Servo motor fault self-diagnosis method, system and readable storage medium

Abstract

The present application discloses a servo motor fault self-diagnosis method, system and readable storage medium, which relates to the field of motor fault diagnosis technology. The method includes obtaining the correlation between various sensor data, and the response probability and time of deviation from the baseline value in response to each fault event, generating a reference data queue that characterizes the fault-affected transmission path; obtaining various sensor data in real time and identifying abnormal data, marking the fault type and start time of the abnormal data; retrieving sensor data related to the abnormal data and analyzing whether it deviates from the baseline value and the deviation time, generating a feature data queue, and confirming the fault-affected transmission path based on the feature data queue; selecting one of the transmission nodes to implement the impact blocking operation, and judging the reliability of the fault source diagnosis result according to the data offset direction of each transmission node. The above scheme improves the fault diagnosis accuracy while blocking the propagation of the fault impact as needed, protecting the key components in the servo system, and maintaining the long-term stable operation of the servo motor.

Description

Technical Field

[0001] The present application relates to the technical field of motor fault diagnosis, and in particular to a servo motor fault self-diagnosis method, system and readable storage medium. Background Art

[0002] Servo motor systems are one of the key drive devices in modern industrial automation and robotics. Their operational stability directly impacts the performance of the entire equipment or production line. Therefore, servo motor system fault diagnosis and self-diagnosis systems play an important role in ensuring production efficiency, reducing equipment downtime, and improving equipment reliability.

[0003] To improve the accuracy of servo motor fault diagnosis, signal analysis-based fault diagnosis methods are currently the primary approach. The general process can be summarized as follows: various signals from the system during operation, such as current, voltage, speed, and temperature, are collected. Abnormal signals are then analyzed using signal processing techniques such as Fourier transforms and time-frequency analysis to ultimately diagnose the fault. For example, when a motor's bearings are worn, frequency domain analysis of the vibration signal reveals a significant increase in energy in the high-frequency bands of the signal, thereby confirming bearing wear. In recent years, with the increasing application of machine learning in fault diagnosis, the collection of large amounts of historical servo motor operating data and the use of training algorithms such as neural networks can also enable timely and accurate identification of different types of fault information.

[0004] In practice, the inventors found that by using the signal analysis method, machine learning method and other methods such as mathematical model prediction method as mentioned above to diagnose servo motor faults, although the accuracy of servo motor fault diagnosis can be greatly improved, when the servo motor fault characterization phenomenon and the fault location are relatively close, the accuracy of the diagnostic results of the above-mentioned diagnostic methods will be greatly reduced.

[0005] In actual applications, when a fault occurs at a certain location in a servo motor, it often triggers fault alarms at other locations. For example, when a motor bearing wears out, in addition to abnormal vibration signals, it also causes an abnormal increase in the internal temperature of the motor. This temperature increase will further lead to effects such as positioning accuracy drift, increased lubricant viscosity, and subsequently increased load current. Obviously, the various fault diagnosis methods mentioned above are only suitable for finding the type of fault and the location of occurrence, and do not have an effective effect in blocking the impact of the fault. Some derived fault effects can actually cause permanent damage to the performance of the internal components of the motor. For example, high temperature can cause irreversible changes in the magnetic flux of the permanent magnets inside the motor, which will ultimately pose a hidden danger to the long-term stable operation of the servo motor. Summary of the Invention

[0006] In order to further improve the accuracy of servo motor fault diagnosis results in practical applications, and at the same time solve the problem that the influence of faults cannot be effectively blocked during the servo motor fault diagnosis process, thereby burying hidden dangers for the long-term stable operation of the servo motor, the purpose of this application is to provide a servo motor fault self-diagnosis method, which uses various sensors to collect and obtain various sensor data during the operation of the servo motor and perform feature extraction, determine the correlation between the various sensor data, and then determine the influence transmission path of each type of fault, and finally obtain the fault type and location based on the above-mentioned influence transmission path by reverse analysis of each sensor data, improve the fault diagnosis accuracy, and effectively block the propagation of the fault influence as needed, protect the key components in the servo system, thereby maintaining the long-term stable operation of the servo motor; to implement the above-mentioned servo motor fault self-diagnosis method, the purpose of this application is to provide a servo motor fault self-diagnosis system, which realizes accurate diagnosis of motor faults by collecting sensor data inside and outside the servo system; finally, to facilitate the promotion and use of the above-mentioned servo motor fault self-diagnosis method, it is proposed to protect a computer-readable storage medium on which a program module for implementing the above-mentioned servo motor fault self-diagnosis method is loaded. The specific scheme is as follows:

[0007] A servo motor fault self-diagnosis method comprising:

[0008] Analyze and obtain the correlation between various sensor data, as well as the response probability and response time of various sensor data deviating from the baseline value in response to various fault events;

[0009] Various sensor data associated with the same fault event are used as transmission nodes, and are arranged in order according to the response probability and response time to generate a reference data queue for characterizing the transmission path affected by the fault;

[0010] Generate a verification model based on historical operation data combined with clustering algorithm to determine whether each sensor data is abnormal;

[0011] Acquire various sensor data of the servo motor in real time and extract data features, analyze and identify abnormal data according to the verification model, and mark the fault type and start time corresponding to the abnormal data;

[0012] Based on the correlation between various types of sensor data, the sensor data associated with the abnormal data is retrieved, and the associated sensor data is analyzed to determine whether the above-mentioned associated sensor data deviates from the baseline value within a set time period and the deviation time, and the characteristic data queue is generated according to the deviation time sequence;

[0013] Comparing the characteristic data queue with each stored reference data queue to match and confirm at least one transmission path affected by the fault;

[0014] Select a transmission path affected by the fault, select a transmission node in the transmission path affected by the fault as a blocking node and perform an impact blocking operation on it, and monitor the data deviation direction of each transmission node on the transmission path affected by the fault in real time:

[0015] If the data of each transmission node after the current blocking node deviates towards the reference value, the current blocking operation is maintained, and the fault type corresponding to the abnormal data is regarded as the fault source and an early warning is recorded;

[0016] If the data of the transmission nodes after the current blocking node do not deviate towards the reference value, then count whether the number of transmission nodes after the current blocking node is less than the set value:

[0017] If the number of transmission nodes after the current blocking node is not less than the set value, a transmission node in the transmission path affected by the fault is selected as the blocking node and the blocking operation is performed on it, and the data deviation direction of each transmission node on the transmission path affected by the fault is monitored;

[0018] If the number of transmission nodes after the currently blocked node is less than the set value, determine whether there are any remaining fault-affected transmission paths to be selected:

[0019] If there are currently remaining transmission paths affected by the fault available for selection, a new transmission path affected by the fault is selected, and a transmission node in the transmission path affected by the fault is selected as a blocking node and an impact blocking operation is performed on it;

[0020] If there are no remaining fault-affected transmission paths available for selection, the identified abnormal data will be marked as interference data and a warning will be recorded.

