Motor state monitoring method and system based on data drive

By segmenting the steady-state working conditions of the motor and training the abnormality detection model, combining the working condition matching and judgment model, the problem of difficult to determine the abnormal state of the motor in the existing technology is solved, and efficient and accurate motor status monitoring and management is achieved .

CN113030726BActive Publication Date: 2024-11-15CHINA TOBACCO HENAN IND CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202110330952.X
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2021-03-24
Publication Date
2024-11-15
Estimated Expiration
2041-03-24

AI Technical Summary

Technical Problem

The prior art is difficult to effectively determine whether the motor is in an abnormal state, and it is impossible to predict the future development trend of the equipment or systematically manage the motor health status.

Method used

By segmenting the steady-state working conditions of the motor, the working condition abnormality detection model is trained, and matching the current working conditions with the motor based on the steady-state working conditions list file, the motor status monitoring results are automatically obtained using the working condition abnormality judgment model.

Benefits of technology

It improves the efficiency and accuracy of motor abnormality monitoring, can effectively determine whether the motor is in an abnormal state, and provides the system's motor health status management.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN113030726B_ABST
    Figure CN113030726B_ABST
Patent Text Reader

Abstract

The present invention discloses a motor state monitoring method and system based on data drive, the method comprising: segmenting the motor working condition after stable operation to obtain a steady-state working condition list; training a working condition abnormality detection model corresponding to the steady-state working condition list according to the segmentation result; matching the current working condition of the motor with each combined working condition in the steady-state working condition list file; obtaining the motor state monitoring result according to the matching result and the corresponding working condition abnormality detection model. The present invention is conducive to obtaining effective motor operating conditions by segmenting the motor working condition, and training a working condition abnormality detection model based on data analysis, and analyzing the current motor working condition based on the working condition matched with the current working condition of the motor in the steady-state working condition list file and the working condition abnormality judgment model corresponding to the matched working condition, and automatically and directly obtaining the motor state monitoring result, which solves the disadvantage of the prior art that it is difficult to effectively determine whether the motor is in an abnormal state.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of motor diagnosis, and in particular to a data-driven motor state monitoring method and system. Background Art

[0002] Motor status monitoring and fault diagnosis technology is a technology that understands and grasps the status of the motor during use, determines whether it is normal or abnormal as a whole or in part, and finds faults and their causes. Existing technical solutions are generally status monitoring and diagnostic devices based on electrical characteristic analysis technology. By measuring the current and voltage signals when the motor is running under load, analyzing the spectrum, harmonics, electrical parameters and other characteristics, it further detects the status changes caused by bearing faults, misalignment faults, load faults, mechanical looseness, insulation and a series of electrical and mechanical faults, and then determines the fault of the entire transmission system.

[0003] However, traditional technical solutions focus on the current status of the equipment and cannot predict the future development trend of the equipment, let alone systematically manage the health status of the equipment. How to effectively determine whether the motor is in an abnormal state is an important problem to be solved in monitoring the motor status and diagnosing motor faults based on motor data.

[0004] Therefore, there is an urgent need for a data-driven motor state monitoring method and system. Summary of the invention

[0005] The purpose of the present invention is to provide a data-driven motor state monitoring method and system to solve the problems in the above-mentioned prior art. It can train an abnormal operating condition detection model based on data analysis, and analyze the current motor operating condition based on the operating condition matching the current motor operating condition in the steady-state operating condition list file and the operating condition abnormality judgment model corresponding to the matching operating condition, so as to automatically and directly obtain the motor state monitoring result.

[0006] The present invention provides a data-driven motor state monitoring method, which includes:

[0007] The motor operating conditions after the motor runs stably are divided to obtain a steady-state operating condition list, wherein the steady-state operating condition list includes several combined operating conditions of different stable speed operating conditions and different current operating conditions;

[0008] According to the segmentation result of the motor working condition, a number of working condition abnormality detection models corresponding to each combined working condition in the steady-state working condition list are trained, wherein the number of the working condition abnormality detection models corresponds to the number of the combined working conditions;

[0009] Matching the current working condition of the motor with each combined working condition in the steady-state working condition list file;

[0010] The motor state monitoring result is obtained according to the matching result of the current working condition of the motor and the working condition abnormality detection model corresponding to the matching result.

[0011] In the data-driven motor state monitoring method as described above, preferably, the motor operating conditions after the motor runs stably are segmented to obtain a steady-state operating condition list, specifically including:

[0012] The data in the working condition file is preliminarily divided according to the stable speed vector to obtain several preliminary groups;

[0013] marking the current data in each of the preliminary groups with a steady-state current label;

[0014] Each preliminary grouping is divided twice according to the steady-state current label to obtain a final division result.

