A method of motor fault analysis

Abstract

The application discloses a motor fault analysis method, a sensor is arranged on a motor body, the sensor collects motor operation parameters in real time, and motor fault modes are positioned and recognized; a vibration sensor is arranged at the bottom of the motor body and is used for detecting vibration data of the motor body in the running process; a speed sensor for detecting rotating shaft speed data of the motor body is arranged on the motor body, a torque sensor for detecting rotating torque data of the motor body is arranged on the motor body, and a temperature sensor for detecting temperature data of the motor body is arranged on the motor body; a current sensor for detecting motor input current stability data is further arranged on the motor body; and a fault analysis device is used for analyzing the type of motor fault. By inputting standard values into a trained fault prediction model, the fault type and the risk level of the fault type are predicted, subjective statistical calculation is avoided, the function of prediction can be realized, and the unscientific nature of subjective statistical calculation can be effectively avoided.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to the technical field of motor fault analysis, and particularly relates to a motor fault analysis method. BACKGROUND

[0002] In industrial production, equipment failures often occur, and after the equipment fails, the daily production work is seriously affected, therefore, how to effectively analyze the fault reason and predict the fault is more and more concerned by people.

[0003] Failure Mode Effects and Criticality Analysis (FMECA) is to determine the influence of each failure mode on the product work according to the analysis of the failure mode, find out the single point failure, and determine the criticality according to the severity of the failure mode and the probability of its occurrence. The so-called single point failure refers to the local failure that causes product failure and has no redundancy or alternative work procedure as a remedy. FMECA includes failure mode and effect analysis and criticality analysis.

[0004] However, the current FMECA analysis method has the following problems: more checking (containment) than prevention (control); the risk priority number (RPN) is the product of severity S, frequency O and detectability D, and the calculation method lacks scientificity; S, O and D are subjective scores without definable distance measure.

[0005] It should be noted that the information disclosed in the above background section is only used to strengthen the understanding of the background of the present disclosure, and therefore can include information that does not constitute prior art known to those of ordinary skill in the art. SUMMARY

[0006] The purpose of the present disclosure is to provide a fault analysis method, device, storage medium and computer system, thereby at least to some extent overcoming the situation that more checking (containment) than prevention (control) due to the limitations and defects of the related art; the RPN is the product of severity S, frequency O and detectability D, and the calculation method lacks scientificity; S, O and D are subjective scores without definable distance measure.

[0007] Other characteristics and advantages of the present disclosure will become apparent from the following detailed description, or will be learned by practice of the present disclosure.

[0008] The present disclosure provides a motor fault analysis method, comprising: a sensor is arranged on a motor body, the sensor collects motor operating parameters in real time, and locates and identifies a motor failure mode.

[0009] A vibration sensor is arranged at the bottom of the motor body, and is used to detect vibration data of the motor body during operation;

[0010] A speed sensor is arranged on the motor body, and is used to detect rotation speed data of the motor body, and a torque sensor is arranged on the motor body, and is used to detect rotation torque data of the motor body, and a temperature sensor is arranged on the motor body, and is used to detect temperature data of the motor body;

[0011] A current sensor is further arranged on the motor body, and is used to detect input current stability data of the motor;

[0012] The fault analysis device is used to receive and analyze the type of the motor fault based on the vibration data, the rotation speed data, the rotation torque data, the temperature data and the input current stability data.

[0013] Optionally, a plurality of variables causing the motor fault are determined;

[0014] The plurality of variable systems are converted into a plurality of independent single-variable systems by using an intelligent decoupling algorithm;

[0015] The system is decoupled into a plurality of subsystems, and then is independently analyzed and diagnosed, so as to realize failure cause tracing under single fault or multiple faults.

