A control method and system for a DC brushless reduction motor

Abstract

The present invention relates to the technical field of auxiliary control of DC brushless geared motors, and discloses a control method and system for DC brushless geared motors. The method includes: based on the real-time data during motor start and stop, including detection data (such as temperature, vibration, noise, etc.), operation data (such as current, voltage, torque, etc.), and load data, by comparing and analyzing with the preset historical detection data, to judge whether the motor has a fault: when a large data difference is detected, the severity of the motor fault is evaluated through a fault scoring mechanism. If the fault score exceeds the preset threshold, an early warning is triggered and the corresponding control method is selected according to the fault type to perform fault handling or adjust the control strategy to ensure the stable operation of the motor. The present invention optimizes the motor maintenance and fault diagnosis processes by combining fault scoring and dynamic adjustment of control strategies, through real-time monitoring and analysis of operation data and load data, and improves the efficiency and accuracy of fault response.

Description

Technical Field

[0001] The present invention relates to the technical field of auxiliary control of DC brushless reduction motors, and more particularly, to a control method and system for DC brushless reduction motors. Background Art

[0002] With the rapid development of modern industrial automation, DC brushless reduction motors are widely used in various mechanical equipment, such as automated production lines, robots, aerospace and other fields. These motors have the advantages of high efficiency, reliability and low maintenance cost, and have become an important power source especially in applications that require precise control.

[0003] However, the operation of the motor is affected by various factors, such as current, voltage fluctuations, load changes, environmental temperature, etc. These factors may cause the performance of the motor to decline or faults to occur. To ensure the stable operation of the motor and timely detect potential faults and give early warnings are the keys to improving the reliability of the equipment and extending its service life. At present, traditional motor fault diagnosis methods usually rely on manual inspection and regular maintenance, and cannot real-time monitor the operation status of the motor, nor can they timely detect potential fault hazards. This method not only increases the labor cost, but also may miss the best repair opportunity, resulting in an increase in equipment downtime and affecting production efficiency.

[0004] Therefore, there is an urgent need to invent a control technology for DC brushless reduction motors to solve the problems that traditional motor fault diagnosis methods rely on manual inspection and regular maintenance, cannot real-time monitor the motor status and timely detect potential faults, resulting in an increase in equipment downtime and a decrease in production efficiency. Summary of the Invention

[0005] In view of this, the present invention proposes a control method and system for DC brushless reduction motors, aiming to solve the problems that traditional motor fault diagnosis methods rely on manual inspection and regular maintenance, cannot real-time monitor the motor status and timely detect potential faults, resulting in an increase in equipment downtime and a decrease in production efficiency.

[0006] The present invention proposes a control method for DC brushless reduction motors, including:

[0007] Obtain the detection data, operation data and load data when the motor starts and stops, and determine whether the motor operation fails according to the relationship between the detection data and the set preset detection data, where:

[0008] When it is determined that the motor operation fails, the relationship between the detection data and the set preset detection data determines the fault score of the motor;

[0009] Determine the fault type of the motor according to the operation data and load data;

[0010] Determine whether to give a warning to the motor according to the relationship between the fault score and the preset fault score configured. Among them, when the fault score is lower than the preset fault score, determine the warning method according to the score difference between the fault score and the preset fault score; when the fault score is higher than or equal to the preset fault score, determine the control method according to the fault type.

[0011] Further, when setting the preset detection data, it includes:

[0012] Obtain the historical detection data of each motor. Among them, the detection data includes the temperature detection data, vibration detection data and noise detection data of the motor;

[0013] Obtain the distance metric between the historical detection data, and establish a distance matrix for the historical detection data according to the distance metric;

[0014] Perform iterative clustering on the historical detection data according to the distance matrix. According to the clustering results, obtain the feature vectors of the normal detection data and the feature vectors of the abnormal detection data;

[0015] Determine the preset detection data according to the feature vector of the normal detection data;

[0016] Obtain the distance value between the feature vector of the normal detection data and the feature vector of the abnormal detection data, and determine it as the preset difference amount.

[0017] Further, when determining whether a fault occurs in the motor operation according to the relationship between the detection data and the set preset detection data, it includes:

[0018] Obtain the data difference between the detection data and the preset detection data. According to the relationship between each data difference and the preset difference amount, determine whether a fault occurs in the motor operation:

[0019] When the data difference is lower than the preset difference amount, it is determined that no fault occurs in the motor operation;

[0020] When the data difference is equal to or higher than the preset difference amount, it is determined that a fault occurs in the motor operation, and according to the relationship between the data difference and the preset difference amount, determine the fault score of the motor.