[0021] Through this technical solution, various sensor data can be monitored in real time during servo motor operation. Abnormal data is first identified using an existing verification model, and then the accuracy of the fault source corresponding to the abnormal data is determined based on the transmission path of each type of fault. During the secondary confirmation of the fault source, the fault influence is blocked at specific transmission nodes, improving the accuracy of fault diagnosis while effectively blocking the propagation of the fault influence as needed, protecting key components in the servo system and maintaining the long-term stable operation of the servo motor.

[0022] Furthermore, selecting a transmission node in the transmission path affected by the fault as a blocking node and performing an impact blocking operation on the node includes:

[0023] Obtaining the blocking coefficient of each transmission node, the probability loss coefficient of the fault affecting the transmission between two adjacent transmission nodes, and the importance level coefficient of each transmission node;

[0024] Extract the blocking coefficient of each transmission node in the transmission path affected by the currently matched fault and the probability loss coefficient between two adjacent transmission nodes, and calculate and select one of the target transmission nodes for the impact blocking operation;

[0025] The blocking coefficient is defined as the probability that the fault impact will spread to the next transmission node after the impact blocking operation is implemented on the transmission node;

[0026] The probability loss coefficient is defined as: the probability that the failure impact will successfully spread from the current transmission node to the next transmission node under natural conditions;

[0027] The importance level coefficient is defined as: the correlation coefficient between the offset value of the sensor data at the transmission node and the positioning accuracy of the servo motor;

[0028] The impact blocking operation includes: maintaining the value of any one or a combination of temperature, current, voltage, and vibration amplitude to fluctuate within a reference range;

[0029] The calculation and selection of one of the target transmission nodes for impact blocking operations include:

[0030] Configuring a blocking expectation parameter L for each transmission node on the transmission path affected by the fault;

[0031] Sort by the size of each blocking expectation parameter L, and select the transmission node with the highest ranking to perform the impact blocking operation;

[0032] Among them, the blocking expectation parameter L=w1·P1+w2·P2+k; P1 is the blocking coefficient, P2 is the probability loss coefficient, w1 and w2 are the weights corresponding to the blocking coefficient and the probability loss coefficient respectively, and k is the correction coefficient;

[0033] The above-mentioned constraint conditions for the blocking expectation parameter L are: In the above constraints, M is the importance level coefficient of each transmission node located after the blocking node on the transmission path affected by the fault, n is the number of transmission nodes located after the blocking node on the transmission path affected by the fault, and X is a set threshold.

[0034] Through the above technical solution, a transmission node can be accurately selected to implement the impact blocking operation, which not only ensures the effectiveness of the blocking but also ensures that the impact of the fault is limited to the set range, providing an accurate basis for subsequent verification of the fault source. At the same time, it can protect the key components in the system from the impact of the fault and maintain the long-term and stable operation of the servo system.

[0035] Furthermore, the correlation between the various types of sensor data includes a time series relationship and a numerical relationship between the various types of sensor data after normalization;

[0036] The fault self-diagnosis method further includes:

[0037] Configure at least one decision threshold and its corresponding control instruction for each transmission node;

[0038] Obtain the starting node of the transmission path affected by the current fault and its corresponding starting time and value;

[0039] Based on the timing and numerical relationships between the transmission nodes, the expected numerical value of each transmission node on the transmission path affected by the fault at each time is calculated and generated;

[0040] Compare the expected value corresponding to each transmission node with the determination threshold at the transmission node, and output the set control instruction according to the comparison result;

[0041] The control instructions include maintaining the current action, outputting an alarm message, executing a set operation action, or shutting down the servo system.

[0042] Through the above technical solution, when a fault occurs at a certain point in the servo motor, the system can evaluate in advance the impact of the fault on various sensor data, that is, various operating parameters, accurately output alarm information or shut down the servo system in advance to ensure that key components in the servo system are not affected by the fault.

[0043] Furthermore, in the fault self-diagnosis method, matching and confirming at least one fault-affected transmission path further comprises:

[0044] Based on the historical operation data of the servo motor and the correlation analysis method, the correlation coefficient between various sensor data is obtained;

[0045] Selecting at least two pieces of sensor data with a correlation coefficient greater than a set value and located in the same fault-affected transmission path, obtaining a temporal relationship and a numerical relationship between the at least two pieces of sensor data, combining them to generate a feature array, and storing the array in association with the corresponding fault-affected transmission path;

[0046] Based on the correlation between various types of sensor data, sensor data associated with the abnormal data is retrieved, the associated sensor data is compared with the feature array, and at least one fault-affected transmission path is matched and confirmed.

[0047] Through the above technical solution, the transmission path affected by the fault can be quickly determined, and the fault source can be quickly reconfirmed, shortening the time for fault self-diagnosis.

[0048] Furthermore, the fault self-diagnosis method further includes:

[0049] According to the data type of the sensor data, the sensor data is divided into electrical data and mechanical data;

[0050] Establishing an electrical parameter correction model and a mechanical parameter correction model for characterizing the correlation between changes in external parameters and fluctuations in the electrical and mechanical data;

[0051] Acquire the external parameters of the servo system in real time during operation, and calculate and generate the correction coefficient k corresponding to the expected blocking parameter L of each transmission node according to the correction model;

[0052] The external parameters associated with the electrical data include the ambient temperature of the servo system, the electromagnetic radiation intensity, and the electrical parameters of the servo system power supply connection;

[0053] The external parameters associated with the mechanical data include vibration amplitude.

[0054] Through the above technical solution, the expected blocking parameters can be adjusted in time according to the changes in external parameters, so that the transmission affected by the fault can be effectively blocked, and then the accuracy of the matching fault-affected transmission path can be accurately determined. Finally, the fault source is reconfirmed to improve the accuracy of fault diagnosis. It can also more accurately block the further spread of the fault impact and protect the key components in the servo system from being affected by the fault.