[0016] Optionally, the method comprises:

[0017] Historical fault data and operation data of the equipment are collected, wherein the operation data comprises vibration data, rotation speed data, rotation torque data, temperature data and input current stability data;

[0018] A neural network algorithm is used to train the relationship between the sensor signals and the failure modes, and then the failure modes are predicted according to the collected signals and the trained model;

[0019] The step of collecting historical fault data and operation data of the equipment comprises:

[0020] The historical fault data and the operation data of the equipment are collected, and comprise various sensor data and operation records of the equipment;

[0021] The step of using a neural network algorithm to train the relationship between the sensor signals and the failure modes, and then predicting the failure modes according to the collected signals and the trained model comprises:

[0022] The collected data are cleaned and preprocessed, including removing abnormal values, processing missing values and standardizing data;

[0023] The preprocessed data is subjected to feature extraction, and the neural network model is trained using the extracted features and historical data to obtain a trained training model;

[0024] The motor operating parameters detected by the sensor are input into the trained training model to predict the fault of the motor body, and then the motor fault mode is located and identified.

[0025] Optionally, the method further comprises:

[0026] Based on the positioning and identification of the motor fault mode, a maintenance strategy for the motor body is formulated.

[0027] Optionally, in the motor body stop state, a target detection strategy is configured; wherein the target detection strategy comprises:

[0028] Line detection, motor overload detection, motor body mechanism jam detection, motor shaft damage detection, motor winding burnout detection, motor capacitor fault detection, control circuit fault detection.

[0029] Optionally, the historical fault data of the equipment is obtained, the statistical clustering algorithm is used to analyze the historical data, the severity matrix is established, and the severity level of the fault mode is determined, comprising:

[0030] Based on the severity of the severity level, a weight is configured for each of the severity levels;

[0031] Based on the frequency of occurrence of each fault type, the possibility of occurrence of a certain fault mode is represented;

[0032] By configuring the severity level with weight and the frequency of occurrence of the fault type, a severity matrix is constructed, and the severity level of the fault mode is determined based on the severity matrix;

[0033] According to the severity level of the fault mode, corresponding processing measures are configured for fault modes of the same level.

[0034] Optionally, the risk priority number method and the hazard matrix method are combined to analyze the hazard of the fault mode;

[0035] RPN = ESR x OPR

[0036] PRN is the product of the severity level ESR of the fault mode and the occurrence probability level OPR of the fault mode, based on the high and low of the RPN value, the hazard of the fault is determined, the higher the RPN value, the greater the hazard.

[0037] Optionally, the probability density function and the mean time between failures MTBF of the original fault type of the equipment are determined.

[0038] The original function of the original failure probability density function is obtained, and the sampling statistical method is used to calculate the failure probability density function after the basic repair;

[0039] The average life and standard deviation after the first repair are calculated, and the second, the third, and the nth are calculated.

[0040] The first failure time of the device in the actual operation condition is determined, and the service life Y of the device is continued; the expected repair cost of the failure is estimated by multiplying the failure cost by the probability of the failure; wherein the failure cost refers to the economic loss caused after the failure, and the failure repair cost is related to the number of failures.

[0041] The sensor is arranged on the motor body, the sensor collects the motor operating parameters in real time, and the motor fault mode is positioned and identified; a vibration sensor is arranged at the bottom of the motor body, and the vibration sensor is used to detect the vibration data of the motor body during operation; a speed sensor for detecting the rotating shaft speed data of the motor body is arranged on the motor body, and a torque sensor for detecting the rotating torque data of the motor body is arranged, and a temperature sensor for detecting the temperature data of the motor body is arranged; a current sensor for detecting the input current stability data of the motor is also arranged on the motor body; the fault analysis device is used to receive and analyze the type of the motor fault based on the vibration data, the rotating speed data, the rotating torque data, the temperature data and the current stability data. By inputting the standard value into the trained fault prediction model, the fault type and the risk level of the fault type are predicted, avoiding subjective statistical calculation, which not only realizes the prediction function, but also effectively avoids the unscientificness of subjective statistical calculation. It can be seen that the scheme of the application can overcome the limitation and defect of the related art to a certain extent, and is more for checking (controlling) than for preventing (controlling); wherein RPN is the product of severity S, frequency O and detection D, and the calculation method lacks scientificity; S, O and D are subjective scores, and there is no definable distance measurement.

[0042] It should be understood that the above general description and the following detailed description are only exemplary and explanatory, and cannot limit the present disclosure. BRIEF DESCRIPTION OF DRAWINGS

[0043] The drawings incorporated into the specification and forming part of the specification, show embodiments consistent with the present disclosure, and together with the specification, serve to explain the principles of the present disclosure. It is obvious that the drawings in the following description are only some embodiments of the present disclosure, and other drawings can be obtained by those skilled in the art without creative labor.