[0021] Further, when determining the fault score of the motor according to the relationship between the data difference and the preset difference amount, it includes:

[0022] ;

[0023] Among them, S is the fault score of the motor, wi is the weight coefficient of the i-th data item, α i is the non-linear coefficient of the i-th data item, t is the temperature data item, v is the vibration data item, c is the noise data item, e -βiis the smoothing coefficient, △di is the data difference, and △Di is the preset difference amount.

[0024] Further, when obtaining the operating data and load data of the motor during startup and shutdown, it includes:

[0025] When obtaining the operating data, it includes obtaining the starting current, starting voltage, starting torque, start-stop time interval, and start-stop speed of the motor;

[0026] When obtaining the load data, it includes obtaining the inertial load, friction load, torque load, and current load carried by the motor.

[0027] Further, when determining the fault type of the motor based on the operating data and load data, it includes:

[0028] Obtain the historical operating data and historical load data of the motor, and establish a motor fault diagnosis model based on the relationship between the decision tree, neural network, historical operating data, and historical load data;

[0029] Substitute the operating data and load data of the motor into the motor fault diagnosis model to determine the fault type of the motor, and determine the execution operation during control according to the fault type.

[0030] Further, when establishing a motor fault diagnosis model based on the relationship between the decision tree, neural network, historical operating data, and historical load data, it includes:

[0031] Establish a motor state correlation formula based on the historical operating data and the corresponding motor state, and historical load data and the corresponding motor state;

[0032] Based on the decision tree, screen each motor state correlation formula to determine the abnormal data between each motor state correlation formula:

[0033] ;

[0034] Among them, A is the contribution value of the motor correlation formula, H(F) is the entropy of the motor state correlation formula in the data set F; Fj is the jth subset after the division of the motor state correlation formula, and K is the total number of data subsets;

[0035] Obtain the median of the contribution values between the contribution values of each motor correlation formula, and determine the abnormal data according to the relationship between the contribution value and the median of the contribution values:

[0036] When the contribution value of the motor correlation formula is less than the median of the contribution values, then determine the historical operating data and historical load data in the motor correlation formula as abnormal data;

[0037] When the contribution value associated with the motor is greater than or equal to the median value of the contribution values, the historical operation data and historical load data in the motor - associated formula are determined to be non - abnormal data;

[0038] Based on the determined abnormal data, the motor fault type and feature vector are determined based on a neural network, and a motor fault diagnosis model is established according to the motor fault type and feature vector.

[0039] Furthermore, when determining the motor fault type and feature vector based on the determined abnormal data using a neural network, it includes:

[0040] ;

[0041] where f(X) is the abnormal data, is the probability of the K - th fault type, and are weight coefficients, and and are both non - zero, σ is the activation function, xi is the i - th feature of the feature data, wij is the weight from the input feature xi to the j - th neuron in the hidden layer, bj is the bias of the j - th neuron, h is the number of motor fault types, n is the number of input features, and j is the neuron index in the hidden layer.

[0042] Furthermore, when determining the execution operation during control according to the fault type, it includes:

[0043] According to the fault type, a first execution operation and a second execution operation are determined. The motor is controlled according to the first execution operation, and the real - time operation data and real - time load data of the motor during control are obtained;

[0044] The operation data difference between the real - time operation data and the historical operation data in the adjacent time period is obtained, and the load data difference between the real - time load data and the historical load data in the adjacent time period is obtained;

[0045] According to the relationship between the operation data difference and the load data difference, it is determined whether to use the second execution operation to control the motor:

[0046] When the operation data difference is lower than or equal to the configured preset operation data difference, and / or the load data difference is lower than or equal to the configured preset load data difference, it is determined not to use the second execution operation to control the motor;

[0047] When the operation data difference is higher than the preset operation data difference, and / or the load data difference is higher than the preset load data difference, it is determined to use the second execution operation to control the motor.

[0048] Compared with the prior art, the beneficial effects of the present invention are as follows: By obtaining the detection data, operation data, and load data of the motor and comparing these real-time data with the set preset detection data, it is possible to efficiently identify whether the motor is abnormal. This method not only avoids the traditional method relying on manual inspection but also can perform real-time monitoring at each stage of the motor operation, ensuring that potential faults can be detected and processed in the initial stage, reducing the downtime and maintenance costs caused by faults. In addition, by introducing a fault scoring mechanism, the fault state of the motor can be quantified. When a fault occurs during the operation of the motor, a fault score is calculated based on the relationship between the detection data and the preset data. This score not only helps to determine whether the motor has a fault but also can accurately evaluate the severity of the fault. After comparing with the preset fault score, it is possible to determine whether to activate the warning mechanism based on the score difference. This warning method based on the score difference makes the fault diagnosis of the motor more refined and intelligent, avoiding the previous extensive judgment relying solely on fixed thresholds. Finally, when the fault score is higher than or equal to the preset value, the corresponding control method can be automatically selected and executed according to the fault type. This control method can not only take the most appropriate countermeasures for different types of faults but also dynamically adjust according to the real-time operation state of the motor to ensure the stability and safety of the motor operation. When dealing with more complex faults, a backup control scheme will be automatically selected, and the control parameters of the motor will be further optimized based on historical load data or other relevant data to achieve the best operation state.