[0015] In the data-driven motor state monitoring method as described above, preferably, the data in the working condition file is preliminarily divided according to the stable speed vector to obtain several preliminary groups, specifically including:

[0016] According to the stable speed vector, the data of different speeds in the working condition file are divided into several preliminary groups, and the data of the same speed is taken as the same preliminary group, wherein the number of the preliminary groups corresponds to the number of the stable speed vectors, and in the same preliminary group, the same speed working condition corresponds to several different current working conditions,

[0017] The secondary segmentation of each preliminary grouping according to the steady-state current label to obtain the final segmentation result specifically includes:

[0018] In each of the preliminary groups, the data marked with the same steady-state current label are regarded as the same fine group;

[0019] According to each of the detailed groups in each of the preliminary groups, a steady-state operating condition list is obtained, wherein the steady-state operating condition list includes combined operating conditions of different speeds and different currents;

[0020] The steady-state operating condition list is output as a CSV file including speed variables and current variables, wherein the CSV file includes operating conditions corresponding to each speed, and each speed corresponds to several current operating conditions.

[0021] In the data-driven motor state monitoring method as described above, preferably, the segmentation result of the motor working condition is used to train a number of working condition anomaly detection models corresponding to each combined working condition in the steady-state working condition list, specifically including:

[0022] The working condition sequence number and the 16-dimensional feature matrix file corresponding to each combined working condition in the steady-state working condition list are used as training data and input into the preset machine learning model, wherein each row of the 16-dimensional feature matrix represents a sample, the first column is time information, the second column represents speed, the third column represents steady state, the fourth column represents temperature, and the remaining columns represent 12 time domain and frequency domain features;

[0023] The method of isolating abnormal points is adopted, and the preset machine learning model is trained using the training data to obtain an operating condition anomaly detection model corresponding to each combined operating condition in the steady-state operating condition list.

[0024] In the data-driven motor state monitoring method as described above, preferably, matching the current working condition of the motor with each combined working condition in the steady-state working condition list file specifically includes:

[0025] Matching the speed and current of the motor in the current working condition with the stable speed working condition and current working condition in the steady-state working condition list file respectively;

[0026] The matching result of the current working condition of the motor is obtained according to the matching result of the rotational speed and the current, so as to determine the matching working condition that matches the current working condition of the motor from among the combined working conditions in the steady-state working condition list.

[0027] The data-driven motor state monitoring method as described above, wherein preferably, the matching result of the current working condition of the motor is obtained according to the matching result of the speed and the current, so as to determine the matching working condition matching the current working condition of the motor in each combined working condition in the steady-state working condition list, specifically includes:

[0028] When the speed and current in the current working condition of the motor are successfully matched, the working condition sequence number corresponding to the combined working condition matching the current working condition of the motor in the steady-state working condition list is output as the matching result of the current working condition of the motor, and a 13-dimensional feature parameter list is output at the same time, in which the first column is temperature and the next 12 columns are time domain and frequency domain features.

[0029] The data-driven motor state monitoring method as described above, wherein preferably, the motor state monitoring result is obtained based on the matching result of the current working condition of the motor and the working condition abnormality detection model corresponding to the matching result, specifically including:

[0030] Determine, from among the several operating condition abnormality judgment models corresponding to the several combined operating conditions in the steady-state operating condition list, the operating condition abnormality judgment model corresponding to the matching operating condition as the matching operating condition abnormality judgment model;

[0031] A motor state monitoring result is obtained according to the matching working condition and the matching working condition abnormality judgment model.

[0032] In the data-driven motor state monitoring method as described above, preferably, the abnormal condition judgment model corresponding to the matching working condition is determined from among the several abnormal condition judgment models corresponding to the several combined working conditions in the steady-state working condition list as the matching abnormal condition judgment model, specifically including:

[0033] Establishing a correspondence between the model serial number of each of the abnormal working condition judgment models and the working condition serial number of each of the combined working conditions in the steady-state working condition list;

[0034] According to the operating condition serial number corresponding to the combined operating condition matching the current operating condition of the motor in the steady-state operating condition list, and the correspondence between the model serial number and the operating condition serial number, the operating condition abnormality judgment model corresponding to the matching operating condition is determined as the matching operating condition abnormality judgment model.

