Servo motor fault diagnosis method and system based on data model analysis

Abstract

The present invention discloses a servo motor fault diagnosis method and system based on data model analysis, including using a current reconstruction method to analyze the current performance of the servo motor, and drawing a current linear diagram of a filter, and constructing a motor mathematical model of the servo motor in a synchronous rotating coordinate system according to the current linear diagram; constructing a torque-speed model and a position-speed model of the servo motor based on a data-driven method according to Newton's laws and kinematic relationships, and combining them to obtain a position model; and performing fault diagnosis on the servo motor according to a discrete multi-phase current set. Through in-depth analysis of real-time current signals, fault features can be more accurately identified, and a multi-dimensional model can be established to make fault diagnosis more comprehensive and systematic. By classifying fault feature quantities and evaluating reliability, the credibility of fault identification is improved, and a function panel provides an intuitive display of fault diagnosis results, which is convenient for operators to understand and make decisions.

Description

Technical Field

[0001] The present invention relates to the technical field of servo motor fault diagnosis, and in particular to a servo motor fault diagnosis method and system based on data model analysis. Background Art

[0002] A servo motor is a motor that can precisely control position, speed, and acceleration. It is usually used in conjunction with a position sensor and a control system. Data model analysis is a method of fault diagnosis based on collected data that can use a variety of algorithms and techniques to process and analyze data.

[0003] At present, the Chinese invention patent with application number CN202410962399.5 discloses a servo motor fault diagnosis method and system, which extracts features from a running sensor data set, generates multiple sensor data features, performs cluster analysis on the running sensor data set, generates a data distribution diagram based on the cluster analysis results, performs fault diagnosis on the target servo motor, generates fault diagnosis data, generates a fault signal based on the fault diagnosis data, and performs intelligent processing on the target servo motor based on the fault signal. The fault diagnosis of traditional motors requires shutdown inspection, which affects production efficiency, and lacks intelligence and automation based on manual experience, resulting in untimely and accurate fault detection and processing. The technical effect of accurately diagnosing and predicting servo motor faults is achieved through real-time monitoring of motor operation data, feature extraction, intelligent learning, etc.; however, the existing technology cannot accurately identify fault features through in-depth analysis of real-time current signals, cannot reduce the possibility of misjudgment, cannot make fault diagnosis more comprehensive and systematic from the perspective of multi-dimensional mathematical models, and cannot improve the credibility of fault identification through classification and reliability evaluation of fault feature quantities. Summary of the invention

[0004] The technical problem solved by the present invention is that the existing technology cannot accurately identify fault characteristics through in-depth analysis of real-time current signals, cannot reduce the possibility of misjudgment, cannot make fault diagnosis more comprehensive and systematic from the perspective of multi-dimensional mathematical models, and cannot improve the credibility of fault identification through classification and reliability evaluation of fault characteristic quantities.

[0005] To solve the above technical problems, the present invention provides the following technical solution: a servo motor fault diagnosis method based on data model analysis, comprising the following steps:

[0006] Step S1: Analyze the current performance of the servo motor by using a current reconstruction method, draw a current linear diagram of the filter, and construct a motor mathematical model of the servo motor in a synchronous rotating coordinate system according to the current linear diagram;

[0007] Step S2: constructing a torque-speed model and a position-speed model of the servo motor based on a data-driven method according to Newton's laws and kinematic relationships;

[0008] Step S3: combining the motor mathematical model, the torque-speed model and the position-speed model to obtain the position model of the servo motor;

[0009] Step S4: using the filter to define a phase sequence variable according to the position model, combining the phase sequence variable with the reconstructed signal after non-repetitive encoding and the corresponding current signal to multiply and calculate a discrete multi-phase current set, and performing fault diagnosis on the servo motor according to the discrete multi-phase current set.

[0010] Preferably, the step S1 includes the following sub-steps:

[0011] Step S11: using a digital filter or an analog filter according to the control requirements of the servo motor, wherein the filter is used to smooth the current signal and remove high-frequency noise;

[0012] Step S12: numbering the current signals to be input, the numbers are sorted according to the magnitudes of the current signals, and the sorted current signals are input into the filter for filtering by a current reconstruction method, and obtaining the reconstructed signal includes obtaining the dynamic response of the servo motor according to the motor control algorithm from the filtered signal, and the dynamic response is the reconstructed signal;

[0013] Step S13: performing linear calculation on the reconstructed signal by fitting a linear equation using the least square method, obtaining a linear function of the relationship between the magnitude of the current signal and time, frequency and the reconstructed signal, and drawing the linear function using a drawing tool to obtain a current linear graph;

[0014] Step S14: using sensors to collect dynamic data of the servo motor, and calculating a motor mathematical model of the servo motor in a synchronous rotating coordinate system according to a dynamic equation and the current linear diagram.