[0049] On the other hand, the present application also provides a control system for a DC brushless reduction motor, including:

[0050] An acquisition module configured to acquire the detection data, operation data, and load data when the motor starts and stops;

[0051] An analysis module electrically connected to the acquisition module. The analysis module is configured to determine whether a fault occurs in the motor operation according to the relationship between the detection data and the set preset detection data, where: when it is determined that a fault occurs in the motor operation, the relationship between the detection data and the set preset detection data determines the fault score of the motor; the analysis module is also configured to determine the fault type of the motor according to the operation data and the load data;

[0052] A central control module electrically connected to the analysis module. The central control module is configured to determine whether to give a warning to the motor according to the relationship between the fault score and the configured preset fault score. Among them, when the fault score is lower than the preset fault score, the score difference between the fault score and the preset fault score is determined to determine the warning method; when the fault score is higher than or equal to the preset fault score, the control method is determined according to the fault type.

[0053] It is understandable that a control method and system for a DC brushless reduction motor in each of the above embodiments of the present invention have the same beneficial effects and will not be elaborated herein. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] By reading the detailed description of the preferred embodiments below, various other advantages and benefits will become clear to those of ordinary skill in the art. The drawings are only for the purpose of illustrating the preferred embodiments and are not considered to be a limitation of the present invention. Moreover, throughout the drawings, the same reference numerals are used to represent the same components. In the drawings:

[0055] Figure 1 is a flowchart of a control method for a DC brushless reduction motor provided by an embodiment of the present invention;

[0056] Figure 2 is a functional block diagram of a control system for a DC brushless reduction motor provided by an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0057] Hereinafter, exemplary embodiments of the present disclosure will be described in more detail with reference to the drawings. Although the exemplary embodiments of the present disclosure are shown in the drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided so that the present disclosure can be more thoroughly understood and the scope of the present disclosure can be fully conveyed to those skilled in the art. It should be noted that, without conflict, the embodiments in the present invention and the features in the embodiments can be combined with each other. The present invention will be described in detail below with reference to the drawings and in combination with the embodiments.

[0058] As Figure 1 shown, in some embodiments of the present application, this embodiment provides a control method for a DC brushless reduction motor, including:

[0059] Step S100: Obtain the detection data, operation data, and load data when the motor starts and stops, and determine whether the motor is operating abnormally according to the relationship between the detection data and the set preset detection data, where: when it is determined that the motor is operating abnormally, the relationship between the detection data and the set preset detection data determines the fault score of the motor.

[0060] Specifically, when setting the preset detection data, it includes: obtaining the historical detection data of the motor, where the detection data includes the temperature detection data, vibration detection data, and noise detection data of the motor. Obtaining the distance metric between the historical detection data, and establishing a distance matrix for the historical detection data according to the distance metric. Performing iterative clustering on the historical detection data according to the distance matrix, and obtaining the feature vectors of the normal detection data and the abnormal detection data according to the clustering results. Determining the preset detection data according to the feature vector of the normal detection data. Obtaining the distance value between the feature vector of the normal detection data and the feature vector of the abnormal detection data, and determining it as the preset difference amount.

[0061] Specifically, when determining whether the motor operation has a fault according to the relationship between the detection data and the set preset detection data, it includes: obtaining the data difference between the detection data and the preset detection data, and determining whether the motor operation has a fault according to the relationship between each data difference and the preset difference amount: when the data difference is lower than the preset difference amount, it is determined that the motor operation has no fault. When the data difference is equal to or higher than the preset difference amount, it is determined that the motor operation has a fault, and the fault score of the motor is determined according to the relationship between the data difference and the preset difference amount.

[0062] Specifically, when determining the fault score of the motor according to the relationship between the data difference and the preset difference amount, it includes:

[0063] ;

[0064] where S is the fault score of the motor, wi is the weight coefficient of the i-th data item, α i is the non-linear coefficient of the i-th data item, t is the temperature data item, v is the vibration data item, c is the noise data item, e -βi is the smoothing coefficient, △di is the data difference, and △Di is the preset difference amount.

