likelihood of temperature difference caused by a change in unit current, and a greater likelihood of worsening heat conduction issues due to wear within the carbon brush. A larger Q value usually indicates a greater temperature difference caused by changes in current during the current operating period. Therefore, a larger Q value often indicates a stronger trend in the temperature rise-current coupling effect.

[0093] The seventh step is to combine the above temperature rise current coupling effect value and the maximum Lyapunov exponent of the vibration signal of the target carbon brush during the current operating period to form a possible feature vector of wear failure manifestation of the target carbon brush at the current moment.

[0094] It should be noted that a larger maximum Lyapunov exponent in the vibration signal of the target carbon brush during the current operating period often indicates a greater likelihood of enhanced chaotic characteristics in the friction system. This suggests a more unstable frictional state between the carbon brush and the commutator. The shift from stable sliding to irregular collisions in the friction between the carbon brush and the commutator is a sign of deteriorating mechanical conditions, indicating that the carbon brush is approaching the failure threshold. Therefore, a larger element in the wear failure manifestation characteristic vector often indicates a greater likelihood of entering a positive feedback stage of accelerated wear leading to deterioration, significantly increasing the risk of failure.

[0095] Step S6: Determine the health status of the target carbon brush at the current moment based on the potential abnormal wear feature vector and the wear fault manifestation possible feature vector.

[0096] As an example, this step may include the following steps: The first step is to normalize the cumulative product of all elements in the aforementioned potential abnormal wear feature vector and the aforementioned wear fault manifestation possible feature vector to obtain the target abnormal state index of the target carbon brush at the current moment.

[0097] The second step is to determine whether the target carbon brush has a potential or actual abnormality at the current moment if the above-mentioned target abnormality indicators are greater than the preset abnormality threshold.

[0098] The preset abnormal threshold can be a pre-set threshold, which can be 0.6.

[0099] The third step is to determine that the target carbon brush is in good health at the current moment if the above-mentioned target abnormal state indicators are less than or equal to the preset abnormal threshold.

[0100] Optionally, determining the health status of the target carbon brush at the current moment based on the potential abnormal wear feature vector and the wear failure manifestation potential feature vector may further include the following steps: For potential abnormal wear feature vectors and wear fault manifestation potential feature vectors, the former describes the subtle changes in the contact state between the carbon brush and the commutator under certain operating conditions and environmental conditions when electrical parameters do not fluctuate significantly; while the latter describes the changes in the brush from wear to gradual evolution into fault manifestation.

[0101] Therefore, these two feature vectors can be used as key extracted features for the real-time operation of the brush under the corresponding working conditions and environments. By using them to train the monitoring model, the health status of the carbon brush at any time can be quantified and evaluated to help maintenance personnel judge the true health status of the carbon brush.

[0102] It can collect historical data under different working conditions and environments, extract potential abnormal wear feature vectors and wear fault manifestation feature vectors for each sample; label the samples according to their actual health status (0 for normal, 1 for slight abnormality, and 2 for severe abnormality), and use these two feature vectors together with conventional explicit features such as current, voltage, and temperature as feature inputs to train a random forest classifier, obtain the basic model corresponding to each combination, and output the probability of the sample belonging to each category (P0, P1, P2).

[0103] After real-time monitoring data is processed synchronously, potential abnormal wear feature vectors and wear fault manifestation feature vectors at the current moment are obtained. The corresponding basic model is called by matching the current working condition-environment label to obtain real-time probabilities P0, P1, and P2.

[0104] The standard carbon brush wear amount is introduced for dynamic correction: when the current carbon brush wear amount exceeds 1.2 times the standard carbon brush wear amount, P1 and P2 are multiplied by weights of 1.1 and 1.3 respectively; otherwise, they are multiplied by weights of 0.9 and 0.7 respectively.

[0105] Ultimately, the health status assessment results corresponding to the real-time monitoring data of the carbon brushes of the brushed motor are obtained, providing a quantitative reference for maintenance personnel.

[0106] Because this invention extracts the subtle changes in the contact state between the carbon brush and the commutator when the electrical parameters do not fluctuate significantly, as well as the changes in the brush as it wears down and gradually evolves into a fault, the full-cycle health status of the carbon brush of a brushed motor can often be accurately captured, and a synchronously quantified evaluation result can be obtained. Based on this evaluation result, maintenance personnel can often accurately judge the carbonization situation to determine whether to replace the node.