[0015] Preferably, the step S14 comprises:

[0016] The dynamic data includes the moment of inertia, load torque, angular velocity, inductance, resistance and voltage applied to the servo motor of the servo motor;

[0017] The mathematical expression of the kinetic equation is:

[0018] ;

[0019] in, is the moment of inertia, is the angular velocity, is the output torque, is the load torque, For time;

[0020] The mathematical expression of output torque is:

[0021] ;

[0022] in, is the torque constant, is the current signal;

[0023] The relationship between the current and voltage of the servo motor is obtained by the electrical equation, and the mathematical expression of the electrical equation is:

[0024] ;

[0025] in, is the voltage applied to the servo motor, is the inductor, is the resistance, is the electromotive force constant;

[0026] In the synchronous rotating coordinate system, the current signal of the servo motor is converted into a coordinate system synchronized with the rotor of the servo motor. The mathematical expression of the converted current signal coordinate system is:

[0027] ;

[0028] ;

[0029] in, is the original format data of the current signal, It is the format data of the current signal in the coordinate system synchronized with the rotor of the servo motor;

[0030] The mathematical expression for transforming the torque coordinate system is:

[0031] ;

[0032] According to the dynamic equation, the mathematical expression of the motor output torque, the mathematical expression of the converted current signal coordinate system and the mathematical expression of the converted torque coordinate system, the state space relationship between the moment of inertia, angular velocity, load torque, inductance, resistance and the voltage applied to the servo motor and the current signal is obtained, and the mathematical expression of the motor mathematical model for obtaining the state space relationship is:

[0033] ;

[0034] ;

[0035] in, is the current signal, is the angular velocity corresponding to the current signal, Output of the motor mathematical model.

[0036] Preferably, step S2 comprises:

[0037] Constructing the torque-velocity model includes:

[0038] The corresponding load, effective radius and speed obtained by the sensor experimental test of the servo motor according to the numbered current signal are collected. The mathematical expression related to the output torque, speed and acceleration is:

[0039] ;

[0040] ;

[0041] in, is the acceleration;

[0042] The kinematic relationship of the servo motor is described by the relationship between speed and displacement, and its mathematical expression is:

[0043] ;

[0044] in, is the linear speed, is the effective radius;

[0045] The relationship between the torque and the speed is established according to the mathematical expression of the relationship between the output torque, the speed and the acceleration, the mathematical expression of the kinematic relationship of the servo motor and the polynomial regression method, including:

[0046] The established polynomial regression mathematical expression is:

[0047] ;

[0048] in, and is the dependent variable parameter, is the order of the polynomial;

[0049] Substitute the mathematical expression relating the output torque to the rotational speed and acceleration and the mathematical expression of the kinematic relationship of the servo motor into the polynomial regression mathematical expression respectively to obtain a torque-speed model.

[0050] Preferably, constructing the position-velocity model comprises:

[0051] The corresponding displacement and linear velocity of the servo motor obtained by the sensor experimental test according to the numbered current signal are collected, and the relationship between position and velocity is established by using a linear regression model, and its mathematical expression is:

[0052] ;

[0053] in, k and c is the constant term parameter of the linear regression model;

[0054] The collected displacement and linear velocity are substituted into the linear regression model for training to obtain a position-velocity model.

[0055] Preferably, step S3 includes the following sub-steps:

[0056] Step S31: discretizing the motor mathematical model obtained in step S1, including convolving and calculating the motor mathematical model with a discrete phase matrix to obtain a discrete motor mathematical model;

[0057] Step S32: transforming the discrete motor mathematical model by a matrix calculation unit;

[0058] Step S33: The transformed discrete motor mathematical model is subjected to formula calculation using a numerical substitution method with the torque-speed model and the position-speed model to obtain a position model.

[0059] Preferably, the step S32 includes:

[0060] The matrix calculation unit is constructed by:

[0061] Encapsulate basic code blocks in a matrix calculation unit to perform basic operations on the matrix, the basic code blocks are retrieved from the Internet, the basic operations include matrix transposition, inverse matrix operation, determinant operation, unit row transformation, unit column transformation, block calculation and vector operation, and logically combine the basic code blocks according to calculation rules to obtain corresponding combined code blocks, the calculation rules are obtained from historical calculation records, the combined code blocks include distance calculation code blocks, dot product code blocks, angle code blocks, residual calculation code blocks and estimation code blocks, the distance calculation code block is used to calculate the distance value between two matrices, the dot product code block is used to calculate the value of the dot product of two matrices, the angle code block is used to calculate the three-dimensional angle and direction value between two matrices, the residual calculation code block is used to calculate the residual value between two matrices, and the estimation code block is used to calculate the position estimation value of the matrix;

[0062] The motor mathematical model obtained in step S1 is input into the matrix calculation unit, and the motor mathematical model is subjected to logical calculation of the combined code blocks to obtain a matrix model with the same dimension as the torque-speed model and the position-speed model.