[0065] It is understandable that preset detection data is constructed based on historical detection data such as temperature, vibration, and noise. After these historical data are processed by distance metric, a distance matrix is formed, and then iterative clustering analysis is performed on the data to identify normal and abnormal detection data. Through this analysis process, the feature vectors of normal data and abnormal data can be extracted, and these feature vectors are used to determine the preset detection data, thereby providing a basis for subsequent fault judgment and scoring. During the fault diagnosis process, the real-time detection data obtained will be compared with the preset detection data to calculate the data difference. This difference is used to judge whether the motor is operating normally. When the difference between the detection data and the preset data is lower than the set preset difference amount, it can be determined that the motor is operating without faults; while when the data difference is equal to or higher than the preset difference amount, it is considered that the motor has a fault, and the fault score of the motor is determined through this difference. This method can not only monitor the state of the motor in real time but also evaluate the health status of the motor by dynamically calculating the fault score. Further, in order to more accurately evaluate the motor fault, the calculation of the fault score involves the weighted sum and non-linear processing of multiple data items. Specifically, for temperature, vibration, and noise data, weight coefficients Wi are assigned to these data items respectively, and they are adjusted in combination with the non-linear coefficient α i for adjustment. These data items have different importance in the process of calculating the fault score, and the impacts of temperature, vibration, and noise are considered separately, so as to be able to finely reflect the specific situation of the motor fault. This method enables the contributions of different data items to the fault score to be reasonably reflected. In addition, in order to improve the stability and accuracy of the score, the calculation of the fault score also considers the smoothing coefficient e -βi , which helps to reduce the influence of data fluctuations on the results and makes the score more stable and reliable. The introduction of this smoothing coefficient effectively suppresses the errors that may be brought by noise data and further improves the accuracy and stability of fault diagnosis. Through the combination of these mathematical models and coefficients, the fault state of the motor can be more accurately reflected and a basis for subsequent fault handling can be provided. Finally, through comprehensive analysis of the relationship between the motor detection data and the preset data, with the help of clustering, weighting, non-linear adjustment, and smoothing processing, the fault degree of the motor is evaluated in real time and accurately, and a fault score is generated.

[0066] It can be seen that by real-time monitoring various detection data, operation data, and load data of the motor and combining with preset standard data, it is possible to effectively judge whether the motor has a fault and conduct a scoring. This method can help the motor management system detect potential fault risks in a timely manner, avoid the further development of faults, reduce the downtime caused by faults, and improve the reliability of the motor. By real-time monitoring and comparing with the preset data, the system can issue a warning before the fault occurs, thus providing an opportunity for maintenance personnel for early intervention and preventing serious consequences caused by the fault. In addition, by using historical data for clustering analysis and combining distance metrics and clustering results, it is possible to extract the feature vectors of normal and abnormal data, providing a more accurate basis for subsequent fault diagnosis. This method can ensure that the performance of the motor under different working conditions can be accurately identified, and reasonable preset data can be generated according to different types of abnormal data. This data-driven method reduces the error of human judgment and improves the diagnostic accuracy. By calculating the relationship between the data difference and the preset difference amount, the fault scoring mechanism is implemented. This can not only help judge whether the motor has a fault, but also evaluate the severity of the fault. When the fault score is higher than the preset value, corresponding control measures can be initiated to prevent the further aggravation of the fault, thus ensuring the stable operation of the motor. This scoring mechanism makes the fault diagnosis of the motor more intelligent and precise, avoiding the blindness of traditional methods. In addition, by using the method of weighted and non-linear adjustment, the weights of different data items in the fault scoring are refined. Data items such as temperature, vibration, and noise have different impacts on the health status of the motor. The introduction of non-linear coefficients and weight coefficients can adjust the contribution degree of each data item according to the actual situation, further improving the accuracy of the fault scoring. This enables the entire motor monitoring system to better adapt to various complex working conditions and make accurate judgments. Finally, the smoothing process of the fault score effectively reduces the noise in the data and avoids misjudgment caused by data fluctuations. Through the smoothing process, it is possible to better handle the instantaneous changes in sensor data and ensure the stability and reliability of the score.

[0067] Step S200: Determine the fault type of the motor according to the operation data and load data.

[0068] Step S300: Determine whether to give a warning to the motor according to the relationship between the fault score and the configured preset fault score. Among them, when the fault score is lower than the preset fault score, the warning method is determined according to the score difference between the fault score and the preset fault score. When the fault score is higher than or equal to the preset fault score, the control method is determined according to the fault type.

[0069] Specifically, when obtaining the operating data and load data of the motor during start and stop, it includes: when obtaining the operating data, it includes obtaining the starting current, starting voltage, starting torque, time interval of start and stop, and start and stop speed of the motor. When obtaining the load data, it includes obtaining the inertial load, friction load, torque load, and current load carried by the motor.

