Signal separation and fault diagnosis method for permanent magnet synchronous motor and transmission system

Abstract

The invention discloses a signal separation and fault diagnosis method for a permanent magnet synchronous motor and a transmission system, and relates to the technical field of motor control and signal processing, and the method comprises the following steps: building an experiment platform, and collecting a rotating speed signal, a dq axis current signal and a rotor angle signal; analyzing a harmonic transfer mechanism, deducing a closed-loop transfer function, and reversely deducing a resistance torque signal based on an electromagnetic torque model; inputting a transfer function to obtain a transmission system related rotation speed component omega tr; the method has the beneficial effects that high-precision separation of the signals is realized, the fault positioning efficiency is improved (the diagnosis time is shortened), an additional vibration sensor is not needed, the cost is reduced, the system robustness is enhanced, and the method is suitable for large-scale popularization and application. The method is suitable for industrial automation and intelligent manufacturing fields.

Description

Technical Field

[0001] This invention relates to the field of motor control and signal processing technology, and more specifically, to a method for signal separation and fault diagnosis based on a permanent magnet synchronous motor (PMSM) and its transmission system. Background Technology

[0002] Permanent magnet synchronous motors (PMSMs) are widely used in industrial, transportation, and home appliance fields due to their advantages such as high efficiency, low noise, and small size. However, during motor operation, they are usually coupled with a transmission system, such as ball screws or gearboxes. This coupling complicates fault diagnosis and dynamic characteristic analysis because the signals from the motor and the transmission system are superimposed, making it difficult to separate their individual contributions.

[0003] Existing research on signal separation between motors and transmission systems has the following problems:

[0004] Insufficient signal separation accuracy: Traditional methods typically rely on simplified dynamic models or frequency domain filtering, which cannot accurately separate the coupled signals between the motor and the transmission system, especially under complex load conditions, resulting in low reliability of the separation results. Limited dynamic characteristic analysis: Existing research focuses primarily on the transfer function of a single component, failing to comprehensively describe the dynamic behavior within the motor and lacking in-depth research on the closed-loop characteristics of the motor and transmission system. Low fault diagnosis efficiency: Due to the ineffective separation of signals from the motor and transmission system, the extraction and location of fault features are inefficient, leading to inaccurate diagnostic results that fail to meet practical engineering needs. Limitations of load observation techniques: Load torque serves as a crucial reference signal for signal separation and fault diagnosis. Traditional methods often employ static estimation or simple filtering algorithms, which are ill-suited to handling dynamic load changes and complex disturbance environments.

[0005] In recent years, high-frequency harmonic injection technology and load torque observation methods based on state-space theory have been introduced into the field of motors. These technologies provide new ideas for solving signal separation and fault diagnosis problems, but there is still no mature solution on how to effectively combine and apply them to the coupling scenario of permanent magnet synchronous motors and drive systems.

[0006] This invention proposes an innovative solution to the above problems. By combining a load torque observer and a high-frequency harmonic injection algorithm, high-precision separation of motor and drive system signals is achieved; by obtaining the dynamic transfer function, the dynamic characteristics of the motor are comprehensively characterized; and by optimizing fault feature extraction and diagnostic algorithms, diagnostic efficiency and accuracy are significantly improved. This solution provides a completely new technical means for performance optimization and fault diagnosis of complex motor systems. Summary of the Invention

[0007] This invention provides a fault diagnosis method for permanent magnet synchronous motors and transmission systems based on signal separation, aiming to solve the problem of low fault diagnosis accuracy caused by signal coupling between the motor and transmission system in existing technologies. This method is based on the control mechanism of permanent magnet synchronous motors, combined with high-frequency signal injection technology and load torque observer design, to achieve precise separation of motor and transmission system signals, and completes fault diagnosis through feature extraction and analysis.

[0008] Through the above method, the present invention can not only accurately distinguish the signal characteristics of permanent magnet synchronous motors and transmission systems, but also effectively improve the sensitivity and accuracy of fault diagnosis, and is especially suitable for health status monitoring and maintenance of complex mechanical systems.

[0009] A method for signal separation and fault diagnosis of a permanent magnet synchronous motor and its transmission system is characterized by separating the original speed signal into a transmission system speed signal and a motor-related speed signal. These two signals eliminate coupling information, thus further narrowing down the fault range. The method includes the following steps:

[0010] Step 1: After assembling the motor and transmission system, build the experimental platform, including a permanent magnet synchronous motor (PMSM), transmission components (ball screw), and measuring devices. High-precision sensors are used to collect signals generated during the operation of the motor and transmission system in real time, mainly including motor speed ω, dq-axis currents id and iq, and rotor angle θ. r These signals serve as input for subsequent analysis, providing fundamental data for signal separation and fault diagnosis.