[0055] Furthermore, the fault self-diagnosis method further includes:

[0056] Establish the correlation between the probability of abnormality of various sensor data and the frequency and range of abnormal data of the set category;

[0057] Calculate the frequency of occurrence of each category of abnormal data and the average value of the abnormal range in the historical operation data, and generate the abnormal probability of each type of sensor data based on the above correlation relationship;

[0058] The probability loss coefficient is adjusted based on the above abnormal probability.

[0059] Through the above technical solution, the potential impact of various historical faults on the system is incorporated into the assessment of the failure-affected transmission success rate, which helps to improve the accuracy of the confirmation of the failure-affected transmission path and provide a more accurate basis for quickly and accurately finding the source of the fault.

[0060] A servo motor fault self-diagnosis system, comprising:

[0061] a database unit configured to store historical operating data of the servo motor, correlation data between various sensor data, real-time operating data, and a reference data queue corresponding to each fault event;

[0062] An abnormal data identification unit includes a verification model generated based on historical operating data combined with a clustering algorithm, which is used to obtain various sensor data of the servo motor in real time and extract data features, analyze and identify abnormal data based on the verification model, and mark the fault type and start time corresponding to the abnormal data;

[0063] a characteristic data queue generating unit configured to be data-connected to the abnormal data identifying unit, retrieve sensor data associated with the abnormal data based on the correlation between various types of sensor data, analyze and determine whether the associated sensor data deviates from the reference value within a set time period and the time of deviation, and generate a characteristic data queue by arranging the data in order of the deviation time;

[0064] a fault-affected transmission path confirmation unit configured to compare the characteristic data queue with each stored reference data queue to match and confirm at least one fault-affected transmission path;

[0065] An impact blocking verification unit is configured to select a transmission path affected by the fault and select one of the transmission nodes to perform an impact blocking operation, monitor the data deviation direction of each transmission node on the transmission path affected by the fault in real time, determine the fault source based on the data deviation direction combined with the set steps, and record an early warning;

[0066] The early warning output unit is configured to be data-connected with the impact blocking verification unit to receive and output early warning information.

[0067] Furthermore, the impact blocking verification unit has a built-in blocking node selection module, including:

[0068] A reference coefficient acquisition submodule configured to acquire and temporarily store a blocking coefficient of each transmission node, a probability loss coefficient of a fault affecting transmission between two adjacent transmission nodes, and an importance level coefficient of each transmission node;

[0069] A reference coefficient extraction submodule is configured to extract the blocking coefficient of each transmission node in the fault-affected transmission path currently matched and confirmed, and the probability loss coefficient between two adjacent transmission nodes;

[0070] A node selection submodule is configured to configure a blocking expectation parameter L for each transmission node on the transmission path affected by the fault, sort the nodes according to the magnitude of the blocking expectation parameters L, and select the transmission node with the highest ranking to perform the impact blocking operation;

[0071] The blocking expected parameter L is obtained by the blocking expected parameter L calculation formula disclosed in the servo motor fault self-diagnosis method as described above.

[0072] Furthermore, the system further includes an instruction presetting unit, including:

[0073] A first storage module is configured to obtain and store a determination threshold corresponding to each transmission node and a control instruction corresponding to the determination threshold;

[0074] A second storage module is configured to acquire and store the temporal relationship between the various types of sensor data and the normalized numerical relationship between the various types of sensor data;

[0075] A starting node acquisition module configured to acquire the starting node of the transmission path affected by the current fault and its corresponding starting time and numerical value;

[0076] An expected numerical value generation module is configured to calculate and generate an expected numerical value of each transmission node on the transmission path affected by the fault at each moment based on the timing relationship and numerical relationship between the transmission nodes;

[0077] An instruction confirmation output module is configured to compare the expected numerical value corresponding to each transmission node with the determination threshold value at the transmission node, and output a set control instruction according to the comparison result;

[0078] The control instructions include maintaining the current action, outputting an alarm message, executing a set operation action, or shutting down the servo system.

[0079] A computer-readable storage medium is loaded with a program module for implementing the servo motor fault self-diagnosis method as described above.

[0080] The above technical solution is helpful to promote the use of the servo motor self-diagnosis method of the present application.

[0081] In summary, this application includes at least one of the following beneficial technical effects:

[0082] (1) Based on historical operating data, a verification model for identifying abnormal data is obtained to form a preliminary judgment of the fault source. Then, based on the correlation between historical operating data and various sensor data, the fault-affecting transmission path is obtained. By comparing various sensor data collected in real time with the fault-affecting transmission path, a secondary verification of the fault source is achieved, thereby improving the accuracy of fault diagnosis;

[0083] (2) By introducing and implementing the impact blocking operation during the fault self-diagnosis process, the spread of the fault impact can be accurately blocked, the important components of the servo system are protected from the impact of the fault, and the long-term stable operation of the servo motor is maintained;

[0084] (3) By configuring the judgment threshold and control instructions step by step along the fault impact transmission path, it is possible to output alarm information or perform corresponding protection actions in a timely manner before the fault impact causes further damage to the servo system, thereby preventing the fault at a certain point from causing damage to other components. BRIEF DESCRIPTION OF THE DRAWINGS

[0085] Figure 1 Schematic diagram of the method for confirming the transmission path affected by the fault;

[0086] Figure 2 Schematic diagram of the method for confirming the source of the fault;

[0087] Figure 3 It is a schematic diagram of the transmission nodes affected by the fault arranged along the time axis;

[0088] Figure 4 This is a schematic diagram of the functional modules of the fault self-diagnosis system of this application.

[0089] Figure numerals: 1. Database unit; 2. Abnormal data identification unit; 3. Feature data queue generation unit; 4. Fault impact transmission path confirmation unit; 5. Impact blocking verification unit; 6. Early warning output unit; 7. Reference coefficient acquisition submodule; 8. Reference coefficient extraction submodule; 9. Node selection submodule. DETAILED DESCRIPTION

[0090] Embodiments of the present application are described in detail below, examples of which are illustrated in the accompanying drawings.

[0091] Throughout this specification, reference to the terms "certain embodiments," "one embodiment," "some embodiments," "illustrative embodiments," "examples," "specific examples," or "some examples" means that the specific features, structures, materials, or characteristics described in conjunction with the embodiments or examples are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.

[0092] The present application embodiment first discloses a servo motor fault self-diagnosis method, such as Figure 1 As shown, it mainly includes the following steps:

[0093] S100 , analyzing and obtaining correlations between various types of sensor data, and response probabilities and response times of various types of sensor data deviating from baseline values ​​in response to various types of fault events.