[0070] Specifically, when determining the fault type of the motor based on the operating data and load data, it includes: obtaining the historical operating data and historical load data of the motor, and establishing a motor fault diagnosis model based on the relationship between the decision tree, neural network, historical operating data, and historical load data. Substitute the operating data and load data of the motor into the motor fault diagnosis model to determine the fault type of the motor, and determine the execution operation during control according to the fault type.

[0071] Specifically, when establishing a motor fault diagnosis model based on the relationship between the decision tree, neural network, historical operating data, and historical load data, it includes: establishing a motor state correlation formula based on the historical operating data and the corresponding motor state, and historical load data and the corresponding motor state. Screen each motor state correlation formula based on the decision tree to determine the abnormal data between each motor state correlation formula:

[0072] 。

[0073] Where A is the contribution value of the motor correlation formula, H(F) is the entropy of the motor state correlation formula in the dataset F; Fj is the j-th subset after the division of the motor state correlation formula, and K is the total number of data subsets; obtain the median value of the contribution values between the contribution values of each motor correlation formula, and determine the abnormal data according to the relationship between the contribution value and the median value of the contribution value: when the contribution value of the motor correlation formula is less than the median value of the contribution value, then determine the historical operating data and historical load data in this motor correlation formula as abnormal data; when the contribution value of the motor correlation formula is greater than or equal to the median value of the contribution value, then determine the historical operating data and historical load data in this motor correlation formula as non-abnormal data; according to the determined abnormal data, determine the motor fault type and characteristic vector based on the neural network, and establish a motor fault diagnosis model according to the motor fault type and characteristic vector.

[0074] Specifically, when determining the motor fault type and characteristic vector based on the determined abnormal data using a neural network, it includes:

[0075] 。

[0076] Where f(X) is the abnormal data, is the probability of the K-th fault type, and are the weight coefficients, and and None of them is zero, σ is the activation function, xi is the i-th feature of the feature data, wij is the weight from the input feature xi to the j-th neuron in the hidden layer, bj is the bias of the j-th neuron, h is the number of motor fault types, n is the number of input features, and j is the neuron index in the hidden layer.

[0077] Specifically, when determining the execution operation during control according to the fault type, it includes: determining the first execution operation and the second execution operation according to the fault type, controlling the motor according to the first execution operation, and obtaining the real-time operation data and real-time load data of the motor during control. Obtaining the operation data difference between the real-time operation data and the historical operation data in the adjacent period, and obtaining the load data difference between the real-time load data and the historical load data in the adjacent period. Determining whether to use the second execution operation to control the motor according to the relationship between the operation data difference and the load data difference: when the operation data difference is lower than or equal to the configured preset operation data difference, and / or the load data difference is lower than or equal to the configured preset load data difference, it is determined not to use the second execution operation to control the motor. When the operation data difference is higher than the preset operation data difference, and / or the load data difference is higher than the preset load data difference, it is determined to use the second execution operation to control the motor.