[0011] Step Two: First, the harmonic transmission mechanism of the electromechanical system is analyzed, and the coupling relationship and frequency characteristics between the speed signal and the resistance torque signal are studied. For components in the electromechanical system that can be described by theoretical formulas, the Laplace transform method is used to derive the component transfer function; for complex nonlinear components, the component transfer function is calculated by inputting a known harmonic excitation signal, measuring the system's output response, and combining the frequency characteristics. Finally, the closed-loop transfer function H from the motor shaft torque T to the speed is derived. T_ω (s);

[0012] Step 3: Based on the collected motor current id, iq and rotor angle θ r The electromagnetic torque response signal Te is calculated by inputting it into the motor magnetic field model. Then, using the known electromagnetic torque Te and motor speed ω, the resistance torque signal Tr on the motor shaft is derived.

[0013] Step 4: Input the calculated resistance torque signal Tr into the derived closed-loop transfer function H. T_ω In (s), the signal component ω that indicates the influence of the transmission system on the motor speed is obtained. trThis signal component reflects the operating state and fault characteristics of the transmission system. That is, ω tr (s)=H T_ω (s)·T r (s).

[0014] Step 5: Subtract the transmission system-related speed signal ω from the actual measured motor speed signal ω. tr This yields the speed signal ω, which is only related to the motor. m This separation process eliminates the coupling between the motor and drive system signals, achieving signal separation between the two.

[0015] Step Six: Separate the signal ω tr and ω m Frequency domain analysis was performed to observe the amplitude differences of signals with the same frequency and the spectral characteristics of signals with different frequencies. The analysis results can verify the effectiveness of signal separation and provide a basis for identifying fault characteristics of transmission systems and motors.

[0016] Step 7: Based on the transmission system signal ω tr and motor signal ω m The method diagnoses potential faults. Experiments verify the accuracy and practicality of the separation algorithm, and further improve the diagnostic effect by optimizing model parameters, ensuring the reliability and robustness of the method in real-world applications.

[0017] In the above scheme, the method for obtaining the transfer function in step two is as follows:

[0018] First, let the system fundamental frequency angle be θ. base The target harmonic frequency is a multiple of the fundamental frequency, k·θ base The target harmonic frequency θ is extracted using angular domain single-point Fourier transform (SPFT). inj =k·θ base A pure real harmonic signal I is injected into the input of the target loop. inj,real =Acos(k·θ) base ), where A is the amplitude of the injected signal, and the response of the system output signal is recorded. The real and imaginary parts of the response at the target frequency are extracted by SPFT and labeled Re0 and Im0, respectively, to measure the dynamic response of the system to real disturbances. The signal components are recorded, and their real and imaginary parts are denoted as Re0 and Im0, respectively. These initial values ​​serve as the reference for subsequent injected signals.

[0019] The second step is to inject a pure real harmonic signal I into the input of the target loop. inj,real =Acos(k·θ) base), where A is the amplitude of the injected signal, and the response of the system output signal is recorded. The real and imaginary responses of the target frequency are extracted by SPFT and labeled as Re1 and Im1, respectively, to measure the dynamic response of the system to real disturbances.

[0020] The third step is to inject a pure imaginary harmonic signal I into the input of the target loop. inj,real =Asin(k·θ) base ), where A is the amplitude of the injected signal, and the response of the system output signal is recorded. The real and imaginary responses of the target frequency are extracted by SPFT and labeled as Re3 and Im3, respectively, to measure the dynamic response of the system to the imaginary disturbance.

[0021] The fourth step involves calculating the transfer function matrix H of the target loop based on the initial signal (Re0, Im0), and the responses (Re1, Im1) and (Re3, Im3) after the injection of the real and imaginary parts of the signal, respectively. Each element of the matrix represents the gain of the real and imaginary parts of the input signal on the real and imaginary parts of the output signal, respectively.

[0022]

[0023] Step 5: Following the above method, conduct experiments on the target loops, including speed loop current to current loop current, speed loop current to speed, q-axis voltage to current loop current, and q-axis voltage to speed, and obtain the corresponding transfer function matrices. The transfer function from speed loop current to current loop current is H. Iq2_Iq (s), the transfer function from the speed loop current to the speed is H. Iq2_ω (s), the q-axis voltage to current loop current transfer function is H Uq_Iq (s), the q-axis voltage to speed transfer function is H Uq_ω (s), gradually establish a complete system dynamic characteristic model.

[0024] The sixth step involves utilizing the dynamic characteristics of the FOC control system and combining the obtained transfer function matrices of each loop to derive the load torque Tr to the relevant speed ω of the transmission system. tr Transfer function:

[0025]

[0026] The characteristics of the target harmonic multiple are determined so that the transfer function describes the dynamic behavior of a specific harmonic.

[0027] Compared with the prior art, the present invention has the following advantages:

[0028] 1. This invention achieves precise signal separation between permanent magnet synchronous motor and transmission system through high-frequency harmonic injection technology and dynamic transfer function analysis. It can effectively cope with nonlinear coupling of signals under complex working conditions and significantly improve the accuracy and reliability of signal separation.