[0094] The above correlation analysis and acquisition methods specifically include theoretical analysis methods, and based on the historical operation data of the servo motor, the correlation between the various sensor data is obtained through time-frequency domain analysis and machine learning methods. The theoretical analysis method specifically aims to obtain the physical parameters associated with various sensor data, such as by obtaining the input voltage U in and current I inThe size, as well as the output torque τ and angular velocity ω of the servo motor, can be used to know the heat generated by the servo system during the above-mentioned set action. Finally, the temperature change at the set position in the servo motor can be calculated. Without considering the heat loss, its theoretical calculation model can be summarized as follows:

[0095]

[0096] In the theoretical calculation model above, τ represents the servo motor's torque, ω represents its angular velocity, m represents the mass of its heat-generating components, such as the core and coil, and c represents their specific heat capacity. In practical applications, the product of the mass and specific heat capacity of the heat-generating components of a given servo motor is typically a fixed value. Therefore, the theoretical calculation model can be used to determine the relationship between the temperature T at a specific location in the servo motor and related physical parameters.

[0097] The above-mentioned time-frequency domain analysis methods include wavelet coherence analysis, Fourier transform method, etc. For example, the distribution area of ​​the servo motor vibration energy is obtained through frequency domain analysis method; machine learning methods mainly include principal component analysis (PCA), canonical correlation analysis (CCA) and recursive neural networks. The principal component analysis method is mainly used to discover and obtain the relationship between sensor data that are not obviously correlated in theory, such as the relationship between the power supply current and vibration amplitude of the servo motor. The canonical correlation analysis method is used to discover and obtain the linear relationship between two groups of variables whose theoretical correlation is relatively complex and difficult to generate a theoretical calculation model, such as the relationship between the humidity of the servo motor working environment and the local temperature of the motor; recursive neural networks are used to learn complex nonlinear relationships between time series.

[0098] It can be seen from the above disclosure that the correlation between various types of sensor data in the embodiments of the present application includes the temporal relationship between the sensor data and the numerical relationship between the various types of sensor data after normalization.

[0099] In the embodiments of the present application, the response probability and response time of various sensor data deviating from the baseline value in response to various fault events are primarily obtained based on an analysis of the servo motor's historical operating data. In practical applications, the servo motor's various sensor data are stored in a specific data storage device. Various fault events are characterized as abnormalities in specific sensor data during data storage. For example, abnormalities in the motor's vibration amplitude or high-frequency energy of the vibration signal can be used to determine the fault type as bearing or motor shaft wear. The baseline values ​​of various sensor data are calculated by averaging historical data, such as the average vibration velocity (mm / s) and the average input current (A). When analyzing historical operating data, the acquisition time of the abnormal data is first obtained, and then the acquisition time of other abnormal data is detected to ultimately generate the response time. In practical applications, the response time can be the average of the response times after multiple fault events. The response probability is the probability that each type of sensor data responds to the same fault event in multiple fault events. It is calculated by dividing the number of responses by the number of fault events.

[0100] S200 , various types of sensor data associated with the same fault event are used as transmission nodes, and are arranged in order according to the response probability and response time to generate a reference data queue for characterizing the transmission path affected by the fault.

[0101] like Figure 3 As shown in , since the same fault event can cause abnormalities in multiple categories of sensor data, and the diffusion direction of the fault impact will change due to changes in external parameters, the fault impact transmission path corresponding to the same fault event will have multiple branches. Figure 3 As shown in the figure, each circle represents a different transmission node, and each transmission node corresponds to a corresponding blocking coefficient P1. The arrows between the transmission nodes indicate the direction in which the fault impact spreads downward from the current transmission node, and each arrow corresponds to a corresponding probability loss coefficient P2. The coordinates of each transmission node on the time axis are different. The transmission path affected by the fault can be generated by the sequence of the above time coordinates and the correlation between the transmission nodes.

[0102] In the implementation manner of the present application, corresponding attribute data is associated and stored for each transmission node, specifically including: blocking coefficient, probability loss coefficient and importance level coefficient.

[0103] The blocking coefficient is defined as: the probability that the fault impact will spread to the next transmission node after the impact blocking operation is implemented on the transmission node. For example: due to bearing wear, the temperature gradually rises, and the gradual temperature rise will cause the lubricant to heat up and then cause the lubricant viscosity to change. In order to block one of the fault-affected transmission paths corresponding to the above-mentioned bearing wear, the temperature rise trend can be suppressed by increasing the cooling power, thereby avoiding the occurrence of subsequent changes in the lubricant viscosity and the like. For different categories of sensor data, the blocking effect that can be achieved by implementing the impact blocking operation is different. For changes in voltage or current, it can be achieved by inputting a reverse suppression signal, while sensor data such as vibration amplitude is difficult to change its value by implementing relevant impact blocking operations. For this reason, each transmission node has a corresponding blocking coefficient.

[0104] In the embodiment of the present application, the aforementioned impact blocking operation includes but is not limited to: maintaining the value of any one or a combination of multiple values ​​of temperature, current, voltage, and vibration amplitude to fluctuate within a reference range.

[0105] The probability loss coefficient is defined as the probability that a fault's impact will successfully propagate from the current transmission node to the next transmission node under natural conditions. As previously mentioned, due to interference from external parameters, the impact of a single fault will not propagate along a specific transmission path. The probability loss coefficient is calculated by statistically analyzing historical operating data. In practice, this historical operating data can be the historical operating data of a single servo motor or multiple servo motors of the same type.

[0106] The importance level coefficient is defined as: the correlation coefficient between the offset value of the sensor data at the transmission node and the positioning accuracy of the servo motor, that is, whether the positioning accuracy of the servo motor will change when the sensor data at a certain transmission node changes. The larger the importance level coefficient, the closer the relationship between the servo motor positioning accuracy and the transmission node.

[0107] S300: Generate a verification model based on historical operation data and a clustering algorithm to determine whether each sensor data is abnormal.

[0108] S400, acquiring various sensor data of the servo motor in real time and extracting data features, analyzing and identifying abnormal data according to the verification model, and marking the fault type and start time corresponding to the abnormal data.

[0109] The above-mentioned verification model is a mathematical model. Based on historical operating data combined with a clustering algorithm, it can quickly find the association between each related sensor data and can quickly identify abnormal data. The execution process of the clustering algorithm includes data acquisition, data preprocessing, and selecting a specific clustering algorithm for implementation verification. The sensor data collected in the above-mentioned data acquisition process include but are not limited to voltage, current, speed, temperature, vibration amplitude, sound spectrum characteristics, etc. For ease of processing, each sensor data is normalized and feature extracted. The clustering algorithm preferably adopts the K-means clustering algorithm, which is simple and efficient, and can quickly and real-time discover abnormal data.