[0063] Preferably, step S4 includes the following sub-steps:

[0064] Step S41: defining a phase sequence variable according to the position model through a filter, wherein the phase sequence variable is obtained by calculating the phase difference and conforms to the Boolean operation formula of the series circuit;

[0065] Step S42: non-repetitively encode the reconstructed signal, and multiply the non-repetitively encoded reconstructed signal and the corresponding current signal by two by two with the phase sequence variable, and perform Fourier transform on the product to obtain a discrete multi-phase current set;

[0066] Step S43: calculating the fault characteristic quantity corresponding to each phase current according to the discrete multi-phase current set, and performing fault diagnosis according to the fault characteristic quantity;

[0067] Step S44: constructing a servo motor fault diagnosis system based on data model analysis.

[0068] Preferably, the step S43 includes the following sub-steps:

[0069] Step S431: using an array function to convert the phase current data included in the discrete multi-phase current set into a NumPy current array, respectively calculating the mean, variance and root mean square value of the NumPy current array, performing a spectrum analysis operation according to the mean, variance and root mean square value of the NumPy current array, and outputting a fault feature quantity, wherein the spectrum analysis operation includes calculating the frequency component of the NumPy current array by fast Fourier transform, and obtaining a harmonic distortion degree quantity, wherein the harmonic distortion degree quantity is the fault feature quantity;

[0070] Step S432: collecting the servo motor state corresponding to the historical fault type, comparing the current feature quantity with the feature quantity under the normal state, and performing fault identification includes:

[0071] Unbalanced fault is the current imbalance of the three phases;

[0072] Harmonic faults are high-order harmonics identified through spectrum analysis;

[0073] A short circuit fault occurs when the current in one phase rises significantly and does not match the other phases;

[0074] Step S433: Classifying the fault according to the fault characteristic quantity includes:

[0075] Calculate the reliability value corresponding to each fault characteristic value respectively, where the reliability value is the time proportion of the corresponding phase of the DC bus current within the time step h;

[0076] When the reliability value is lower than a preset time threshold, the fault characteristic quantity is unreliable, and a compensation voltage vector is inserted to improve reliability;

[0077] When the reliability value is higher than the preset time threshold, the fault characteristic quantity is reliable, and a secondary diagnosis step is entered by judging the fault characteristic quantity, and the secondary diagnosis step includes:

[0078] When the fault characteristic quantity is greater than the time threshold, it means that the DC bus current sensor detects a fault, and the current signal phase corresponding to the fault characteristic quantity is the inspection phase of the DC bus current sensor. Each inspection phase represents a corresponding fault phase, and each fault phase corresponds to a fault diagnosis type. The fault diagnosis types include normal, short circuit, overload and unbalanced.

[0079] Servo motor fault diagnosis system based on data model analysis, including function panel, calculation module and transmission module:

[0080] The function panel is used as a startup interface for visualizing fault diagnosis, and displays the fault diagnosis results obtained by real-time calculation in a coordinate system;

[0081] The calculation module is used to perform fault diagnosis on the servo motor using a servo motor fault diagnosis method based on data model analysis, and the calculation module is embedded in the main thread of the servo motor fault diagnosis system based on data model analysis;

[0082] The transmission module is used for:

[0083] After the slave thread of the servo motor fault diagnosis system based on data model analysis parses the real-time data of the servo motor, the parameter data required by the motor mathematical model, torque-speed model and position-speed model are sent to the calculation module located in the main thread with zero delay through signal broadcasting for fault diagnosis operation, the operation of the slave thread is interrupted by a preset timer, including sending an interrupt signal to the main thread at a unit time interval, and the fault diagnosis result output by the calculation module is sent to the coordinate system of the function panel;

[0084] After the calculation module completes the fault diagnosis calculation at the current moment, the diagnosis result is input into the memory space;

[0085] The real-time data includes the current signal, dynamic data, load, effective radius, rotation speed, displacement and linear speed of the servo motor;

[0086] The parameter data includes real-time data, dependent variable parameters and constant term parameters

[0087] The beneficial effects of the present invention are as follows: through in-depth analysis of real-time current signals, fault characteristics can be identified more accurately and the possibility of misjudgment can be reduced. By combining the motor mathematical model, torque-speed model and position-speed model, a multi-dimensional model is established to make fault diagnosis more comprehensive and systematic. It is suitable for servo motors of different types and specifications and provides solutions for various industrial applications. The credibility of fault identification is improved by classifying fault characteristic quantities and evaluating reliability. The function panel provides an intuitive display of fault diagnosis results, which is convenient for operators to understand and make decisions. BRIEF DESCRIPTION OF THE DRAWINGS

[0088] Figure 1 A basic flow chart of a servo motor fault diagnosis method based on data model analysis is provided as an embodiment of the present invention. DETAILED DESCRIPTION

[0089] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings. It is obvious that the described embodiments are only part of the embodiments of the present invention, but not all of the embodiments.