[0078] It is understandable that by obtaining the operating data (such as starting current, starting voltage, starting torque, start-stop time interval, etc.) and load data (such as inertial load, friction load, torque load, current load, etc.) when the motor starts and stops, and combining these data with historical operating data and load data, a correlation model of the motor state is established through the decision tree algorithm. The core of this process is to detect abnormal data by constructing a state correlation formula, and to provide support for subsequent fault diagnosis by screening out data points that deviate greatly from the normal operating state. The motor state correlation formula established based on the decision tree helps to determine the abnormal data in the data set, and these abnormal data reflect the abnormal performance of the motor under specific conditions. By establishing the motor state correlation formula, using the decision tree to screen the motor state, calculating the contribution value of each state correlation formula, and partitioning the data through the entropy value, the abnormal data can be determined. The basis for judging abnormal data is the comparison between the contribution value of the motor correlation formula and the median of its contribution values. If it is less than the median, it is considered abnormal data. Then, based on the determined abnormal data, the neural network is used to further analyze the fault type and feature vector of the motor, and finally a motor fault diagnosis model is constructed to achieve accurate motor fault identification and prediction. Specifically, by analyzing the historical operating data and load data through the decision tree, a motor state correlation formula is established. The motor state correlation formula divides the motor data in different time periods and clusters the historical data under each type of state. Based on these data, the contribution value A of each motor correlation formula is calculated, and by calculating the entropy H(F) of each correlation formula, the complexity of the data set is evaluated. In this process, the data set F is divided into multiple subsets Fj, and each subset Fj contains different historical operating data and load data. Then, by calculating the median of the contribution values of each motor correlation formula, the abnormal data is further screened out. If the contribution value A of a certain motor correlation formula is less than the median of the contribution values, the historical operating data and historical load data in this motor correlation formula are considered abnormal data. On the contrary, if the contribution value of the motor correlation formula is greater than or equal to the median of the contribution values, these data are considered normal and do not belong to abnormal data. Through this method, normal data and abnormal data can be effectively distinguished, and more accurate input can be provided for subsequent fault diagnosis. After the abnormal data is selected, these data are input into the neural network model to further classify the fault types of the motor, so as to provide a basis for control decisions. The application of the neural network in fault type judgment plays a crucial role. By training the abnormal data, the neural network can output the probability of the fault type and perform a non-linear mapping on the data through the activation function σ, so as to determine the fault type and feature vector of the motor. The key to this process is that through continuous learning and optimization of the model, the specific fault type of the motor can be identified according to the historical data and feature vector, improving the accuracy and real-time performance of fault diagnosis. According to the motor fault type, the execution operation of the motor will be determined according to the preset control strategy next.The execution operation is divided into a first execution operation and a second execution operation. The first execution operation is a priority operation for performing preliminary control according to the fault type of the motor. During the control process, by obtaining the motor operation data and load data in real time and comparing them with the historical data, it is judged whether the second execution operation needs to be performed. Specifically, by calculating the difference between the real-time data and the historical data, it is judged whether the conditions for executing the second operation are met. If the difference is greater than the set threshold, the second operation is executed for further adjustment. Through this comprehensive control method, the control strategy can be dynamically adjusted during the motor operation based on the real-time data and the fault score. When a fault occurs in the motor, by means of multi-level data analysis and judgment, the most appropriate execution operation is selected, which can not only improve the operation efficiency of the motor, but also detect and solve the fault in a short time, ensuring the stable operation of the motor under different working conditions. This method can improve the accuracy and efficiency of motor fault diagnosis and treatment by combining decision trees, neural networks and refined control strategies, minimize the downtime of the equipment to the greatest extent, and extend the service life of the motor.

[0079] It can be seen that by combining the operating data and load data of the motor, a diagnostic model based on decision trees and neural networks provides a comprehensive and accurate method for motor fault detection and control. A main beneficial effect of this method is that by real-time monitoring the operating state and load state of the motor, it can accurately identify whether there is a fault in the motor and take corresponding control measures according to different fault scores. Compared with traditional maintenance methods, it can give early warnings of potential faults in real time, reduce the dependence on manual inspections, and greatly improve the efficiency and timeliness of fault diagnosis. Secondly, by setting a preset fault score and judging based on the difference between the actual fault score and the preset one, it can clearly define when to give a warning for the motor. When the fault score is higher than or equal to the preset fault score, a more in-depth control strategy is decided based on the fault type. This difference judgment mechanism ensures the stability of the motor operation, avoids premature warnings or ignoring potential problems, and thus reduces the risks during the motor operation under a more efficient control strategy. Through the analysis and modeling of the motor historical data based on decision trees and neural network models, it can extract the characteristics of motor faults from a large amount of historical data and accurately classify the fault types. This data-driven approach improves the accuracy of diagnosis, avoids the errors of human judgment, and makes the identification of motor fault types more scientific and systematic. In addition, based on the state-correlation screening method, it can effectively identify abnormal data that does not conform to the normal working mode, further improving the sensitivity of fault prediction. Combining with the feature vector processing of neural networks, it can extract more distinguishable features from complex abnormal data, making the identification of fault types more accurate. This multi-level and multi-dimensional data analysis method improves the reliability of fault diagnosis. Especially when facing the motor operation under various working conditions, it can maintain a high diagnostic accuracy and ensure that the motor is always in the best operating state. Finally, based on the comparison of the differences between real-time data and historical data, it can judge whether to perform a second control operation according to the changes in the differences of operating data and load data. This mechanism of dynamically adjusting control operations can first perform lightweight adjustments when the motor has minor faults. If the problem is not solved, more complex control strategies are then adopted. This step-by-step optimized control method not only reduces the risk of motor faults, but also improves the long-term operation stability of the motor, minimizing the maintenance cost and downtime to the greatest extent.