[0035] The data-driven motor state monitoring method as described above, wherein preferably, obtaining the motor state monitoring result according to the matching working condition and the matching working condition abnormality judgment model specifically includes:

[0036] According to the relationship between the temperature in the 13-dimensional feature parameter list and the preset temperature threshold, determine whether it is necessary to use a matching operating condition abnormality judgment model to obtain a motor state monitoring result;

[0037] When the temperature is less than or equal to the preset temperature threshold, it is determined that the motor state is normal;

[0038] When the temperature is greater than a preset temperature threshold, the motor state monitoring result is obtained by using a matching working condition abnormality judgment model, and the working condition sequence number corresponding to the combined working condition matching the current working condition of the motor in the steady-state working condition list and the 13-dimensional feature parameter list are input into the matching working condition abnormality judgment model to obtain the cumulative number of abnormalities output by the matching working condition abnormality judgment model;

[0039] If the accumulated number of abnormalities meets the preset abnormal condition, it is determined that the motor state is abnormal; if the accumulated number of abnormalities does not meet the preset abnormal condition, it is determined that the motor state is normal.

[0040] The present invention also provides a data-driven motor state monitoring system using the above method, comprising:

[0041] A working condition segmentation module, used for segmenting the motor working condition after the motor runs stably to obtain a steady-state working condition list, wherein the steady-state working condition list includes several combined working conditions of different stable speed working conditions and different current working conditions;

[0042] A model training module, used for training a plurality of abnormal working condition detection models corresponding to each combined working condition in the steady-state working condition list according to the segmentation result of the motor working condition, wherein the number of the abnormal working condition detection models corresponds to the number of the combined working conditions;

[0043] A working condition matching module, used for matching the current working condition of the motor with each combined working condition in the steady-state working condition list file;

[0044] The abnormality detection module is used to obtain the motor state monitoring result according to the matching result of the current working condition of the motor and the working condition abnormality detection model corresponding to the matching result.

[0045] The present invention provides a data-driven motor state monitoring method and system, which is beneficial to obtaining effective motor operating conditions by segmenting the motor working conditions, and training a working condition abnormality detection model based on data analysis, and analyzing the current motor working conditions based on the working conditions matching the current motor working conditions in a steady-state working condition list file and a working condition abnormality judgment model corresponding to the matching working conditions, and automatically and directly obtaining the motor state monitoring result, thereby solving the problem in the prior art that it is difficult to effectively determine whether the motor is in an abnormal state, and improving the efficiency and accuracy of monitoring the abnormal state of the motor. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to make the purpose, technical solutions and advantages of the present invention more clear, the present invention will be further described below with reference to the accompanying drawings, in which:

[0047] Figure 1 A flow chart of an embodiment of a motor state monitoring method based on data drive provided by the present invention;

[0048] Figure 2 This is a structural block diagram of an embodiment of a data-driven motor state monitoring system provided by the present invention. DETAILED DESCRIPTION

[0049] Various exemplary embodiments of the present disclosure will now be described in detail with reference to the accompanying drawings. The description of the exemplary embodiments is merely illustrative and is in no way intended to limit the present disclosure and its application or use. The present disclosure can be implemented in many different forms and is not limited to the embodiments described herein. These embodiments are provided to make the present disclosure thorough and complete and to fully express the scope of the present disclosure to those skilled in the art. It should be noted that unless otherwise specifically stated, the relative arrangement of the components and steps, the composition of the materials, the numerical expressions and the numerical values ​​set forth in these embodiments should be interpreted as being merely exemplary and not as limiting.

[0050] The words "first", "second" and similar words used in this disclosure do not indicate any order, quantity or importance, but are only used to distinguish different parts. The words "include" or "comprises" and similar words mean that the elements before the word include the elements listed after the word, and do not exclude the possibility of including other elements. "Up", "down" and the like are only used to indicate relative positional relationships. When the absolute position of the described object changes, the relative positional relationship may also change accordingly.

[0051] In the present disclosure, when a specific component is described as being located between a first component and a second component, there may or may not be an intermediate component between the specific component and the first component or the second component. When a specific component is described as being connected to other components, the specific component may be directly connected to the other components without an intermediate component, or may not be directly connected to the other components but have an intermediate component.

[0052] All terms (including technical terms or scientific terms) used in the present disclosure have the same meanings as those understood by ordinary technicians in the field to which the present disclosure belongs, unless otherwise specifically defined. It should also be understood that terms defined in general dictionaries, such as general dictionaries, should be interpreted as having meanings consistent with their meanings in the context of the relevant technology, and should not be interpreted in an idealized or extremely formal sense, unless explicitly defined herein.