[0044] Figure 1A structural schematic diagram of a motor fault analysis device in an exemplary embodiment of the present disclosure is shown schematically.

[0045] Figure 2 An architecture diagram of an open-loop feedforward decoupling algorithm in an exemplary embodiment of the present disclosure is shown schematically.

[0046] Figure 3 A hazard matrix diagram in an exemplary embodiment of the present disclosure is shown schematically.

[0047] Figure 4 A training diagram of a fault prediction model in an exemplary embodiment of the present disclosure is shown schematically.

[0048] Figure 5 A flowchart of a motor fault analysis method in an exemplary embodiment of the present disclosure is shown schematically. DETAILED DESCRIPTION

[0049] Example implementations are now described with reference to the drawings. Example implementations can, however, be implemented in many different forms and should not be construed as limited to the examples set forth herein; rather, these implementations are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the example implementations to those skilled in the art. The described features, structures, or characteristics can be combined in any suitable manner in one or more implementations.

[0050] In addition, the drawings are only schematic and are non-limiting. Like references signs denote like parts throughout the drawings and textual description, whereupon repeated description of like parts can be omitted. Some of the blocks in the drawings are functional blocks, which do not necessarily correspond to physical or logical independent entities. These functional blocks can be implemented in software, or in one or more hardware modules or integrated circuits, or in different network and / or processor devices and / or microcontroller devices.

[0051] The block diagrams shown in the drawings are merely functional blocks, which do not necessarily correspond to physically independent entities. That is, these functional blocks can be implemented in software, or in one or more hardware modules or integrated circuits, or in different network and / or processor devices and / or microcontroller devices.

[0052] The flowcharts shown in the drawings are merely exemplary illustrations, and do not necessarily include all contents and operations / steps, nor are they necessarily executed in the described order. For example, some operations / steps can be further divided, and some operations / steps can be combined or partially combined, so that the actual execution order can be changed according to the actual situation.

[0053] The motor fault analysis device provided by the present disclosure comprises a motor body and a fault analysis device, as shown in Figure 1 The motor body is provided with a vibration sensor at the bottom thereof, which is used to detect vibration data of the motor body during operation. The motor body is provided with a speed sensor for detecting rotation speed data of the motor body, a torque sensor for detecting rotation torque data of the motor body, and a temperature sensor for detecting temperature data of the motor body. The motor body is also provided with a current sensor for detecting input current stability data of the motor. The fault analysis device is used to receive and analyze the type of motor fault based on the vibration data, the rotation speed data, the rotation torque data, the temperature data, and the current stability data.

[0054] In the present example embodiment, the present disclosure provides a motor fault analysis method, comprising: providing a sensor on the motor body, which collects motor operating parameters in real time, locates and identifies motor fault modes; the motor body is provided with a vibration sensor at the bottom thereof, which is used to detect vibration data of the motor body during operation. The motor body is provided with a speed sensor for detecting rotation speed data of the motor body, a torque sensor for detecting rotation torque data of the motor body, and a temperature sensor for detecting temperature data of the motor body. The motor body is also provided with a current sensor for detecting input current stability data of the motor. The fault analysis device is used to receive and analyze the type of motor fault based on the vibration data, the rotation speed data, the rotation torque data, the temperature data, and the current stability data.

[0055] In a specific embodiment, a plurality of variables causing motor failure are determined; the plurality of variable systems are converted into a plurality of independent single variable system mappings by using an intelligent decoupling algorithm; the system is decoupled into several subsystems, and then independent analysis and diagnosis are performed to realize failure cause tracing under single fault or multiple faults.

[0056] Specifically: for motor failure caused by multiple variables, the plurality of variable systems are converted into a plurality of independent single variable system mappings by using an intelligent decoupling algorithm, the system is decoupled into several subsystems, and then independent analysis and diagnosis are performed to realize failure cause tracing under single fault or multiple faults.