[0080] In the above embodiments, by obtaining the detection data, operation data, and load data of the motor and comparing these real-time data with the set preset detection data, it is possible to efficiently identify whether the motor is abnormal. This method not only avoids the traditional method relying on manual inspection but also enables real-time monitoring at each stage of the motor operation, ensuring that potential faults can be detected and processed in a timely manner at the initial stage, reducing the downtime and maintenance costs caused by faults. In addition, by introducing a fault scoring mechanism, the fault state of the motor can be quantified. When a fault occurs during the motor operation, a fault score is calculated based on the relationship between the detection data and the preset data. This score not only helps to determine whether the motor has a fault but also can accurately evaluate the severity of the fault. After comparing with the preset fault score, it is possible to determine whether to activate the warning mechanism based on the score difference. This warning method based on the score difference makes the fault diagnosis of the motor more refined and intelligent, avoiding the previous extensive judgment relying solely on fixed thresholds. Finally, when the fault score is higher than or equal to the preset value, the corresponding control method can be automatically selected and executed according to the fault type. This control method can not only take the most appropriate countermeasures for different types of faults but also dynamically adjust according to the real-time operation state of the motor to ensure the stability and safety of the motor operation. When dealing with more complex faults, a backup control scheme will be automatically selected, and the control parameters of the motor will be further optimized based on historical load data or other relevant data to achieve the best operation state.

[0081] In another preferred manner based on the above embodiments, as Figure 2 shown, this embodiment provides a control system for a DC brushless reduction motor, including: an acquisition module, an analysis module, and a central control module.

[0082] Specifically, the acquisition module is configured to obtain the detection data, operation data, and load data when the motor starts and stops. The analysis module is electrically connected to the acquisition module. The analysis module is configured to determine whether the motor operation is faulty according to the relationship between the detection data and the set preset detection data, where: when it is determined that the motor operation is faulty, the relationship between the detection data and the set preset detection data determines the fault score of the motor. The analysis module is further configured to determine the fault type of the motor according to the operation data and the load data. The central control module is electrically connected to the analysis module. The central control module is configured to determine whether to give a warning to the motor according to the relationship between the fault score and the configured preset fault score. Among them, when the fault score is lower than the preset fault score, the score difference between the fault score and the preset fault score is determined to determine the warning method. When the fault score is higher than or equal to the preset fault score, the control method is determined according to the fault type.

[0083] It is understandable that the control method and system for a DC brushless reduction motor in each of the above embodiments of the present invention have the same beneficial effects and will not be elaborated herein.

[0084] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a system, or a computer program product. Therefore, the present application can adopt the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can adopt the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0085] The present application is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to the embodiments of the present application. It should be understood that each flow and / or block in the flowcharts and / or block diagrams can be implemented by computer program instructions, and the combination of flows and / or blocks in the flowcharts and / or block diagrams can also be implemented. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, such that the instructions executed by the processor of the computer or other programmable data processing devices generate a device for implementing the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0086] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory generate a manufactured article including an instruction device, and the instruction device implements the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0087] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process, and thus the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0088] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention has been described in detail with reference to the above embodiments, those of ordinary skill in the art should understand that: modifications or equivalent substitutions can still be made to the specific implementation manners of the present invention, and any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the protection scope of the claims of the present invention.

Claims

1. A control method for a brushless DC reduction motor, characterized in that: include: The detection data, operation data and load data of the motor when it is started and stopped are obtained, and whether the motor operation fails is determined based on the relationship between the detection data and the preset detection data, where: When it is determined that a fault occurs during the operation of the motor, the fault score of the motor is determined based on the relationship between the detection data and the preset detection data; Determine the fault type of the motor based on the operating data and load data; Determine whether to issue an early warning to the motor according to the relationship between the fault score and the preset fault score, wherein when the fault score is lower than the preset fault score, determine the early warning method according to the score difference between the fault score and the preset fault score; when the fault score is higher than or equal to the preset fault score, determine the control method according to the fault type; When setting the preset test data, include: Acquire various historical detection data of the motor, wherein the detection data includes temperature detection data, vibration detection data and noise detection data of the motor; Obtaining the distance measurement between each historical detection data, and establishing a distance matrix for each historical detection data according to the distance measurement; Iteratively cluster the historical detection data according to the distance matrix, and obtain the feature vector of the normal detection data and the feature vector of the abnormal detection data according to the clustering results; Determine preset detection data according to the feature vector of normal detection data; Obtaining a distance value between a feature vector of normal detection data and a feature vector of abnormal detection data, and determining the distance value as a preset difference value; When determining whether a motor operation failure occurs based on the relationship between the detection data and the preset detection data, it includes: Obtain the data difference between the detection data and the preset detection data, and determine whether the motor operation fails based on the relationship between each data difference and the preset difference: When the data difference is lower than the preset difference, it is determined that the motor is running without any fault; When the data difference is equal to or higher than the preset difference amount, it is determined that the motor operation fails, and the fault score of the motor is determined based on the relationship between the data difference and the preset difference amount; When determining the fault score of the motor based on the relationship between the data difference and the preset difference amount, it includes: ; Among them, S is the fault score of the motor, wi is the weight coefficient of the i-th data item, α i is the nonlinear coefficient of the i-th data item, t is the temperature data item, v is the vibration data item, c is the noise data item, e -βi is the smoothing coefficient, △di is the data difference, and △Di is the preset difference amount.