[0053] Technologies, methods, and equipment known to ordinary technicians in the relevant art may not be discussed in detail, but where appropriate, the technologies, methods, and equipment should be considered part of the specification.

[0054] like Figure 1 As shown, the data-driven motor state monitoring method provided in this embodiment includes the following steps in actual execution:

[0055] Step S1, dividing the motor operating conditions after the motor runs stably to obtain a steady-state operating condition list, wherein the steady-state operating condition list includes several combination operating conditions of different stable speed operating conditions and different current operating conditions.

[0056] In order to obtain effective motor operating conditions, before conducting machine learning training, it is necessary to effectively divide the different operating conditions of the motor, adaptively divide the training data file generated in the configuration phase according to several stable speeds in the feature file, and produce a steady-state condition list file.

[0057] In one embodiment of the data-driven motor state monitoring method of the present invention, step S1 may specifically include:

[0058] Step S11: preliminarily divide the data in the operating condition file according to the stable speed vector to obtain a number of preliminary groups.

[0059] Specifically, the data of different speeds in the working condition file are divided according to the stable speed vector to obtain a plurality of preliminary groups, and the data of the same speed is taken as the same preliminary group, wherein the number of the preliminary groups corresponds to the number of the stable speed vectors, and in the same preliminary group, the same speed working condition corresponds to a plurality of different current working conditions,

[0060] Step S12: Mark the current data in each of the preliminary groups with a steady-state current label.

[0061] In a specific implementation, first, the current data in each preliminary group is rounded. Specifically, for all current data in each preliminary group, the current data whose difference from the target current value is within a preset range is classified into the same operating condition category, wherein the number of operating condition categories is consistent with the number of target current values. Further, exemplarily, the preset range includes -5%-5%. It should be noted that the present invention does not specifically limit the target current value and the preset range; then, the current data in each preliminary group is marked with a steady-state current label according to the rounded result of the current data.

[0062] Step S13: performing secondary segmentation on each preliminary grouping according to the steady-state current label to obtain a final segmentation result.

[0063] In one embodiment of the data-driven motor state monitoring method of the present invention, step S13 may specifically include:

[0064] Step S131 : In each of the preliminary groups, data marked with the same steady-state current label are regarded as the same fine group.

[0065] Step S132: obtaining a steady-state operating condition list according to each of the detailed groups in each of the preliminary groups, wherein the steady-state operating condition list includes combined operating conditions of different speeds and different currents.

[0066] Step S133: output the steady-state operating condition list as a CSV file including speed variables and current variables, wherein the CSV file includes the operating conditions corresponding to each speed, and each speed corresponds to several current operating conditions.

[0067] Step S2: Based on the segmentation results of the motor operating conditions, a number of operating condition anomaly detection models corresponding to the combined operating conditions in the steady-state operating condition list are trained, wherein the number of the operating condition anomaly detection models corresponds to the number of the combined operating conditions.

[0068] In one embodiment of the data-driven motor state monitoring method of the present invention, step S2 may specifically include:

[0069] Step S21, use the operating condition serial number and the sixteen-dimensional feature matrix file corresponding to each combination of operating conditions in the steady-state operating condition list as training data, and input them into the preset machine learning model, wherein each row of the sixteen-dimensional feature matrix represents a sample, the first column is time information, the second column represents speed, the third column represents stable state, the fourth column represents temperature, and the remaining columns represent 12 time domain and frequency domain features.

[0070] Step S22: adopting the method of isolating abnormal points, using the training data to train the preset machine learning model, and obtaining an operating condition anomaly detection model corresponding to each combined operating condition in the steady-state operating condition list.

[0071] Existing anomaly detection methods mainly describe normal samples and give a region of normal samples in the feature space. Samples that are not in this region are considered abnormal. The main disadvantage of these anomaly detection methods is that they only optimize the description of normal samples but not the description of abnormal samples, which may cause a large number of false positives or only detect a small number of anomalies.

[0072] The anomaly detection method of the present invention no longer describes normal sample points, but isolates abnormal points and uses statistics to explain that in the data space, sparsely distributed areas indicate that the probability of data occurring in this area is very low, so the data falling in these areas can be considered abnormal. In order to find out which points are easy to be isolated, it is assumed that a random hyperplane is used to cut the data space, and two subspaces can be generated by cutting once. Then continue to use a random hyperplane to cut each subspace, and cycle until there is only one data point in each subspace. It is understandable that those clusters with very high density can be cut many times before cutting stops, but those points with very low density can easily stop in a subspace very early.