[0057] Referring to Figure 2 As shown in FIG. 2, taking a 2-input, 2-output object as an example, an open-loop feedforward decoupling algorithm based on a neural network is adopted. For a certain channel, the influence of the remaining channels on it is regarded as an interference signal, and a feedforward compensation method is used to eliminate the interference signal.

[0058] As Figure 2As shown in the figure, f 11 12 21 22 N 12 N 21 is a neural network decoupling link, for the first main channel f 11 and output y1(k+1), the input u2(k) of the second channel can be regarded as a measurable disturbance, which is eliminated by introducing a feedforward compensation link N 12 , that is, N 12 = f 21 · f 22 -1 .

[0059] After introducing N 12 , N 21 , y1(k+1) is only controlled by r1(k), and the mapping relationship between them is f 11 , y2(k+1) is only controlled by r2(k), and the mapping relationship between them is f 22 , that is, the single variable system after decoupling has the characteristics of the original object main channel.

[0060] In a specific embodiment, it includes: collecting historical failure data and running data of the equipment; wherein the running data includes vibration data, rotation speed data, rotation torque data, temperature data, current stability data; training the relationship between sensor signals and failure modes by using a neural network algorithm, and then predicting the failure mode according to the collected signals and the trained model; wherein the step of collecting historical failure data and running data of the equipment includes: collecting historical failure data and running data of the equipment, including various sensor data and operation records of the equipment; the step of training the relationship between sensor signals and failure modes by using a neural network algorithm, and then predicting the failure mode according to the collected signals and the trained model includes: cleaning and preprocessing the collected data, including removing outliers, processing missing values, and standardizing data; extracting features from the preprocessed data; and training a neural network model using the extracted features and historical data to obtain a trained training model; inputting the motor operating parameters detected by the sensor into the trained training model to predict the fault of the motor body, and then positioning and identifying the motor fault mode.

[0061] Specifically: collecting historical failure data and running data of the equipment, training the relationship between sensor signals and failure modes by using a neural network algorithm, and then predicting the failure mode according to the collected signals and the trained model, which provides a basis for predictive maintenance of equipment in industrial production, including the following steps:

[0062] ​​​① Data collection: Collect historical failure data and operation data of the equipment, including various sensor data, operation records, etc.

[0063] ② Data cleaning and preprocessing: Clean and preprocess the collected data, including removing outliers, handling missing values, standardizing data, etc.

[0064] ③ Feature extraction: Extract useful features from the preprocessed data, such as equipment vibration frequency, temperature change, current change, etc.

[0065] ④ Neural network model construction: Select and construct a suitable neural network model, such as multilayer perceptron (MLP), convolutional neural network (CNN), etc.

[0066] ⑤ Model training: Train the neural network model using historical data, including setting appropriate learning rate, iteration number, etc. Train the algorithm model with multi-sensor data as node input and fault type as node output to realize comprehensive analysis and diagnosis of multi-source data collected by sensors (such as current, voltage, temperature, pressure, speed, etc.).

[0067] The prediction model can accurately predict and diagnose the fault type of "multiple causes and one effect" or "multiple causes and multiple effects".

[0068] In this embodiment, as shown in Figure 4 The prediction model is trained using a training data set, specifically: the training method includes:

[0069] Based on the type of target data, the time of target data occurrence, and the duration of target data in the retrospective cohort of the training data set, a feature sample set is determined.

[0070] The obtained feature sample set is preprocessed to obtain a feature vector corresponding to the feature.

[0071] The obtained feature vector is divided into a training data set and a validation data set.

[0072] The training data set is used to train the fault prediction model, and the validation data set is used to verify the trained fault prediction model, obtaining a trained fault prediction model.

[0073] The trained fault prediction model includes a signal input end, a hidden processing end, and a fault type output end.

[0074] In the example embodiment, the failure prediction model of the application is a multi-sensor coupling technology based on artificial neural network algorithm. The algorithm model is trained with multi-sensor data as node input and failure type as node output, realizing comprehensive analysis and diagnosis of multi-source data (such as current, voltage, temperature, pressure, speed, etc.) collected by sensors, especially precise prediction of "multiple causes and one effect" and "multiple causes and multiple effects".