[0107] refer to Figure 2Based on the same inventive concept as the above-described method embodiments, this invention provides an online monitoring system for the carbon brush health status of a motor. The system includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the steps of an online monitoring method for the carbon brush health status of a motor, specifically including: The target operating load environment cluster acquisition module 201 is used to determine the target operating load environment cluster of the target carbon brush at the current moment based on the load and operating environment of the target carbon brush in the current running segment, which are obtained in advance. The runtime segment filtering module 202 is used to filter out the historical runtime segments belonging to the target runtime environment cluster from all the historical runtime segments of all historical carbon brushes obtained in advance, and use them as reference runtime segments. The abnormal wear factor determination module 203 is used to determine the abnormal wear factor of the target carbon brush at the current moment based on the total wear amount of the target carbon brush in the current running segment and the total wear amount of the historical carbon brush in its reference running segment. The potential abnormal wear feature vector acquisition module 204 is used to determine the potential abnormal wear feature vector of the target carbon brush at the current moment based on the pre-acquired current ripple time domain signal of the target carbon brush in the current operating period and its clamping force and abnormal wear factor at the current moment. The wear fault manifestation possible feature vector acquisition module 205 is used to determine the wear fault manifestation possible feature vector of the target carbon brush at the current moment based on the Lyapunov exponent of the vibration signal of the target carbon brush in the current running period, as well as the temperature change and current change of the target carbon brush in the current running period. The health status monitoring module 206 is used to determine the health status of the target carbon brush at the current moment based on the potential abnormal wear feature vector and the wear fault manifestation possible feature vector.

[0108] Figure 3 This is a schematic diagram of the structure of a computer device provided in an embodiment of the present invention. For example, as shown... Figure 3 As shown, the computer device 300 includes: a memory 301, a processor 302, and a computer program 303 stored in the memory 301 and running on the processor 302. When the processor 302 executes the computer program 303, the computer device can execute any of the aforementioned online monitoring methods for the carbon brush health status of motors.

[0109] Based on the same inventive concept as the above-described method embodiments, the present invention provides a server, including a memory and a processor. The memory stores executable program code, and the processor retrieves and runs the executable program code from the memory, causing the device to execute any of the above-described online monitoring methods for the carbon brush health status of motors.

[0110] Based on the same inventive concept as the above-described method embodiments, the present invention provides a computer program product comprising: computer program code, which, when executed on a computer, causes the computer to perform any of the above-described methods for online monitoring of the carbon brush health status of a motor.

[0111] Based on the same inventive concept as the above-described method embodiments, the present invention provides a computer-readable storage medium storing computer program code, which, when executed on a computer, causes the computer to perform any of the above-described methods for online monitoring of the carbon brush health status of a motor.

[0112] In summary, this invention quantifies multiple indicators related to potential carbon brush anomalies by analyzing the current ripple time-domain signal, load conditions, operating environment conditions, vibration signals, temperature changes, and current changes of the target carbon brush during the current operating period. These indicators include abnormal wear factors, potential abnormal wear feature vectors, and wear fault manifestation feature vectors. This allows for the relatively timely identification of potential carbon brush anomalies and improves the timeliness of monitoring abnormal carbon brush health status.

[0113] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A method for online monitoring of the carbon brush health status of an electric motor, characterized in that, Includes the following steps: Based on the pre-obtained load and operating environment of the target carbon brush during its current running period, the target operating load environment cluster of the target carbon brush at the current moment is determined, where the current moment is the end moment of the current running period; Select the historical runtime segments belonging to the target operating load environment cluster from all historical runtime segments of all historical carbon brushes obtained in advance, and use them as reference runtime segments; Based on the total wear of the target carbon brush in the current operating period and the total wear of historical carbon brushes in their reference operating periods, the abnormal wear factor of the target carbon brush at the current moment is determined. Based on the pre-acquired current ripple time-domain signal of the target carbon brush during the current operating period, as well as its clamping force and abnormal wear factor at the current moment, the potential abnormal wear feature vector of the target carbon brush at the current moment is determined. Based on the Lyapunov exponent of the vibration signal of the target carbon brush during the current operating period, and the temperature and current changes of the target carbon brush during the current operating period, the possible feature vector of wear failure manifestation of the target carbon brush at the current moment is determined. Based on the potential abnormal wear feature vector and the wear failure manifestation feature vector, the health status of the target carbon brush at the current moment is determined.