[0073] Further, in one embodiment of the training method of the motor operating abnormality detection model of the present invention, the step S22 may specifically include:

[0074] Step S221: construct an isolation forest consisting of a plurality of isolation trees based on the training data.

[0075] Further, in one implementation of the method for training the motor operating abnormality detection model of the present invention, the step S221 may specifically include:

[0076] Step A1: Randomly select Ψ samples from the training data and put them into the root node of an isolation tree.

[0077] Step A2: Randomly specify a dimension and randomly generate a cutting point p in the current node data, wherein the cutting point p is generated between the maximum value and the minimum value of the specified dimension in the current node data.

[0078] Step A3: Generate a hyperplane with a randomly generated cutting point p, and then divide the current node data space into 2 subspaces.

[0079] Step A4: Place the data less than p in the specified dimension in the left subtree of the current node, and place the data greater than or equal to p in the right subtree of the current node.

[0080] Step A5: Recursively perform steps A2 and A3 in the subtree branch, and continuously construct new leaf nodes until there is only one data in the leaf node and it is no longer possible to cut or the tree has reached a limited height.

[0081] Step S222: traverse the isolation forest for the data to be detected for anomalies, calculate anomaly scores, and determine whether there are abnormal data in the data to be detected for anomalies according to the anomaly scores.

[0082] Specifically, for each sample x, the results of each tree are calculated and the anomaly score is calculated using the formula:

[0083]

[0084] Among them, h(x) is the height of sample x in each tree, that is, the path length, and c(Ψ) is the average value of the path length when the number of samples Ψ is given, which is used to standardize the path length h(x) of sample x.

[0085] Step S3, matching the current operating condition of the motor with each combined operating condition in the steady-state operating condition list file.

[0086] In one embodiment of the data-driven motor state monitoring method of the present invention, step S3 may specifically include:

[0087] Step S31, matching the speed and current of the motor in the current working condition with the stable speed working condition and current working condition in the steady-state working condition list file respectively.

[0088] In order to determine whether the current motor working condition belongs to a certain working condition in the steady-state working condition list, it is necessary to match its speed and current respectively. Specifically, firstly, the data format of the speed and current in the current working condition of the motor is reorganized; then the speed and current in the current working condition of the motor after the reorganization data format are matched with the stable speed working condition and current working condition in the steady-state working condition list file respectively.

[0089] Furthermore, a method for reorganizing the data format of the speed and current in the current working condition of the motor may be, for example: converting the data format of the speed and current in the current working condition of the motor respectively so that the units of the speed and current in the current working condition of the motor are consistent with the units of the stable speed condition and current condition in the steady-state working condition list file.

[0090] Furthermore, a method for matching the speed and current in the current operating condition of the motor after the data format is reformatted with the stable speed condition and current condition in the steady-state operating condition list file, for example, may be: matching the speed in the current operating condition of the motor after the data format is reformatted with the stable speed condition in the steady-state operating condition list file; if they match, matching the current in the current operating condition of the motor after the data format is reformatted with the current condition in the steady-state operating condition list file.

[0091] Furthermore, the method for matching the speed (or current) of the motor in the current operating condition after the data format is reformatted with the stable speed condition (or current condition) in the steady-state condition list file can be specifically as follows: matching the speed (or current) of the motor in the current operating condition after the data format is reformatted with several stable speed conditions in the steady-state condition list file; if the speed (or current) of the motor in the current operating condition after the data format is reformatted matches one of the stable speed conditions (or current conditions) in the steady-state condition list file, then the successfully matched stable speed condition (or current condition) is used as the matching speed condition (matching current condition); if the speed (or current) of the motor in the current operating condition after the data format is not matched with any of the stable speed conditions (or current conditions) in the steady-state condition list file, then a prompt message of unsuccessful speed (current) matching is output, and the matching process is ended.

[0092] Step S32: obtaining a matching result of the current operating condition of the motor according to the matching results of the rotational speed and the current, so as to determine a matching operating condition that matches the current operating condition of the motor from among the combined operating conditions in the steady-state operating condition list.

[0093] Specifically, if the speed and current in the current operating condition of the motor are matched successfully, the operating condition number corresponding to the combination of the matching speed condition and the matching current condition in the steady-state operating condition list file is used as the matching result of the motor operating condition at the current moment; if the speed in the current operating condition of the motor is not matched successfully, or the speed is matched successfully but the current is not matched successfully, it is determined that the motor operating condition at the moment does not belong to any operating condition in the steady-state operating condition list file.