[0090] Reference Figure 1 , as an embodiment of the present invention, provides a servo motor fault diagnosis method based on data model analysis, comprising the following steps:

[0091] Step S1: Analyze the current performance of the servo motor by using a current reconstruction method, draw a current linear diagram of the filter, and construct a motor mathematical model of the servo motor in a synchronous rotating coordinate system according to the current linear diagram;

[0092] Step S2: constructing a torque-speed model and a position-speed model of the servo motor based on a data-driven method according to Newton's laws and kinematic relationships;

[0093] Step S3: combining the motor mathematical model, the torque-speed model and the position-speed model to obtain the position model of the servo motor;

[0094] Step S4: using the filter to define a phase sequence variable according to the position model, combining the phase sequence variable with the reconstructed signal after non-repetitive encoding and the corresponding current signal to multiply and calculate a discrete multi-phase current set, and performing fault diagnosis on the servo motor according to the discrete multi-phase current set.

[0095] Through in-depth analysis of real-time current signals, fault characteristics can be identified more accurately and the possibility of misjudgment can be reduced. By combining the motor mathematical model, torque-speed model and position-speed model, a multi-dimensional model is established to make fault diagnosis more comprehensive and systematic.

[0096] The step S1 comprises the following sub-steps:

[0097] Step S11: using a digital filter or an analog filter according to the control requirements of the servo motor, wherein the filter is used to smooth the current signal and remove high-frequency noise;

[0098] Step S12: numbering the current signals to be input, the numbers are sorted according to the magnitudes of the current signals, and the sorted current signals are input into the filter for filtering by a current reconstruction method, and obtaining the reconstructed signal includes obtaining the dynamic response of the servo motor according to the motor control algorithm from the filtered signal, and the dynamic response is the reconstructed signal;

[0099] Step S13: performing linear calculation on the reconstructed signal by fitting a linear equation using the least square method, obtaining a linear function of the relationship between the magnitude of the current signal and time, frequency and the reconstructed signal, and drawing the linear function using a drawing tool to obtain a current linear graph;

[0100] Step S14: using sensors to collect dynamic data of the servo motor, and calculating a motor mathematical model of the servo motor in a synchronous rotating coordinate system according to a dynamic equation and the current linear diagram.

[0101] The step S14 comprises:

[0102] The dynamic data includes the moment of inertia, load torque, angular velocity, inductance, resistance and voltage applied to the servo motor of the servo motor;

[0103] The mathematical expression of the kinetic equation is:

[0104] ;

[0105] in, is the moment of inertia, is the angular velocity, is the output torque, is the load torque, For time;

[0106] The mathematical expression of output torque is:

[0107] ;

[0108] in, is the torque constant, is the current signal;

[0109] The relationship between the current and voltage of the servo motor is obtained by the electrical equation, and the mathematical expression of the electrical equation is:

[0110] ;

[0111] in, is the voltage applied to the servo motor, is the inductor, is the resistance, is the electromotive force constant;

[0112] In the synchronous rotating coordinate system, the current signal of the servo motor is converted into a coordinate system synchronized with the rotor of the servo motor. The mathematical expression of the converted current signal coordinate system is:

[0113] ;

[0114] ;

[0115] in, is the original format data of the current signal, It is the format data of the current signal in the coordinate system synchronized with the rotor of the servo motor;

[0116] The mathematical expression for transforming the torque coordinate system is:

[0117] ;

[0118] According to the dynamic equation, the mathematical expression of the motor output torque, the mathematical expression of the converted current signal coordinate system and the mathematical expression of the converted torque coordinate system, the state space relationship between the moment of inertia, angular velocity, load torque, inductance, resistance and the voltage applied to the servo motor and the current signal is obtained, and the mathematical expression of the motor mathematical model for obtaining the state space relationship is:

[0119] ;

[0120] ;

[0121] in, is the current signal, is the angular velocity corresponding to the current signal, Output of the motor mathematical model.

[0122] The synchronous rotating coordinate system means that the coordinate system of the motor is synchronized with the angle of rotation of the rotor, which is convenient for analysis and control. The motor mathematical model provides a theoretical basis for the control and simulation of the servo motor and is suitable for the design and optimization of actual control systems.