[0075] Referring to Figure 4 As shown, X1, X2, X3...Xn represent the standard values of the data collected by a series of sensing devices such as ammeter, voltmeter, temperature sensor...speed sensor after preprocessing, and n is the total number of sensing device measuring points; Y1, Y2, Y3...Ym represent a series of failure types such as fracture, wear, corrosion...deformation, and m is the total number of failure types; f1, f2, f3...fv represent each hidden layer. 0≤X1, X2, X3...Xn≤1, when Xn is 0, it means that the signal detected by the sensor is irrelevant to the failure type, for example, the data collected by the smoke sensor is irrelevant to the failure type of the shaft fracture.

[0076] In a specific embodiment, the method further comprises formulating a maintenance strategy for the motor body based on the positioning and identifying the motor failure mode.

[0077] Specifically, safety, reliability, and maintenance cost are used as constraint conditions, and the above prediction results are used as basis to formulate and recommend preventive and restorative measures. According to the use and historical failure of the system, appropriate preventive measures are selected and adjusted and optimized according to the actual situation; and according to the failure type and the degree of influence, the restorative measures are selected to avoid excessive maintenance and waste of resources.

[0078] In a specific embodiment, a target detection strategy is configured in the motor body stop state; wherein the target detection strategy includes line detection, motor overload detection, motor body mechanism jam detection, motor shaft damage detection, motor winding burnout detection, motor capacitor fault detection, and control circuit fault detection.

[0079] According to engineering practice and experience, commonly used detection methods, detection tools and measurement values are summarized, a detection method, detection tool and measurement value database is established, and corresponding detection methods, detection tools and measurement values are recommended according to the type of equipment and failure mode.

[0080] In the case of insufficient or failure of the sensing element, multiple screening is required to determine the failure mode. Probability and logical judgment are introduced in the recommended algorithm to quickly screen the failure cause.

[0081] For example, when the motor is stalled, the system generates the following strategy as a recommendation until the failure cause is successfully detected.

[0082]

[0083] In one embodiment, historical failure data of the equipment is acquired, statistical clustering algorithm is used to analyze the historical data, a severity matrix is established, and the severity level of the failure mode is determined, including: based on the severity of the severity level, a weight is configured for each of the severity levels; based on the frequency of occurrence of each failure type, the possibility of occurrence of a certain failure mode is represented; by configuring the severity level with the weight and the frequency of occurrence of the failure type, a severity matrix is constructed, and the severity level of the failure mode is determined based on the severity matrix; and according to the severity level of the failure mode, corresponding processing measures are configured for failure modes of the same level.

[0084] The historical failure data of the equipment is acquired, statistical clustering algorithm is used to analyze the historical data, a severity matrix is established, and the severity level of the failure mode is determined.

[0085] The consequences of the failure type may include physical damage, safety risk, environmental problem, quality problem, reliability problem, etc., and according to the severity of these consequences, a corresponding weight is given to each consequence, and the weight of a more serious consequence is higher.

[0086] The frequency of occurrence of each failure type is counted to represent the possibility of occurrence of a certain failure mode.

[0087] A severity matrix is established through the above two dimensions to determine the severity level of the failure mode.

[0088] According to the severity level of the failure mode, different measures are taken for failure modes of different levels to reduce the risk, such as prioritizing the processing of failure modes with high severity, and taking more stringent inspection and testing, etc.

[0089]

[0090] In this embodiment, the operation data of the equipment includes:

[0091] The to-be-predicted equipment is determined, and a detection tool is determined according to the structure, size, operation accuracy, and range of operation data of the to-be-predicted equipment.

[0092] The to-be-predicted equipment is detected by using the determined detection tool.

[0093] In one embodiment, the to-be-predicted equipment is determined, and a detection tool is determined according to the structure, size, operation accuracy, and range of operation data of the to-be-predicted equipment, including:

[0094] An association between the target prediction device and the detection tool is constructed and stored to establish a detection tool database.

[0095] A target device under test is determined, and a target detection tool is indexed in the detection tool database based on the association between the target device under test and the target detection tool.

[0096] The target detection tool is called as the detection tool.

[0097] In the example embodiment, the fault detection method module is improved, a special database of detection tools is established, and the general tool and special tool databases are divided, and specific measurement values are recommended.