[0094] Furthermore, when the speed and current in the current operating condition of the motor are successfully matched, the operating condition sequence number corresponding to the combined operating condition matching the current operating condition of the motor in the steady-state operating condition list is output as the matching result of the current operating condition of the motor, and a 13-dimensional feature parameter list is output at the same time, in which the first column is temperature and the next 12 columns are time domain and frequency domain features.

[0095] Further, after step S2 and before step S3, the data-driven motor state monitoring method further includes:

[0096] The configuration parameters are set to obtain a configuration file, wherein the configuration parameters include a 13-dimensional feature parameter list, a stable speed vector, and a steady-state operating condition list at the current moment.

[0097] Step S4: obtaining a motor state monitoring result according to a matching result of the current working condition of the motor and a working condition abnormality detection model corresponding to the matching result.

[0098] The abnormal working condition judgment model is constructed based on the working condition sequence number and the 16-dimensional feature file, and the 16-dimensional feature file includes time information, speed, stable state, temperature and 12 time domain and frequency domain features. In one embodiment of the data-driven motor state monitoring method of the present invention, the step S4 may specifically include:

[0099] Step S41, determining, from among several operating condition abnormality judgment models corresponding to several combined operating conditions in the steady-state operating condition list, an operating condition abnormality judgment model corresponding to the matching operating condition as a matching operating condition abnormality judgment model.

[0100] In one implementation of the data-driven motor state monitoring method of the present invention, step S41 may specifically include:

[0101] Step S411, establishing a correspondence between the model serial number of each of the abnormal operating condition judgment models and the operating condition serial number of each of the combined operating conditions in the steady-state operating condition list.

[0102] Step S412, according to the operating condition serial number corresponding to the combined operating condition matching the current operating condition of the motor in the steady-state operating condition list, and the correspondence between the model serial number and the operating condition serial number, determine the operating condition abnormality judgment model corresponding to the matching operating condition as the matching operating condition abnormality judgment model.

[0103] Step S42: Obtain a motor state monitoring result according to the matching working condition and the matching working condition abnormality judgment model.

[0104] In one implementation of the data-driven motor state monitoring method of the present invention, step S42 may specifically include:

[0105] Step S421: Determine whether it is necessary to obtain the motor state monitoring result by using a matching operating condition abnormality judgment model according to the relationship between the temperature in the 13-dimensional feature parameter list and the preset temperature threshold.

[0106] Step S422: When the temperature is less than or equal to the preset temperature threshold, it is determined that the motor state is normal.

[0107] Step S423, when the temperature is greater than the preset temperature threshold, the motor state monitoring result is obtained by using the matching operating condition abnormality judgment model, and the operating condition serial number corresponding to the combined operating condition matching the current operating condition of the motor in the steady-state operating condition list and the 13-dimensional feature parameter list are input into the matching operating condition abnormality judgment model to obtain the cumulative number of abnormalities output by the matching operating condition abnormality judgment model.

[0108] Step S424: If the accumulated number of abnormalities meets the preset abnormal condition, it is determined that the motor state is abnormal; if the accumulated number of abnormalities does not meet the preset abnormal condition, it is determined that the motor state is normal.

[0109] Among them, the preset abnormal condition includes that the abnormal cumulative number is greater than or equal to the preset abnormal cumulative number threshold, so that the false alarm rate can be reduced. Exemplarily, the preset abnormal cumulative number threshold is 3-7 times, for example, 5 times. It should be noted that the present invention does not specifically limit the preset abnormal cumulative number threshold. The abnormal cumulative number is the number of times the matching working condition abnormality judgment model continuously outputs abnormal results.

[0110] The data-driven motor state monitoring method provided in the embodiment of the present invention is beneficial to obtaining effective motor operating conditions by segmenting the motor operating conditions. On the basis of data analysis, a working condition abnormality detection model is trained, and based on the working conditions matching the current working conditions of the motor in the steady-state working condition list file and the working condition abnormality judgment model corresponding to the matching working conditions, the current motor working conditions are analyzed to automatically and directly obtain the motor state monitoring results, which solves the problem in the prior art that it is difficult to effectively determine whether the motor is in an abnormal state, and improves the efficiency and accuracy of monitoring the abnormal state of the motor.