[0123] The step S2 comprises:

[0124] Constructing the torque-velocity model includes:

[0125] The corresponding load, effective radius and speed obtained by the sensor experimental test of the servo motor according to the numbered current signal are collected. The mathematical expression related to the output torque, speed and acceleration is:

[0126] ;

[0127] ;

[0128] in, is the acceleration;

[0129] The kinematic relationship of the servo motor is described by the relationship between speed and displacement, and its mathematical expression is:

[0130] ;

[0131] in, is the linear speed, is the effective radius;

[0132] The relationship between the torque and the speed is established according to the mathematical expression of the relationship between the output torque, the speed and the acceleration, the mathematical expression of the kinematic relationship of the servo motor and the polynomial regression method, including:

[0133] The established polynomial regression mathematical expression is:

[0134] ;

[0135] in, and is the dependent variable parameter, is the order of the polynomial;

[0136] Substitute the mathematical expression relating the output torque to the rotational speed and acceleration and the mathematical expression of the kinematic relationship of the servo motor into the polynomial regression mathematical expression respectively to obtain a torque-speed model.

[0137] Constructing the position-velocity model includes:

[0138] The corresponding displacement and linear velocity of the servo motor obtained by the sensor experimental test according to the numbered current signal are collected, and the relationship between position and velocity is established by using a linear regression model, and its mathematical expression is:

[0139] ;

[0140] in, k and c is the constant term parameter of the linear regression model;

[0141] The collected displacement and linear velocity are substituted into the linear regression model for training to obtain a position-velocity model.

[0142] The step S3 comprises the following sub-steps:

[0143] Step S31: discretizing the motor mathematical model obtained in step S1, including convolving and calculating the motor mathematical model with a discrete phase matrix to obtain a discrete motor mathematical model;

[0144] Step S32: transforming the discrete motor mathematical model by a matrix calculation unit;

[0145] Step S33: The transformed discrete motor mathematical model is subjected to formula calculation using a numerical substitution method with the torque-speed model and the position-speed model to obtain a position model.

[0146] The step S32 comprises:

[0147] The matrix calculation unit is constructed by:

[0148] Encapsulate basic code blocks in a matrix calculation unit to perform basic operations on the matrix, the basic code blocks are retrieved from the Internet, the basic operations include matrix transposition, inverse matrix operation, determinant operation, unit row transformation, unit column transformation, block calculation and vector operation, and logically combine the basic code blocks according to calculation rules to obtain corresponding combined code blocks, the calculation rules are obtained from historical calculation records, the combined code blocks include distance calculation code blocks, dot product code blocks, angle code blocks, residual calculation code blocks and estimation code blocks, the distance calculation code block is used to calculate the distance value between two matrices, the dot product code block is used to calculate the value of the dot product of two matrices, the angle code block is used to calculate the three-dimensional angle and direction value between two matrices, the residual calculation code block is used to calculate the residual value between two matrices, and the estimation code block is used to calculate the position estimation value of the matrix;

[0149] The motor mathematical model obtained in step S1 is input into the matrix calculation unit, and the motor mathematical model is subjected to logical calculation of the combined code blocks to obtain a matrix model with the same dimension as the torque-speed model and the position-speed model.

[0150] The step S4 comprises the following sub-steps:

[0151] Step S41: defining a phase sequence variable according to the position model through a filter, wherein the phase sequence variable is obtained by calculating the phase difference and conforms to the Boolean operation formula of the series circuit;

[0152] Step S42: non-repetitively encode the reconstructed signal, and multiply the non-repetitively encoded reconstructed signal and the corresponding current signal by two by two with the phase sequence variable, and perform Fourier transform on the product to obtain a discrete multi-phase current set;

[0153] Step S43: calculating the fault characteristic quantity corresponding to each phase current according to the discrete multi-phase current set, and performing fault diagnosis according to the fault characteristic quantity;

[0154] Step S44: constructing a servo motor fault diagnosis system based on data model analysis, realizing real-time data interaction and online fault diagnosis functions, and completing visualization of the fault diagnosis process and results of the servo motor.

[0155] The step S43 includes the following sub-steps:

[0156] Step S431: using an array function to convert the phase current data included in the discrete multi-phase current set into a NumPy current array, respectively calculating the mean, variance and root mean square value of the NumPy current array, performing a spectrum analysis operation according to the mean, variance and root mean square value of the NumPy current array, and outputting a fault feature quantity, wherein the spectrum analysis operation includes calculating the frequency component of the NumPy current array by fast Fourier transform, obtaining a harmonic distortion degree quantity, and the harmonic distortion degree quantity is the fault feature quantity, and the fault feature quantity is used to describe potential faults in the current signal, wherein the mean reflects the DC component of the current signal, the variance describes the volatility of the signal, and the root mean square value reflects the effective value of the signal;

[0157] Step S432: collecting the servo motor state corresponding to the historical fault type, comparing the current feature quantity with the feature quantity under the normal state, and performing fault identification includes:

[0158] An unbalanced fault is an unbalanced current in the three phases, which may indicate uneven load;

[0159] Harmonic faults are high-order harmonics identified through spectrum analysis, which may indicate nonlinear loads or faults in the equipment;

[0160] A short circuit fault occurs when the current in one phase rises significantly and does not match that of other phases, which may indicate a short circuit;