[0098] The database is shown in the following table:

[0099]

[0100]

[0101]

[0102]

[0103] In one specific embodiment, the risk priority number method and the hazard matrix method are combined to analyze the hazard of the fault mode.

[0104] RPN = ESR x OPR

[0105] PRN is equal to the product of the severity level ESR of the fault mode and the occurrence probability level OPR of the fault mode. Based on the high and low of the RPN value, the hazard of the fault is determined. The higher the RPN value, the greater the hazard.

[0106] Specifically, the risk priority number (RPN) method and the hazard matrix method are combined to analyze the hazard of the fault mode.

[0107] RPN = ESR x OPR

[0108] PRN is equal to the product of the severity level ESR of the fault mode and the occurrence probability level OPR of the fault mode. The higher the RPN value, the greater the hazard, and more attention needs to be given to the fault mode.

[0109]

[0110]

[0111] According to the severity and failure probability level of each failure mode, see Figure 3 .

[0112] Figure 3 In the figure, the marked failure mode distribution points are perpendicular to the diagonal line (dotted line OP), and the distance from the intersection point of the perpendicular line and the diagonal line to the origin is used as a measure of the harm of the failure mode (or product). The longer the distance, the greater the harm, and the sooner the improvement measures should be taken. In the above figure, because the distance O1 is longer than the distance O2, the failure mode M1 is more harmful than the failure mode M2.

[0113] In a specific embodiment, the probability density function and the mean time between failures MTBF of the original failure type of the equipment are determined. The original function of the original failure probability density function is obtained, and the basic repair failure probability density function is calculated by sampling statistical method. The mean life and standard deviation after the first repair are calculated, and the second, the third, and so on are determined. The first failure time of the equipment under actual operation is determined, and the equipment continues to run for Y years. The expected maintenance cost of the failure is estimated by multiplying the failure cost by the probability of the failure occurring. The failure cost refers to the economic loss caused after the failure occurs, and the failure maintenance cost is related to the number of failures.

[0114] Specifically: the failure cost can be evaluated by the failure maintenance cost and the failure probability density. The probability density function and the mean time between failures MTBF of the original failure type of the equipment are determined. The original function of the original failure probability density function is obtained, and the basic repair failure probability density function is calculated by sampling statistical method. The mean life and standard deviation after the first repair are calculated, and the second, the third, and so on are determined. The first failure time of the equipment under actual operation is determined, and the equipment continues to run for Y years. The failure cost refers to the economic loss caused after the failure occurs, and the failure maintenance cost is related to the number of failures. The expected maintenance cost of the failure is estimated by multiplying the failure cost by the probability of the failure occurring.

[0115] In the example embodiment, see Figure 5 , the FMECA analysis device in the application includes a motor failure data acquisition module, a failure data training and learning module, a calculation module, a prediction module, a decision module, and an analysis and evaluation module.

[0116] Specifically, see Figure 5As shown, the operation principle of the above FMECA analysis device is as follows: the sensor monitors the motor state in real time, locates and identifies the motor failure mode; the Kalman filtering algorithm is used to decouple the multi-source mixed waveform, and independent analysis and diagnosis are performed; a neural network model with multi-sensor input and fault type output is trained; the fault type under the influence of single or complex factors is predicted; a general and special tool database is established, and a measurement tool and a measurement value are recommended; the fault reason weight under the fault mode is distributed, and a priority ranking is established; the fuzzy comprehensive evaluation method is used to quantify the hazard degree of the fault mode; the fault cost is evaluated and a maintenance scheme is provided; and a one-key generation full-process analysis report is generated.

[0117] In addition, the above figures are only schematic illustrations of the processes included in the method according to the exemplary embodiments of the present application, and are not for limiting purposes. It is easy to understand that the processes shown in the above figures do not indicate or limit the time sequence of these processes. In addition, it is also easy to understand that these processes can be executed synchronously or asynchronously, for example, in multiple modules.

[0118] Other embodiments of the present disclosure will be apparent to those skilled in the art from consideration of the specification and practice of the application disclosed herein. This application is intended to cover any variations, uses, or adaptations of the present disclosure following the general principles thereof and including such departures from the present disclosure that come within known use or custom in the art. The specification and examples are to be regarded as illustrative only, and the true scope and spirit of the present disclosure are indicated by the claims.