[0111] Accordingly, if Figure 2 As shown, the present invention also provides a data-driven motor state monitoring system, comprising:

[0112] The working condition segmentation module 1 is used to segment the motor working condition after the motor runs stably to obtain a steady-state working condition list, wherein the steady-state working condition list includes several combined working conditions of different stable speed working conditions and different current working conditions;

[0113] Model training module 2, used for training a plurality of abnormal working condition detection models corresponding to each combined working condition in the steady-state working condition list according to the segmentation result of the motor working condition, wherein the number of the abnormal working condition detection models corresponds to the number of the combined working conditions;

[0114] The working condition matching module 3 is used to match the current working condition of the motor with each combined working condition in the steady-state working condition list file;

[0115] The abnormality detection module 4 is used to obtain the motor state monitoring result according to the matching result of the current working condition of the motor and the working condition abnormality detection model corresponding to the matching result.

[0116] The data-driven motor state monitoring system provided by the embodiment of the present invention utilizes a working condition segmentation module to segment the motor working condition, which is conducive to obtaining effective motor operating conditions. On the basis of data analysis, a working condition abnormality detection model is obtained using a model training module, and through a working condition matching module and an abnormality detection module, based on the working condition matching the current working condition of the motor in a steady-state working condition list file and a working condition abnormality judgment model corresponding to the matching working condition, the current motor working condition is analyzed, and the motor state monitoring result is automatically and directly obtained, which solves the problem in the prior art that it is difficult to effectively determine whether the motor is in an abnormal state, and improves the efficiency and accuracy of monitoring the abnormal state of the motor.

[0117] It should be understood that the above Figure 2 The division of the various components of the data-driven motor condition monitoring system shown is only a division of logical functions. In actual implementation, they can be fully or partially integrated into one physical entity, or they can be physically separated. And these components can all be implemented in the form of software calling through processing elements; they can also be all implemented in the form of hardware; some components can also be implemented in the form of software calling through processing elements, and some components can be implemented in the form of hardware. For example, one of the above modules can be a separately established processing element, or it can be integrated in a chip of an electronic device. The implementation of other components is similar. In addition, all or part of these components can be integrated together, or they can be implemented independently. In the implementation process, each step of the above method or the above components can be completed by the hardware integrated logic circuit in the processor element or the instructions in the form of software.

[0118] So far, various embodiments of the present disclosure have been described in detail. In order to avoid obscuring the concept of the present disclosure, some details known in the art are not described. Based on the above description, those skilled in the art can fully understand how to implement the technical solution disclosed here.

[0119] Although some specific embodiments of the present disclosure have been described in detail by way of examples, it should be understood by those skilled in the art that the above examples are for illustration only and are not intended to limit the scope of the present disclosure. It should be understood by those skilled in the art that the above embodiments may be modified or some technical features may be replaced by equivalents without departing from the scope and spirit of the present disclosure. The scope of the present disclosure is defined by the appended claims.

Claims

1. A data-driven motor state monitoring method, characterized in that: include: The motor operating conditions after the motor runs stably are divided to obtain a steady-state operating condition list, wherein the steady-state operating condition list includes several combined operating conditions of different stable speed operating conditions and different current operating conditions; Specifically, the method comprises: preliminarily dividing the data in the working condition file according to the stable speed vector to obtain a plurality of preliminary groups, dividing the data of different speeds in the working condition file according to the stable speed vector to obtain a plurality of preliminary groups, and taking the data of the same speed as the same preliminary group, wherein the number of the preliminary groups corresponds to the number of the stable speed vectors, and in the same preliminary group, the same speed working condition corresponds to a plurality of different current working conditions; marking the current data in each of the preliminary groups with a steady-state current label; Performing secondary segmentation on each preliminary grouping according to the steady-state current label to obtain a final segmentation result, specifically comprising: in each preliminary grouping, treating data marked with the same steady-state current label as the same fine grouping; According to each of the detailed groups in each of the preliminary groups, a steady-state operating condition list is obtained, wherein the steady-state operating condition list includes combined operating conditions of different speeds and different currents; Outputting the steady-state operating condition list as a CSV file including speed variables and current variables, wherein the CSV file includes operating conditions corresponding to each speed, and each speed corresponds to a number of current operating conditions; According to the segmentation result of the motor working condition, a number of working condition abnormality detection models corresponding to each combined working condition in the steady-state working condition list are trained, wherein the number of the working condition abnormality detection models corresponds to the number of the combined working conditions; The method for obtaining the abnormal working condition detection model is as follows: the working condition serial number and the sixteen-dimensional feature matrix file corresponding to each combined working condition in the steady-state working condition list are used as training data, and are input into a preset machine learning model, wherein each row of the sixteen-dimensional feature matrix represents a sample, the first column is time information, the second column represents speed, the third column represents steady state, the fourth column represents temperature, and the remaining columns represent 12 time domain and frequency domain features; the method of isolating abnormal points is adopted, and the preset machine learning model is trained using the training data to obtain the abnormal working condition detection model corresponding to each combined working condition in the steady-state working condition list; Matching the current working condition of the motor with each combined working condition in the steady-state working condition list file; Specifically, the method includes: matching the speed and current in the current working condition of the motor with the stable speed working condition and current working condition in the steady-state working condition list file respectively; obtaining the matching result of the current working condition of the motor according to the matching result of the speed and current, so as to determine the matching working condition that matches the current working condition of the motor in each combination working condition in the steady-state working condition list; Obtaining a motor state monitoring result according to a matching result of the motor's current working condition and a working condition abnormality detection model corresponding to the matching result; Specifically, it includes: determining an operating condition abnormality judgment model corresponding to the matching operating condition among several operating condition abnormality judgment models corresponding to several combined operating conditions in the steady-state operating condition list as a matching operating condition abnormality judgment model; and obtaining a motor state monitoring result according to the matching operating condition and the matching operating condition abnormality judgment model.