[0161] Step S433: Classifying the fault according to the fault characteristic quantity includes:

[0162] Calculate the reliability value corresponding to each fault characteristic value respectively, the reliability value is the time proportion of the DC bus current corresponding phase within the step length h. In order to improve the reliability of fault diagnosis, it is necessary to improve the time proportion of the DC bus current corresponding phase;

[0163] When the reliability value is lower than a preset time threshold, the fault characteristic quantity is unreliable, and a compensation voltage vector is inserted to improve reliability;

[0164] When the reliability value is higher than the preset time threshold, the fault characteristic quantity is reliable, and a secondary diagnosis step is entered by judging the fault characteristic quantity, and the secondary diagnosis step includes:

[0165] When the fault characteristic quantity is greater than the time threshold, it means that the DC bus current sensor detects a fault, and the current signal phase corresponding to the fault characteristic quantity is the inspection phase of the DC bus current sensor. Each inspection phase represents a corresponding fault phase, and each fault phase corresponds to a fault diagnosis type. The fault diagnosis types include normal, short circuit, overload and unbalanced.

[0166] Through the above steps, we can extract fault characteristics from discrete multi-phase current signals and use these characteristics to perform fault diagnosis. This method can improve the reliability of equipment operation and detect potential faults in a timely manner.

[0167] Servo motor fault diagnosis system based on data model analysis, including function panel, calculation module and transmission module:

[0168] The function panel is used as a startup interface for visualizing fault diagnosis, and displays the fault diagnosis results obtained by real-time calculation in a coordinate system;

[0169] The calculation module is used to perform fault diagnosis on the servo motor using the servo motor fault diagnosis method based on data model analysis, because the slave thread in the servo motor fault diagnosis system based on data model analysis is responsible for receiving and analyzing the real-time data collected by the servo motor. In order to ensure the effect of fault diagnosis, the data of the servo motor collected in real time must meet the accurate sampling time and cannot have delay, so the calculation module is embedded in the main thread of the servo motor fault diagnosis system based on data model analysis;

[0170] The transmission module is used for:

[0171] After the slave thread of the servo motor fault diagnosis system based on data model analysis parses the real-time data of the servo motor, the parameter data required by the motor mathematical model, torque-speed model and position-speed model are sent to the calculation module located in the main thread with zero delay through signal broadcasting for fault diagnosis operation, the operation of the slave thread is interrupted by a preset timer, including sending an interrupt signal to the main thread at a unit time interval, and the fault diagnosis result output by the calculation module is sent to the coordinate system of the function panel;

[0172] After the calculation module completes the fault diagnosis calculation at the current moment, the diagnosis result is input into the memory space, and the above operation is repeated to realize the online fault diagnosis function;

[0173] The real-time data includes the current signal, dynamic data, load, effective radius, rotation speed, displacement and linear speed of the servo motor;

[0174] The parameter data includes real-time data, dependent variable parameters and constant term parameters.

[0175] The present invention can more accurately identify fault characteristics and reduce the possibility of misjudgment through in-depth analysis of real-time current signals. It combines motor mathematical models, torque-speed models and position-speed models to establish a multi-dimensional model to make fault diagnosis more comprehensive and systematic. It is applicable to servo motors of different types and specifications and provides solutions for various industrial applications. The reliability of fault identification is improved through classification and reliability evaluation of fault characteristic quantities. The function panel provides an intuitive display of fault diagnosis results to facilitate operator understanding and decision-making.