2. The method for online monitoring of carbon brush health status of an electric motor according to claim 1, characterized in that, The step of determining the target operating load environment cluster of the target carbon brush at the current moment based on the pre-acquired load and operating environment conditions during the current operating period includes: Based on the load rate and dimensional data of different preset environmental dimensions at all times in each historical runtime segment of each historical carbon brush, a load environment feature vector for each historical carbon brush under each historical runtime segment is constructed. Based on the load environment feature vectors of all historical carbon brushes under all historical running periods, clusters are formed for all historical running periods of all historical carbon brushes to obtain initial clusters. Based on the load rate and dimensional data of the target carbon brush at all times during the current running period and under different preset environmental dimensions, a load environment feature vector of the target carbon brush under the current running period is constructed as the current load environment feature vector. The Euclidean distance between the current load environment feature vector and the cluster center of each initial cluster is determined as the reference distance for each initial cluster. The initial cluster with the smallest reference distance is selected from all initial clusters and used as the target operating load environment cluster for the target carbon brush at the current moment.

3. The method for online monitoring of carbon brush health status of an electric motor according to claim 2, characterized in that, The step involves constructing a load environment feature vector for each historical brush during each historical runtime period based on the pre-acquired load rate and dimensional data under different preset environmental dimensions for each historical brush, including: Any historical carbon brush is identified as a marked historical carbon brush, and any historical running segment of the marked historical carbon brush is identified as a marked historical running segment. The time series of load rates of the marked historical carbon brushes collected at all times within the marked historical running period is denoted as the load rate sequence corresponding to the marked historical running period. The time series of dimensional data of the marked historical carbon brushes collected at all times within the marked historical runtime period under the same preset environmental dimension is denoted as the dimensional data sequence of the marked historical runtime period under the preset environmental dimension. The mean, standard deviation, and number of maxima of the load rate sequence corresponding to the marked historical running segment, as well as the mean of the dimensional data sequence of the marked historical running segment under the same preset environmental dimension, constitute the load environment feature vector of the marked historical brush under the marked historical running segment.

4. The method for online monitoring of carbon brush health status of an electric motor according to claim 1, characterized in that, The step of determining the abnormal wear factor of the target carbon brush at the current moment based on the pre-obtained total wear of the target carbon brush during the current operating period and the total wear of historical carbon brushes during their reference operating periods includes: The difference between the length of the target carbon brush at the start time and its length at the end time of the current running segment is determined as the total wear of the target carbon brush in the current running segment, and is used as the current total wear of the carbon brush. Similarly, the difference between the length of the historical carbon brush at the beginning of its reference operating period and its length at the end of its reference operating period is determined as the total wear of the historical carbon brush during its reference operating period, and is used as the reference total wear of the carbon brush. Determine the probability density function of the overall wear of all reference carbon brushes, and take the overall wear of the reference carbon brush corresponding to the maximum function value of the probability density function as the overall wear of the standard carbon brush. The difference between the current total wear of the carbon brush and the total wear of the standard carbon brush is normalized to obtain the initial anomaly factor of the target carbon brush at the current moment; The sum of constant 1 and the initial anomaly factor is determined as the anomaly loss factor of the target carbon brush at the current moment.

5. The method for online monitoring of carbon brush health status of an electric motor according to claim 1, characterized in that, The step of determining the potential abnormal wear feature vector of the target carbon brush at the current moment based on the pre-acquired current ripple time-domain signal of the target carbon brush during the current operating period, as well as its clamping force and abnormal wear factor at the current moment, includes: The absolute value of the difference between the clamping force of the target carbon brush at the current moment and the pre-acquired standard clamping force is normalized to obtain the force deviation factor; The sum of constant 1 and the force deviation factor is determined as the target force deviation index; The product of the target force deviation index and the abnormal loss factor is determined as the abnormal amplification factor; Obtain the current ripple time-domain signal of the target carbon brush during the current operating period; The current ripple time-domain signal is decomposed into wavelet packet decomposition, which divides the signal into 8 characteristic frequency bands. The energy proportion of each characteristic frequency band is determined. The characteristic frequency bands are ordered, and the earlier the characteristic frequency band is, the higher the frequency. The sum of the preset positive factor and the energy proportion of each characteristic frequency band is used to determine the energy representative value corresponding to each characteristic frequency band; The target performance value of micro-discharge intensity is determined based on the abnormal amplification factor and the energy representative values ​​corresponding to the first three characteristic frequency bands. The abnormal index of carbon brush elastic deformation recovery is determined based on the sum of the energy representative values ​​corresponding to the fourth and fifth characteristic frequency bands, and the abnormal amplification factor. The material carbonization anomaly index is determined based on the ratio between the energy representative value corresponding to the 6th characteristic frequency band and the energy representative value corresponding to the 7th characteristic frequency band, as well as the anomaly amplification factor. The target performance value of micro-discharge intensity, the abnormal index of carbon brush elastic deformation recovery, and the abnormal index of material carbonization are used to construct the potential abnormal wear feature vector of the target carbon brush at the current moment.