[0119] It should be understood that the present disclosure is not limited to the precise structures described above and illustrated in the drawings and that various modifications and changes can be made without departing from the scope thereof. The scope of the present disclosure is indicated by the appended claims.

Claims

1. A method of motor fault analysis, characterized by, The motor body is provided with a sensor, which collects motor operating parameters in real time, locates and identifies motor fault modes; A vibration sensor is arranged at the bottom of the motor body, which is used to detect vibration data of the motor body during operation; A speed sensor is arranged on the motor body to detect the rotating shaft speed data of the motor body, a torque sensor is arranged to detect the rotating torque data of the motor body, and a temperature sensor is arranged to detect the temperature data of the motor body; A current sensor is further arranged on the motor body to detect the input current stability data of the motor; A fault analysis device is used to receive and analyze the type of motor fault based on the vibration data, the rotating speed data, the rotating torque data, the temperature data and the current stability data; A plurality of variables causing the motor fault are determined, and an intelligent decoupling algorithm is used to convert the plurality of variable systems into a plurality of independent single-variable system mappings; the system is decoupled into a plurality of subsystems, which are independently analyzed and diagnosed to realize failure cause tracing under single fault or multiple faults; A target detection strategy is configured under the motor body stop state; The target detection strategy includes line detection, motor overload detection, motor body mechanism jamming detection, motor shaft damage detection, motor winding burning detection, motor capacitor fault detection and control circuit fault detection; The probability density function and the mean time between failures (MTBF) of the original fault type of the equipment are determined, the original function of the original fault probability density function is obtained, the basic repair fault probability density function is calculated by sampling and sampling statistical method, the average life and the standard deviation after the first repair are calculated, and the second time, until the nth time; the first failure time of the equipment in the actual running condition is determined, and the equipment continues to run for a certain period of Y; the expected maintenance cost of the fault is estimated by multiplying the fault cost by the probability of the fault; wherein, the fault cost refers to the economic loss caused after the fault occurs, and the fault maintenance cost is related to the number of faults. The historical fault data and operating data of the equipment are collected; wherein, the operating data includes vibration data, rotating speed data, rotating torque data, temperature data and current stability data; 2. The method of claim 1, wherein, The relationship between the sensor signal and the failure mode is trained by using a neural network algorithm, and then the failure mode is predicted according to the collected signal and the trained model; The step of collecting historical fault data and operating data of the equipment includes: Collecting historical fault data and operating data of the equipment, including various sensor data and operation records of the equipment; The step of training the relationship between the sensor signal and the failure mode by using a neural network algorithm, and then predicting the failure mode according to the collected signal and the trained model includes: Cleaning and preprocessing the collected data, including removing outliers, processing missing values and standardizing data; Feature extraction is performed on the preprocessed data; and the neural network model is trained by using the extracted features and historical data to obtain a trained training model; ​ ​ The motor operation parameters detected by the sensor are input into the trained training model to predict the fault of the motor body, and then the motor fault mode is located and identified.

3. The method of claim 2, wherein, The method further comprises: Based on locating and identifying the motor fault mode, a maintenance strategy for the motor body is formulated.

4. The method of claim 1, wherein, Obtaining historical fault data of the equipment, analyzing the historical data by using a statistical clustering algorithm, establishing a severity matrix, and determining the severity level of the fault mode include: Based on the severity of the severity level, a weight is configured for each of the severity levels; Based on the frequency of occurrence of each fault type, the possibility of occurrence of a certain fault mode is represented; By configuring the severity level with the weight and the frequency of occurrence of the fault type, a severity matrix is constructed, and the severity level of the fault mode is determined based on the severity matrix; According to the severity level of the fault mode, corresponding treatment measures are configured for fault modes of the same level.

5. The method of claim 4, wherein, The risk priority number method and the hazard matrix method are combined to analyze the hazard of the fault mode. RPN=ESR×OPR PRN is the product of the severity level ESR of the fault mode and the occurrence probability level OPR of the fault mode. Based on the high and low of the RPN value, the hazard of the fault is determined. The higher the RPN value, the greater the hazard.

Images