[0176] It should be understood by those skilled in the art that the embodiments of the present invention may be provided as methods, systems or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment or an embodiment combining software and hardware aspects. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media containing computer-usable program codes. Among them, the storage medium may be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (Static Random Access Memory, referred to as SRAM), electrically erasable programmable read-only memory (Electrically Erasable Programmable Read-Only Memory, referred to as EEPROM), erasable programmable read-only memory (Erasable Programmable Read Only Memory, referred to as EPROM), programmable read-only memory (Programmable Red-Only Memory, referred to as PROM), read-only memory (Read-Only Memory, referred to as ROM), magnetic memory, flash memory, magnetic disk or optical disk. These computer program instructions may also be stored in a computer readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0177] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A servo motor fault diagnosis method based on data model analysis, characterized in that: The following steps are involved: Step S1: Analyze the current performance of the servo motor by using a current reconstruction method, draw a current linear diagram of the filter, and construct a motor mathematical model of the servo motor in a synchronous rotating coordinate system according to the current linear diagram; Step S2: constructing a torque-speed model and a position-speed model of the servo motor based on a data-driven method according to Newton's laws and kinematic relationships; Step S3: combining the motor mathematical model, the torque-speed model and the position-speed model to obtain the position model of the servo motor; Step S4: using the filter to define a phase sequence variable according to the position model, combining the phase sequence variable with the reconstructed signal after non-repetitive encoding and the corresponding current signal and multiplying them to obtain a discrete multi-phase current set, and performing fault diagnosis on the servo motor according to the discrete multi-phase current set; The step S1 comprises the following sub-steps: Step S11: using a digital filter or an analog filter according to the control requirements of the servo motor, wherein the filter is used to smooth the current signal and remove high-frequency noise; Step S12: numbering the current signals to be input, the numbers are sorted according to the magnitudes of the current signals, and the sorted current signals are input into the filter for filtering by a current reconstruction method, and obtaining the reconstructed signal includes obtaining the dynamic response of the servo motor according to the motor control algorithm from the filtered signal, and the dynamic response is the reconstructed signal; Step S13: performing linear calculation on the reconstructed signal by fitting a linear equation using the least square method, obtaining a linear function of the relationship between the magnitude of the current signal and time, frequency and the reconstructed signal, and drawing the linear function using a drawing tool to obtain a current linear graph; Step S14: using a sensor to collect dynamic data of the servo motor, and calculating a motor mathematical model of the servo motor in a synchronous rotating coordinate system according to a dynamic equation and the current linear diagram; The step S4 comprises the following sub-steps: Step S41: defining a phase sequence variable according to the position model through a filter, wherein the phase sequence variable is obtained by calculating the phase difference and conforms to the Boolean operation formula of the series circuit; Step S42: non-repetitively encode the reconstructed signal, and multiply the non-repetitively encoded reconstructed signal and the corresponding current signal by two by two with the phase sequence variable, and perform Fourier transform on the product to obtain a discrete multi-phase current set; Step S43: calculating the fault characteristic quantity corresponding to each phase current according to the discrete multi-phase current set, and performing fault diagnosis according to the fault characteristic quantity; Step S44: constructing a servo motor fault diagnosis system based on data model analysis.

2. The servo motor fault diagnosis method based on data model analysis according to claim 1, characterized in that: The step S14 comprises: The dynamic data includes the moment of inertia, load torque, angular velocity, inductance, resistance and voltage applied to the servo motor of the servo motor; The mathematical expression of the kinetic equation is: ； in, is the moment of inertia, is the angular velocity, is the output torque, is the load torque, For time; The mathematical expression of output torque is: ； in, is the torque constant, is the current signal; The relationship between the current and voltage of the servo motor is obtained by the electrical equation, and the mathematical expression of the electrical equation is: ； in, is the voltage applied to the servo motor, is the inductor, is the resistance, is the electromotive force constant; In the synchronous rotating coordinate system, the current signal of the servo motor is converted into a coordinate system synchronized with the rotor of the servo motor. The mathematical expression of the converted current signal coordinate system is: ； ； in, It is the original format data of the current signal. It is the format data of the current signal in the coordinate system synchronized with the rotor of the servo motor; The mathematical expression for transforming the torque coordinate system is: ； According to the dynamic equation, the mathematical expression of the motor output torque, the mathematical expression of the converted current signal coordinate system and the mathematical expression of the converted torque coordinate system, the state space relationship between the moment of inertia, angular velocity, load torque, inductance, resistance and the voltage applied to the servo motor and the current signal is obtained, and the mathematical expression of the motor mathematical model for obtaining the state space relationship is: ； ； in, is the current signal, is the angular velocity corresponding to the current signal, Output of the motor mathematical model.

3. The servo motor fault diagnosis method based on data model analysis as claimed in claim 2, characterized in that: The step S2 comprises: Constructing the torque-velocity model includes: The corresponding load, effective radius and speed obtained by the sensor experimental test of the servo motor according to the numbered current signal are collected. The mathematical expression related to the output torque, speed and acceleration is: ； ； in, is the acceleration; The kinematic relationship of the servo motor is described by the relationship between speed and displacement, and its mathematical expression is: ； in, is the linear speed, is the effective radius; The relationship between the torque and the speed is established according to the mathematical expression of the relationship between the output torque, the speed and the acceleration, the mathematical expression of the kinematic relationship of the servo motor and the polynomial regression method, including: The established polynomial regression mathematical expression is: ； in, and is the dependent variable parameter, is the order of the polynomial; Substitute the mathematical expression relating the output torque to the rotational speed and acceleration and the mathematical expression of the kinematic relationship of the servo motor into the polynomial regression mathematical expression respectively to obtain a torque-speed model.

4. The servo motor fault diagnosis method based on data model analysis as claimed in claim 3, characterized in that: Constructing the position-velocity model includes: The corresponding displacement and linear velocity of the servo motor obtained by the sensor experimental test according to the numbered current signal are collected, and the relationship between position and velocity is established by using a linear regression model, and its mathematical expression is: ； in, k and c is the constant term parameter of the linear regression model; The collected displacement and linear velocity are substituted into the linear regression model for training to obtain a position-velocity model.