6. The method for online monitoring of carbon brush health status of an electric motor according to claim 5, characterized in that, The step of determining the target performance value of micro-discharge intensity based on the abnormal amplification factor and the energy representative values ​​corresponding to the first three characteristic frequency bands includes: The sum of the energy representative value corresponding to the first characteristic frequency band and the energy representative value corresponding to the second characteristic frequency band is determined as the normal contact friction current ripple performance value. The ratio of the energy representative value corresponding to the third characteristic frequency band to the normal contact friction current ripple performance value is determined as the initial performance value of the micro-discharge intensity. The product of the abnormal amplification factor and the initial performance value of the micro-discharge intensity is determined as the target performance value of the micro-discharge intensity.

7. The method for online monitoring of carbon brush health status of an electric motor according to claim 1, characterized in that, The step of determining the possible wear fault manifestation feature vector of the target carbon brush at the current moment based on the pre-acquired Lyapunov exponent of the vibration signal of the target carbon brush during the current operating period, and the pre-acquired temperature and current changes of the target carbon brush during the current operating period, includes: The vibration signal of the target carbon brush during the current operating period is acquired, and the surface temperature, internal temperature and current value of the target carbon brush at each moment during the current operating period are acquired. The difference between the surface temperature and the internal temperature of the target carbon brush at each moment during the current operating period is determined as the initial temperature difference factor of the target carbon brush at each moment during the current operating period. The difference between the initial temperature difference factors of the target carbon brush at each adjacent time point in the current operating period is determined as the temperature difference change between each adjacent time point, thus obtaining a temperature difference change sequence. The absolute value of the difference between the current values ​​of the target carbon brush at each adjacent time point within the current operating period is determined as the current change between each adjacent time point, thus obtaining a current change sequence. Based on the temperature difference change and current change between adjacent moments in the current operating period of the target carbon brush, the target change index between adjacent moments is determined, and the target change index sequence is obtained. The temperature rise-current coupling effect value is determined based on the temperature difference change sequence, the current change sequence, and the target change index sequence. The temperature rise current coupling effect value and the maximum Lyapunov exponent of the vibration signal of the target carbon brush during the current operating period are used to construct the wear fault manifestation feature vector of the target carbon brush at the current moment.

8. The method for online monitoring of carbon brush health status of an electric motor according to claim 7, characterized in that, The step of determining the temperature rise-current coupling effect value based on the temperature difference change sequence, the current change sequence, and the target change index sequence includes: By performing linear fitting on the target change index sequence, a target change line is obtained; The temperature rise-current coupling effect value is determined based on the mean of the temperature difference change sequence, the mean of the current change sequence, and the slope of the target change line.

9. The method for online monitoring of carbon brush health status of an electric motor according to claim 1, characterized in that, The process of determining the health status of the target carbon brush at the current moment based on the potential abnormal wear feature vector and the wear fault manifestation possible feature vector includes: Normalize the cumulative product of all elements in the potential abnormal wear feature vector and the wear fault manifestation possible feature vector to obtain the target abnormal state index of the target carbon brush at the current moment. If the target abnormal state index is greater than the preset abnormal threshold, it is determined that the target carbon brush has a potential abnormality or a real abnormality at the current moment. If the target abnormal state index is less than or equal to the preset abnormal threshold, the target carbon brush is determined to be in good health at the current moment.

10. An online monitoring system for the carbon brush health status of an electric motor, characterized in that, It includes a processor and a memory, the processor being used to process instructions stored in the memory to implement an online monitoring method for the carbon brush health status of an electric motor according to any one of claims 1-9.

Images