[0110] S500, based on the correlation between various types of sensor data, retrieve the sensor data associated with the abnormal data, analyze and determine whether the above-mentioned associated sensor data deviates from the baseline value and the deviation time within a set time period, and generate a feature data queue according to the deviation time sequence.

[0111] S600: Compare the characteristic data queue with each stored reference data queue to match and confirm at least one transmission path affected by the fault.

[0112] In the above step S500 , the feature data queue is composed of a plurality of discrete data with a specific time sequence or numerical value relationship. Compared with the stored reference data queue, there is usually a lack of data in the number of data.

[0113] In step S600, the comparison method includes: first, setting a fluctuation time interval for the data in each reference data queue. For example, if transmission node A is defined as the source node, its data abnormal time is t1, and the time of the next transmission node B is set to t1+t2+Δt, where t2 is the time required for the fault impact to spread from transmission node A to transmission node B, and Δt is the fluctuation time interval corresponding to transmission node B. The data in the generated feature data queue is compared with the data in the reference data queue, and the fault-affected transmission path is found through the matching relationship between the time of each transmission node. After the fault-affected transmission path with a matching time node is found, the numerical relationship between the sensor data corresponding to each transmission node is used to further match and confirm the fault-affected transmission path, thereby ensuring the accuracy of the matching confirmation result.

[0114] In an optimized implementation, matching and confirming at least one transmission path affected by the fault further includes:

[0115] S610, obtaining correlation coefficients between various sensor data based on historical operation data of the servo motor and a correlation analysis method;

[0116] S620, selecting at least two pieces of sensor data having a correlation coefficient greater than a set value and located in the same fault-affected transmission path, obtaining a temporal relationship and a numerical relationship between the at least two pieces of sensor data, combining them to generate a feature array, and storing the array in association with the corresponding fault-affected transmission path;

[0117] S630 , based on the correlation between various types of sensor data, retrieve sensor data associated with the abnormal data, compare the associated sensor data with the feature array, and match and confirm at least one transmission path affected by the fault.

[0118] The above technical solution can quickly determine the transmission path affected by the fault, and then quickly conduct a second confirmation of the fault source, shortening the time for fault self-diagnosis.

[0119] From steps S100 to S600, two fault diagnosis operations are included. The first is to identify and mark abnormal data in step S400, and the second is to confirm the transmission path affected by the fault in step S600. Step S400 directly uses the verification model to diagnose the fault, and step S600 confirms the fault source by confirming the transmission path affected by the fault.

[0120] In order to further determine the accuracy of the transmission path affected by the fault, combined with Figure 2 As shown, the fault self-diagnosis method described in the embodiment of the present application also includes:

[0121] S700, selecting a transmission path affected by the fault;

[0122] S710: Select a transmission node in the transmission path affected by the fault as a blocking node and perform an impact blocking operation on it, and monitor the data deviation direction of each transmission node on the transmission path affected by the fault in real time:

[0123] S720: If the data of each transmission node located after the blocking node deviates toward the reference value, maintain the current impact blocking operation, and simultaneously identify the fault type corresponding to the abnormal data as the fault source and record an early warning;

[0124] S730: If the data of each transmission node after the blocking node does not deviate toward the reference value, count whether the number of transmission nodes after the current blocking node is less than the set value:

[0125] S731, if the number of transmission nodes after the current blocking node is not less than the set value, repeat steps S710-S731;

[0126] S732: If the number of transmission nodes after the currently blocked node is less than the set value, determine whether there are any remaining faults affecting the transmission path to be selected.

[0127] S7321, if there are currently remaining transmission paths affected by the fault available for selection, repeat step S700 to reselect a transmission path affected by the fault, and then repeat steps S710-S731;

[0128] S7322: If there is no remaining fault-affected transmission path available for selection, the abnormal data identified in step S400 is marked as interference data and a warning is recorded.

[0129] In the implementation manner of the present application, the servo system fault warning method includes on-site sound and light warning, such as using LED indicator lights in red, orange, yellow and green colors to indicate the current operating status of the servo system, and also includes sending fault information to the handheld communication terminal of a specific person, or sending warning information to the host in the monitoring room through a communication cable to remind relevant maintenance personnel to check the system operating status.

[0130] In the above step S710, a transmission node in the transmission path affected by the fault is selected as a blocking node and an impact blocking operation is performed on it, specifically including:

[0131] S711, obtaining a blocking coefficient of each transmission node, a probability loss coefficient of a failure affecting transmission between two adjacent transmission nodes, and an importance level coefficient of each transmission node;

[0132] S712 , extracting the blocking coefficient of each transmission node in the currently matched and confirmed fault-affected transmission path and the probability loss coefficient between two adjacent transmission nodes, and calculating and selecting one of the target transmission nodes for impact blocking operation.

[0133] Furthermore, step S712 specifically includes:

[0134] S7121, configuring a blocking expectation parameter L for each transmission node on the transmission path affected by the fault;

[0135] S7122: Sort the nodes according to the size of the expected blocking parameters L, and select the transmission node with the highest ranking to perform the impact blocking operation.

[0136] The blocking expectation parameter L = w1·P1+w2·P2+k; P1 is the blocking coefficient, P2 is the probability loss coefficient, w1 and w2 are the weights corresponding to the blocking coefficient and the probability loss coefficient, respectively, and k is the correction coefficient used to correct the blocking expectation parameter L corresponding to each transmission node according to external parameters.

[0137] The above-mentioned constraint conditions for the blocking expectation parameter L are: In the above constraints, M is the importance level coefficient of each transmission node located after the blocking node on the transmission path affected by the fault, n is the number of transmission nodes located after the blocking node on the transmission path affected by the fault, and X is a set threshold. From the above constraints, it can be seen that whether a transmission node is selected for implementing the impact blocking operation is not only related to its corresponding blocking expectation parameter L, that is, the success rate of implementing the impact blocking, but also to its position on the transmission path affected by the fault. Obviously, the later the position, the smaller the f(M) value, that is, the smaller the importance level coefficient of the transmission node, and the less effective the impact blocking operation will be.

[0138] Based on the above solution, a transmission node can be accurately selected to implement the impact blocking operation, which not only ensures the effectiveness of the blocking but also ensures that the fault impact is limited to the set range, providing an accurate basis for subsequent verification of the fault source. At the same time, it can protect the key components in the system from the impact of the fault and maintain the long-term stable operation of the servo system.