2. The data-driven motor state monitoring method according to claim 1, characterized in that: The step of obtaining the matching result of the current working condition of the motor according to the matching result of the rotational speed and the current, so as to determine the matching working condition matching the current working condition of the motor from among the combined working conditions in the steady-state working condition list, specifically includes: When the speed and current in the current working condition of the motor are successfully matched, the working condition sequence number corresponding to the combined working condition matching the current working condition of the motor in the steady-state working condition list is output as the matching result of the current working condition of the motor, and a 13-dimensional feature parameter list is output at the same time, in which the first column is temperature and the next 12 columns are time domain and frequency domain features.

3. Steady-state operating condition list The data-driven motor state monitoring method according to claim 2 is characterized in that: The determining, from among the several operating condition abnormality judgment models corresponding to the several combined operating conditions in the steady-state operating condition list, the operating condition abnormality judgment model corresponding to the matching operating condition as the matching operating condition abnormality judgment model specifically includes: Establishing a correspondence between the model serial number of each of the abnormal working condition judgment models and the working condition serial number of each of the combined working conditions in the steady-state working condition list; According to the operating condition serial number corresponding to the combined operating condition matching the current operating condition of the motor in the steady-state operating condition list, and the correspondence between the model serial number and the operating condition serial number, the operating condition abnormality judgment model corresponding to the matching operating condition is determined as the matching operating condition abnormality judgment model.

4. The data-driven motor state monitoring method according to claim 3 is characterized in that: The obtaining of the motor state monitoring result according to the matching working condition and the matching working condition abnormality judgment model specifically includes: According to the relationship between the temperature in the 13-dimensional feature parameter list and the preset temperature threshold, determine whether it is necessary to use a matching operating condition abnormality judgment model to obtain a motor state monitoring result; When the temperature is less than or equal to the preset temperature threshold, it is determined that the motor state is normal; When the temperature is greater than a preset temperature threshold, the motor state monitoring result is obtained by using a matching working condition abnormality judgment model, and the working condition sequence number corresponding to the combined working condition matching the current working condition of the motor in the steady-state working condition list and the 13-dimensional feature parameter list are input into the matching working condition abnormality judgment model to obtain the cumulative number of abnormalities output by the matching working condition abnormality judgment model; If the accumulated number of abnormalities meets the preset abnormal condition, it is determined that the motor state is abnormal; if the accumulated number of abnormalities does not meet the preset abnormal condition, it is determined that the motor state is normal.

5. A data-driven motor condition monitoring system using the method according to any one of claims 1 to 4, characterized in that: include: A working condition segmentation module, used for segmenting the motor working condition after the motor runs stably to obtain a steady-state working condition list, wherein the steady-state working condition list includes several combined working conditions of different stable speed working conditions and different current working conditions; A model training module, used for training a plurality of abnormal working condition detection models corresponding to each combined working condition in the steady-state working condition list according to the segmentation result of the motor working condition, wherein the number of the abnormal working condition detection models corresponds to the number of the combined working conditions; A working condition matching module, used for matching the current working condition of the motor with each combined working condition in the steady-state working condition list file; The abnormality detection module is used to obtain the motor state monitoring result according to the matching result of the current working condition of the motor and the working condition abnormality detection model corresponding to the matching result.