2. The control method for a brushless DC reduction motor according to claim 1, characterized in that: When obtaining the running data and load data of the motor when it starts and stops, it includes: When obtaining the operating data, it includes obtaining the motor's starting current, starting voltage, starting torque, start-stop time interval, and start-stop speed; When acquiring load data, it includes acquiring the inertial load, friction load, torque load and current load carried by the motor.

3. The control method for a brushless DC reduction motor according to claim 2, characterized in that: Determine the motor fault type based on the operating data and load data, including: Acquire each historical operation data and each historical load data of the motor, and establish a motor fault diagnosis model based on a decision tree, a neural network, and the relationship between each historical operation data and each historical load data; The motor's operating data and load data are substituted into the motor fault diagnosis model to determine the motor's fault type, and the execution operation during control is determined based on the fault type.

4. The control method for a brushless DC reduction motor according to claim 3, characterized in that: When establishing a motor fault diagnosis model based on the relationship between the decision tree, neural network, historical operation data and historical load data, it includes: Establishing a motor state correlation equation based on historical operation data and corresponding motor states and historical load data and corresponding motor states; Based on the decision tree, the motor state associations are screened to determine the abnormal data between the motor state associations: ； Where A is the contribution value of the motor correlation, H(F) is the entropy of the motor state correlation in the data set F; Fj is the jth subset after the motor state correlation is divided, and K is the total number of data subsets; Get the median contribution value between the contribution values ​​of each motor correlation equation, and determine the abnormal data based on the relationship between the contribution value and the median contribution value: When the contribution value of the motor correlation equation is less than the median value of the contribution value, the historical operation data and the historical load data in the motor correlation equation are determined to be abnormal data; When the contribution value of the motor correlation formula is greater than or equal to the median value of the contribution value, it is determined that the historical operation data and the historical load data in the motor correlation formula are not abnormal data; According to the determined abnormal data, the motor fault type and characteristic vector are determined based on the neural network, and a motor fault diagnosis model is established according to the motor fault type and characteristic vector.

5. The control method for a brushless DC reduction motor according to claim 4, characterized in that: According to the determined abnormal data, the motor fault type and characteristic vector are determined based on the neural network, including: ； Among them, f(X) is abnormal data, is the probability of the Kth fault type, and is the weight coefficient, and and are all non-zero, σ is the activation function, xi is the i-th feature of the feature data, wij is the weight of the input feature xi to the j-th neuron in the hidden layer, bj is the bias of the j-th neuron, h is the number of motor fault types, n is the number of input features, and j is the neuron index of the hidden layer.

6. The control method for a brushless DC reduction motor according to claim 4, characterized in that: Depending on the fault type, the control operation to be performed is determined, including: Determine a first execution operation and a second execution operation according to the fault type, control the motor according to the first execution operation, and obtain real-time operation data and real-time load data of the motor during control; Obtaining the running data difference between the real-time running data and the historical running data of the adjacent time periods, and obtaining the load data difference between the real-time load data and the historical load data of the adjacent time periods; According to the relationship between the running data difference and the load data difference, determine whether to use the second execution operation to control the motor: When the operation data difference is lower than or equal to the configured preset operation data difference, and / or the load data difference is lower than or equal to the configured preset load data difference, it is determined not to select the second execution operation to control the motor; When the operation data difference is higher than the preset operation data difference, and / or the load data difference is higher than the preset load data difference, it is determined to select the second execution operation to control the motor.

7. A control system for a brushless DC reduction motor, which is adopted in a control method for a brushless DC reduction motor as claimed in any one of claims 1 to 6, characterized in that: include: An acquisition module is configured to acquire detection data, operation data and load data when the motor is started and stopped; An analysis module is electrically connected to the acquisition module, and the analysis module is configured to determine whether a fault occurs in the operation of the motor according to a relationship between the detection data and the preset detection data, wherein: when it is determined that a fault occurs in the operation of the motor, a fault score of the motor is determined according to the relationship between the detection data and the preset detection data; the analysis module is further configured to determine a fault type of the motor according to the operation data and the load data; The central control module is electrically connected to the analysis module. The central control module is configured to determine whether to issue an early warning for the motor based on the relationship between the fault score and a configured preset fault score. When the fault score is lower than the preset fault score, the score difference between the fault score and the preset fault score is determined to determine the early warning method; when the fault score is higher than or equal to the preset fault score, the control method is determined according to the fault type.

Images