[0139] In the embodiment of the present application, in order to achieve fault diagnosis and real-time processing, the fault self-diagnosis method further includes:

[0140] A100: Configure at least one determination threshold and its corresponding control instruction for each transmission node. For example, when the shaft temperature exceeds the determination threshold, the cooling power of the associated refrigeration device is increased.

[0141] A200, obtain the starting node of the transmission path affected by the current fault and its corresponding starting time and value;

[0142] A300 calculates and generates the expected value of each transmission node on the transmission path affected by the fault at each time based on the timing relationship and numerical relationship between the transmission nodes;

[0143] A400 compares the expected numerical value corresponding to each transmission node with the determination threshold at that transmission node and outputs a set control instruction based on the comparison result. In actual applications, each transmission node may correspond to multiple determination thresholds, forming multiple determination intervals. The control instruction is output based on the determination interval in which the expected numerical value falls.

[0144] The control instructions described in each of the above steps include, but are not limited to, maintaining the current action, outputting an alarm, executing a set operation, or shutting down the servo system. Therefore, if a fault occurs at a certain point in the servo motor, the system can proactively assess the impact of the fault on various sensor data, i.e., various operating parameters, and accurately output an alarm or prematurely shut down the servo system, ensuring that key components in the servo system are not affected by the fault.

[0145] In practical applications, the operation of the servo system will be affected by external parameters such as ambient temperature, electromagnetic radiation intensity, and vibration amplitude. Under different external parameter conditions, the probability of various faults occurring and the transmission path affected by the faults will change. In order to adapt to the changes in external parameters and ensure the accuracy of fault diagnosis, the fault self-diagnosis method described in the embodiment of the present application also includes:

[0146] B100, classifies the sensor data into electrical data and mechanical data according to the data type of the sensor data.

[0147] B200, establishes an electrical parameter correction model and a mechanical parameter correction model to characterize the correlation between external parameter changes and the fluctuations of the above-mentioned electrical data and mechanical data. For example, the resistance of the servo motor winding is related to the temperature, and the resistance value of the winding can be fine-tuned by detecting the temperature parameters.

[0148] B300, obtaining external parameters of the servo system in real time during operation, and calculating and generating a correction coefficient k corresponding to the expected blocking parameter L of each transmission node according to the correction model.

[0149] In the embodiment of the present application, the external parameters associated with the electrical data include the ambient temperature of the servo system, the electromagnetic radiation intensity, and the electrical parameters of the servo system power supply terminal connection, and the external parameters associated with the mechanical data include the vibration amplitude.

[0150] The above technical solution can timely adjust the blocking expectation parameters according to the changes in external parameters, so that the transmission affected by the fault can be effectively blocked, and then accurately determine whether the matching fault-affected transmission path is correct.

[0151] Furthermore, the fault self-diagnosis method further includes:

[0152] C100, establishes the correlation between the probability of abnormal occurrence of various types of sensor data and the frequency and range of abnormal data of the set category;

[0153] C200, counts the frequency of occurrence of each category of abnormal data and the average value of the abnormal range in the historical operation data, and generates the abnormal probability of each type of sensor data based on the above correlation relationship;

[0154] C300: Adjust the probability loss coefficient based on the abnormal probability.

[0155] During servo system operation, abnormal data can occur for a variety of reasons, including the fault itself and interference. Abnormal data not caused by faults often has a potential correlation between its frequency and the fault itself. Step C100 obtains this potential correlation from historical operating data. This technical solution incorporates the potential impact of various historical faults on the system into the assessment of the success rate of fault-impact transmission, helping to improve the accuracy of fault-impact transmission path identification and providing a more precise basis for quickly and accurately locating the fault source.

[0156] A servo motor fault self-diagnosis system, such as Figure 4 As shown, it mainly includes: a database unit 1, an abnormal data identification unit 2, a feature data queue generation unit 3, a fault impact transmission path confirmation unit 4, an impact blocking verification unit 5 and an early warning output unit 6.

[0157] Database unit 1 is configured to store historical servo motor operating data, correlation data between various sensor data types, real-time operating data, and reference data queues corresponding to various fault events. In a specific application, database unit 1 is configured as local memory and cloud storage. The local memory is used to collect and store the real-time operating data of the servo system, and the cloud storage is mainly used to store historical operating data of one or more servo motors and related data processing algorithm modules, such as clustering algorithm models and neural network models. The servo system's host computer is connected to the cloud storage data via a communication cable or wireless communication module.

[0158] Abnormal data identification unit 2 includes a mathematical model generated based on historical operating data and a clustering algorithm. This model is used to acquire various sensor data from the servo motor in real time and extract data features. The model then analyzes and identifies abnormal data based on the model, marking the corresponding fault type and onset time. Feature data queue generation unit 3 is configured to connect to abnormal data identification unit 2, acquire the abnormal data, and then retrieve sensor data associated with the abnormal data based on correlations between the various sensor data types. This model analyzes and determines whether the associated sensor data deviates from a baseline value within a set time period and the time of deviation. The feature data queue is then generated by arranging the data in chronological order based on the deviation time.

[0159] The fault-affected transmission path confirmation unit 4 is configured to be data-connected with the feature data queue generation unit 3, receive the generated feature data queue, and then compare the feature data queue with the stored reference data queues to match and confirm at least one fault-affected transmission path. In actual applications, since different fault impacts have similarities in external representations, multiple fault-affected transmission paths can usually be matched.

[0160] The impact blocking verification unit 5 is configured to be data-connected with the fault-affected transmission path confirmation unit 4, and is used to receive and select one of the fault-affected transmission paths, and then select one of the transmission nodes to implement the impact blocking operation, monitor the data offset direction of each transmission node on the fault-affected transmission path in real time, determine the fault source based on the data offset direction combined with the set steps, and record the early warning.

[0161] The warning output unit 6 is configured to be data-connected to the impact blocking verification unit 5 to receive and output warning information. The warning output unit 6 includes on-site warning devices, such as LED sound and light alarms, as well as displays and speakers installed in the monitoring center, and mobile communication terminals installed on maintenance personnel to notify relevant maintenance personnel of system failures.

[0162] In order to improve the effect of the blocking operation, the blocking verification unit 5 is equipped with a blocking node selection module. Figure 4 As shown, it specifically includes a reference coefficient acquisition submodule 7, a reference coefficient extraction submodule 8 and a node selection submodule 9.