5. The servo motor fault diagnosis method based on data model analysis as claimed in claim 4, characterized in that: The step S3 comprises the following sub-steps: Step S31: discretizing the motor mathematical model obtained in step S1, including convolving and calculating the motor mathematical model with a discrete phase matrix to obtain a discrete motor mathematical model; Step S32: transforming the discrete motor mathematical model by a matrix calculation unit; Step S33: The transformed discrete motor mathematical model is subjected to formula calculation using a numerical substitution method with the torque-speed model and the position-speed model to obtain a position model.

6. The servo motor fault diagnosis method based on data model analysis according to claim 5, characterized in that: The step S32 comprises: The matrix calculation unit is constructed by: Encapsulate basic code blocks in a matrix calculation unit to perform basic operations on the matrix, the basic code blocks are retrieved from the Internet, the basic operations include matrix transposition, inverse matrix operation, determinant operation, unit row transformation, unit column transformation, block calculation and vector operation, and logically combine the basic code blocks according to calculation rules to obtain corresponding combined code blocks, the calculation rules are obtained from historical calculation records, the combined code blocks include distance calculation code blocks, dot product code blocks, angle code blocks, residual calculation code blocks and estimation code blocks, the distance calculation code block is used to calculate the distance value between two matrices, the dot product code block is used to calculate the value of the dot product of two matrices, the angle code block is used to calculate the three-dimensional angle and direction value between two matrices, the residual calculation code block is used to calculate the residual value between two matrices, and the estimation code block is used to calculate the position estimation value of the matrix; The motor mathematical model obtained in step S1 is input into the matrix calculation unit, and the motor mathematical model is subjected to logical calculation of the combined code blocks to obtain a matrix model with the same dimension as the torque-speed model and the position-speed model.

7. The servo motor fault diagnosis method based on data model analysis according to claim 6, characterized in that: The step S43 includes the following sub-steps: Step S431: using an array function to convert the phase current data included in the discrete multi-phase current set into a NumPy current array, respectively calculating the mean, variance and root mean square value of the NumPy current array, performing a spectrum analysis operation according to the mean, variance and root mean square value of the NumPy current array, and outputting a fault feature quantity, wherein the spectrum analysis operation includes calculating the frequency component of the NumPy current array by fast Fourier transform, and obtaining a harmonic distortion degree quantity, wherein the harmonic distortion degree quantity is the fault feature quantity; Step S432: collecting the servo motor state corresponding to the historical fault type, comparing the current feature quantity with the feature quantity under the normal state, and performing fault identification includes: Unbalanced fault is the unbalanced current of three phases; Harmonic faults are high-order harmonics identified through spectrum analysis; A short circuit fault occurs when the current in one phase rises significantly and does not match the other phases; Step S433: Classifying the fault according to the fault characteristic quantity includes: Calculate the reliability value corresponding to each fault characteristic value respectively, where the reliability value is the time proportion of the corresponding phase of the DC bus current within the time step h; When the reliability value is lower than a preset time threshold, the fault characteristic quantity is unreliable, and a compensation voltage vector is inserted to improve reliability; When the reliability value is higher than the preset time threshold, the fault characteristic quantity is reliable, and a secondary diagnosis step is entered by judging the fault characteristic quantity, and the secondary diagnosis step includes: When the fault characteristic quantity is greater than the time threshold, it means that the DC bus current sensor detects a fault, and the current signal phase corresponding to the fault characteristic quantity is the inspection phase of the DC bus current sensor. Each inspection phase represents a corresponding fault phase, and each fault phase corresponds to a fault diagnosis type. The fault diagnosis types include normal, short circuit, overload and unbalanced.

8. A servo motor fault diagnosis system based on data model analysis, characterized in that: Including function panel, calculation module and transmission module: The function panel is used as a startup interface for visualizing fault diagnosis, and displays the fault diagnosis results obtained by real-time calculation in a coordinate system; The calculation module is used to perform fault diagnosis on the servo motor using the servo motor fault diagnosis method based on data model analysis according to claim 1, and the calculation module is embedded in the main thread of the servo motor fault diagnosis system based on data model analysis; The transmission module is used for: After the slave thread of the servo motor fault diagnosis system based on data model analysis parses the real-time data of the servo motor, the parameter data required by the motor mathematical model, the torque-speed model and the position-speed model are sent to the calculation module located in the main thread with zero delay through signal broadcasting for fault diagnosis operation, the operation of the slave thread is interrupted by a preset timer, including sending an interrupt signal to the main thread at a unit time interval, and the fault diagnosis result output by the calculation module is sent to the coordinate system of the function panel; After the calculation module completes the fault diagnosis calculation at the current moment, the diagnosis result is input into the memory space; The real-time data includes the current signal, dynamic data, load, effective radius, rotation speed, displacement and linear speed of the servo motor; The parameter data includes the real-time data, dependent variable parameters and constant term parameters.

Images