[0163] The reference coefficient acquisition submodule 7 is configured to be data-connected to the database unit 1 and is used to acquire and temporarily store the blocking coefficient of each transmission node, the probability loss coefficient of the failure affecting transmission between two adjacent transmission nodes, and the importance level coefficient of each transmission node. The definitions of each reference coefficient have been explained above and will not be repeated here.

[0164] The reference coefficient extraction submodule 8 is configured to be data-connected to the fault-affected transmission path confirmation unit 4 and is configured to receive and extract the blocking coefficient of each transmission node in the currently matched and confirmed fault-affected transmission path, as well as the probability loss coefficient between two adjacent transmission nodes. The node selection submodule 9 is configured to assign a blocking expectation parameter L to each transmission node on the fault-affected transmission path, sort the nodes based on their respective blocking expectation parameters L, and select the transmission node with the highest ranking to perform the impact blocking operation. The blocking expectation parameter L is obtained using the blocking expectation parameter L calculation formula disclosed in step S7122 as described above.

[0165] In order to deal with the possible impact of faults and reduce the damage caused by faults to important components of the servo system, in an embodiment of the present application, the servo motor fault self-diagnosis system also includes an instruction preset unit, specifically including a first storage module, a second storage module, a starting node acquisition module, an expected numerical value generation module and an instruction confirmation output module.

[0166] Specifically, the first storage module is configured to acquire and store the determination threshold corresponding to each transmission node, as well as the control instructions corresponding to the determination threshold. The second storage module is configured to acquire and store the temporal relationships between the various types of sensor data, as well as the normalized numerical relationships between the various types of sensor data. In practice, preferably, the first and second storage modules are configured to be data-connected to the database unit 1 to directly access the stored relevant data.

[0167] The starting node acquisition module is configured to obtain the starting node of the transmission path affected by the current fault and its corresponding starting time and numerical value. The expected numerical value generation module is configured to calculate and generate the expected numerical value of each transmission node on the transmission path affected by the fault at each moment based on the timing relationship and numerical relationship between each transmission node. The instruction confirmation output module is configured to be data-connected to the expected numerical value generation module, and is used to compare the expected numerical value corresponding to each transmission node with the judgment threshold at the transmission node, and output the set control instruction based on the comparison result. In the embodiment of the present application, the above-mentioned control instruction includes but is not limited to maintaining the current action, outputting alarm information, executing the set operation action or shutting down the servo system. The above-mentioned control instructions can be executed separately or simultaneously.

[0168] In order to facilitate the promotion of the servo motor fault self-diagnosis method described in this application, a computer-readable storage medium is also disclosed in the embodiment of this application, which is loaded with a program module for implementing the servo motor fault self-diagnosis method as described above. The above-mentioned computer-readable storage medium includes but is not limited to storage devices such as disk storage, CD-ROM, and optical storage.

[0169] Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and cannot be understood as limitations on the present application. Ordinary technicians in this field can change, modify, replace and modify the above embodiments within the scope of the present application.

Claims

1. A servo motor fault self-diagnosis method, characterized in that: include: Analyze and obtain the correlation between various sensor data, as well as the response probability and response time of various sensor data deviating from the baseline value in response to various fault events; Various sensor data associated with the same fault event are used as transmission nodes, and are arranged in order according to the response probability and response time to generate a reference data queue for characterizing the transmission path affected by the fault; Acquire various sensor data of the servo motor in real time and extract data features. Identify abnormal data based on the verification model analysis and mark the fault type and start time corresponding to the abnormal data. Based on the correlation between various types of sensor data, the sensor data associated with the abnormal data is retrieved, and the associated sensor data is analyzed to determine whether the above-mentioned associated sensor data deviates from the baseline value within a set time period and the deviation time, and the characteristic data queue is generated according to the deviation time sequence; Comparing the characteristic data queue with each stored reference data queue to match and confirm at least one transmission path affected by the fault; Select a transmission path affected by the fault, choose one of the transmission nodes as a blocking node and perform the blocking operation on it, and monitor the data deviation direction of each transmission node on the transmission path affected by the fault in real time: If the data of the transmission nodes after the current blocking node deviates towards the reference value, the current blocking operation is maintained, and the fault type corresponding to the abnormal data is regarded as the fault source and an early warning is recorded; If the data of the transmission nodes after the current blocking node do not deviate towards the reference value, then count whether the number of transmission nodes after the current blocking node is less than the set value: If the number of transmission nodes after the current blocking node is not less than the set value, a transmission node in the transmission path affected by the fault is selected as the blocking node and the blocking operation is performed on it, and the data deviation direction of each transmission node on the transmission path affected by the fault is monitored; If the number of transmission nodes after the currently blocked node is less than the set value, determine whether there are any remaining fault-affected transmission paths to be selected: If there are currently remaining transmission paths affected by the fault available for selection, a new transmission path affected by the fault is selected, and a transmission node in the transmission path affected by the fault is selected as a blocking node and an impact blocking operation is performed on it; If there are no remaining fault-affected transmission paths available for selection, the identified abnormal data will be marked as interference data and a warning will be recorded.

2. The servo motor fault self-diagnosis method according to claim 1, characterized in that: Selecting a transmission node in the transmission path affected by the fault as a blocking node and performing an impact blocking operation on the node includes: Obtaining the blocking coefficient of each transmission node, the probability loss coefficient of the fault affecting the transmission between two adjacent transmission nodes, and the importance level coefficient of each transmission node; Extract the blocking coefficient of each transmission node in the transmission path affected by the currently matched fault and the probability loss coefficient between two adjacent transmission nodes, and calculate and select one of the target transmission nodes for the impact blocking operation; The blocking coefficient is defined as the probability that the fault impact will spread to the next transmission node after the impact blocking operation is implemented on the transmission node; The probability loss coefficient is defined as: the probability that the failure impact will successfully spread from the current transmission node to the next transmission node under natural conditions; The importance level coefficient is defined as: the correlation coefficient between the offset value of the sensor data at the transmission node and the positioning accuracy of the servo motor; The impact blocking operation includes: maintaining the value of any one or a combination of temperature, current, voltage, and vibration amplitude to fluctuate within a reference range; The calculation and selection of one of the target transmission nodes for impact blocking operations include: Configuring a blocking expectation parameter L for each transmission node on the transmission path affected by the fault; Sort by the size of each blocking expectation parameter L, and select the transmission node with the highest ranking to perform the impact blocking operation; Among them, the blocking expectation parameter L=w1·P1+w2·P2+k; P1 is the blocking coefficient, P2 is the probability loss coefficient, w1 and w2 are the weights corresponding to the blocking coefficient and the probability loss coefficient respectively, and k is the correction coefficient; The above-mentioned constraint conditions for the blocking expectation parameter L are: In the above constraints, M is the importance level coefficient of each transmission node located after the blocking node on the transmission path affected by the fault, n is the number of transmission nodes located after the blocking node on the transmission path affected by the fault, and X is a set threshold.

3. The servo motor fault self-diagnosis method according to claim 2, characterized in that: The correlation between the various sensor data includes a temporal relationship and a numerical relationship between the various sensor data after normalization; The fault self-diagnosis method further includes: Configure at least one decision threshold and its corresponding control instruction for each transmission node; Obtain the starting node of the transmission path affected by the current fault and its corresponding starting time and value; Based on the timing and numerical relationships between the transmission nodes, the expected numerical value of each transmission node on the transmission path affected by the fault at each time is calculated and generated; Compare the expected value corresponding to each transmission node with the determination threshold at the transmission node, and output the set control instruction according to the comparison result; The control instructions include maintaining the current action, outputting an alarm message, executing a set operation action, or shutting down the servo system.

4. The servo motor fault self-diagnosis method according to claim 2, characterized in that: In the fault self-diagnosis method, matching and confirming at least one fault-affected transmission path further comprises: Based on the historical operation data of the servo motor and the correlation analysis method, the correlation coefficient between various sensor data is obtained; Selecting at least two pieces of sensor data with a correlation coefficient greater than a set value and located in the same fault-affected transmission path, obtaining a temporal relationship and a numerical relationship between the at least two pieces of sensor data, combining them to generate a feature array, and storing the array in association with the corresponding fault-affected transmission path; Based on the correlation between various types of sensor data, sensor data associated with the abnormal data is retrieved, the associated sensor data is compared with the feature array, and at least one fault-affected transmission path is matched and confirmed.

5. The servo motor fault self-diagnosis method according to claim 2, characterized in that: The fault self-diagnosis method further includes: According to the data type of the sensor data, the sensor data is divided into electrical data and mechanical data; Establishing an electrical parameter correction model and a mechanical parameter correction model for characterizing the correlation between changes in external parameters and fluctuations in the electrical and mechanical data; Acquire the external parameters of the servo system in real time during operation, and calculate and generate the correction coefficient k corresponding to the expected blocking parameter L of each transmission node according to the correction model; The external parameters associated with the electrical data include the ambient temperature of the servo system, the electromagnetic radiation intensity, and the electrical parameters of the servo system power supply connection; The external parameters associated with the mechanical data include vibration amplitude.

6. The servo motor fault self-diagnosis method according to claim 2, characterized in that: The fault self-diagnosis method further includes: Establish the correlation between the probability of abnormality of various sensor data and the frequency and range of abnormal data of the set category; Calculate the frequency of occurrence of each category of abnormal data and the average value of the abnormal range in the historical operation data, and generate the abnormal probability of each type of sensor data based on the above correlation relationship; The probability loss coefficient is adjusted based on the above abnormal probability.

7. A servo motor fault self-diagnosis system, characterized in that: include: A database unit (1) is configured to store historical operating data of the servo motor, correlation data between various sensor data, real-time operating data, and a reference data queue corresponding to each fault event; An abnormal data identification unit (2) includes a verification model generated based on historical operating data combined with a clustering algorithm, which is used to obtain various sensor data of the servo motor in real time and extract data features, analyze and identify abnormal data based on the verification model, and mark the fault type and start time corresponding to the abnormal data; A feature data queue generating unit (3) is configured to be data-connected to the abnormal data identifying unit (2), retrieve sensor data associated with the abnormal data based on the correlation between various types of sensor data, analyze and determine whether the associated sensor data deviates from the reference value within a set period and the time of deviation, and generate a feature data queue according to the order of deviation time; A fault-affected transmission path confirmation unit (4) is configured to compare the characteristic data queue with each stored reference data queue to match and confirm at least one fault-affected transmission path; An impact blocking verification unit (5) is configured to select a transmission path affected by the fault and select one of the transmission nodes to perform an impact blocking operation, monitor the data offset direction of each transmission node on the transmission path affected by the fault in real time, determine the fault source based on the data offset direction combined with the set steps, and record an early warning; The early warning output unit (6) is configured to be data-connected to the impact blocking verification unit (5) to receive and output early warning information.

8. The servo motor fault self-diagnosis system according to claim 7, characterized in that: The impact blocking verification unit (5) is built with a blocking node selection module, including: A reference coefficient acquisition submodule (7) is configured to acquire and temporarily store the blocking coefficient of each transmission node, the probability loss coefficient of the fault affecting the transmission between two adjacent transmission nodes, and the importance level coefficient of each transmission node; A reference coefficient extraction submodule (8) is configured to extract the blocking coefficient of each transmission node in the fault-affected transmission path currently matched and confirmed, and the probability loss coefficient between two adjacent transmission nodes; A node selection submodule (9) is configured to configure a blocking expectation parameter L for each transmission node on the transmission path affected by the fault, sort the nodes according to the size of the blocking expectation parameters L, and select the transmission node with the highest ranking to perform the impact blocking operation; The expected blocking parameter L is obtained by the calculation formula of the expected blocking parameter L disclosed in the servo motor fault self-diagnosis method according to claim 2.

9. The servo motor fault self-diagnosis system according to claim 8, characterized in that: The system further includes an instruction presetting unit, including: A first storage module is configured to obtain and store a determination threshold corresponding to each transmission node and a control instruction corresponding to the determination threshold; A second storage module is configured to acquire and store the temporal relationship between the various types of sensor data and the normalized numerical relationship between the various types of sensor data; A starting node acquisition module configured to acquire the starting node of the transmission path affected by the current fault and its corresponding starting time and numerical value; An expected numerical value generation module is configured to calculate and generate an expected numerical value of each transmission node on the transmission path affected by the fault at each moment based on the timing relationship and numerical relationship between the transmission nodes; An instruction confirmation output module is configured to compare the expected numerical value corresponding to each transmission node with the determination threshold value at the transmission node, and output a set control instruction according to the comparison result; The control instructions include maintaining the current action, outputting an alarm message, executing a set operation action, or shutting down the servo system.

10. A computer-readable storage medium, characterized in that A program module for implementing the servo motor fault self-diagnosis method according to any one of claims 1 to 6 is loaded thereon.

Images