[0029] 2. By analyzing the separated harmonic frequencies, this invention can quickly identify the source of harmonic anomalies and clearly determine whether the fault is in the motor or the transmission system, thereby significantly shortening the fault location time and enhancing the accuracy and efficiency of diagnosis.

[0030] 3. This invention does not require the installation of expensive vibration sensors. Fault diagnosis can be completed solely based on motor operating signals (such as speed, current, torque, etc.), which not only reduces system hardware costs but also reduces maintenance complexity and adapts to diverse dynamic working conditions and industrial site requirements. Attached Figure Description

[0031] Figure 1 Flowchart for separating the speed signal into transmission system-related signals and motor-related signals;

[0032] Figure 2 , Figure 3 To obtain the transfer function graph for high-frequency harmonic injection;

[0033] Figure 4 To obtain the block diagram of the torque-to-speed transfer function;

[0034] Figure 5 The diagram shows the harmonic frequencies separated by rotational speed. Detailed Implementation

[0035] The concept, specific structure and technical effects of the present invention will be further described below in conjunction with the accompanying drawings to fully understand the purpose, characteristics and effects of the present invention.

[0036] This invention provides a method for signal separation and fault diagnosis of a permanent magnet synchronous motor and its transmission system, see reference. Figure 2 As shown, the operating principle of this system is as follows:

[0037] The permanent magnet synchronous motor is fixed on the ball screw slide, and the ball screw is used as part of the transmission system. The permanent magnet synchronous motor and the ball screw are connected in sequence through the controller (DSP) to form a complete signal acquisition system.

[0038] Example:

[0039] like Figures 1-5 As shown in the embodiment, this is a method for signal separation and fault diagnosis of a permanent magnet synchronous motor and its transmission system. The system operates as follows:

[0040] Step 1: After assembling the motor and transmission system, build the experimental platform, including a permanent magnet synchronous motor (PMSM), transmission components (ball screw), and measuring devices. High-precision sensors are used to collect signals generated during the operation of the motor and transmission system in real time, mainly including motor speed v, dq-axis currents id and iq, and rotor angle θ. r These signals serve as input for subsequent analysis, providing fundamental data for signal separation and fault diagnosis.

[0041] Step Two: First, the harmonic transmission mechanism of the electromechanical system is analyzed, and the coupling relationship and frequency characteristics between the speed signal and the resistance torque signal are studied. For components in the electromechanical system that can be described by theoretical formulas, the Laplace transform method is used to derive the component transfer function; for complex nonlinear components, the component transfer function is calculated by inputting a known harmonic excitation signal, measuring the system's output response, and combining the frequency characteristics. Finally, the closed-loop transfer function H from the motor shaft torque T to the speed is derived. T_ω (s);

[0042] The method for obtaining the transfer function is as follows:

[0043] First, let the system fundamental frequency angle be θ. base The target harmonic frequency is a multiple of the fundamental frequency, k·θ base The target harmonic frequency θ is extracted using angular domain single-point Fourier transform (SPFT). inj =k·θ base A pure real harmonic signal I is injected into the input of the target loop. inj,real =Acos(k·θ) base ), where A is the amplitude of the injected signal, and the response of the system output signal is recorded. The real and imaginary parts of the response at the target frequency are extracted by SPFT and labeled Re0 and Im0, respectively, to measure the dynamic response of the system to real disturbances. The signal components are recorded, and their real and imaginary parts are denoted as Re0 and Im0, respectively. These initial values ​​serve as the reference for subsequent injected signals.

[0044] The second step is to inject a pure real harmonic signal I into the input of the target loop. inj,real =Acos(k·θ) base ), where A is the amplitude of the injected signal, and the response of the system output signal is recorded. The real and imaginary responses of the target frequency are extracted by SPFT and labeled as Re1 and Im1, respectively, to measure the dynamic response of the system to real disturbances.

[0045] The third step is to inject a pure imaginary harmonic signal I into the input of the target loop. inj,real =Asin(k·θ) base), where A is the amplitude of the injected signal, and the response of the system output signal is recorded. The real and imaginary responses of the target frequency are extracted by SPFT and labeled as Re3 and Im3, respectively, to measure the dynamic response of the system to the imaginary disturbance.

[0046] The fourth step involves calculating the transfer function matrix H of the target loop based on the initial signal (Re0, Im0), and the responses (Re1, Im1) and (Re3, Im3) after the injection of the real and imaginary parts of the signal, respectively. Each element of the matrix represents the gain of the real and imaginary parts of the input signal on the real and imaginary parts of the output signal, respectively.

[0047]

[0048] Step 5: Following the above method, conduct experiments on the target loops, including speed loop current to current loop current, speed loop current to speed, q-axis voltage to current loop current, and q-axis voltage to speed, and obtain the corresponding transfer function matrices. The transfer function from speed loop current to current loop current is H. Iq2_Iq (s), the transfer function from the speed loop current to the speed is H. Iq2_ω (s), the q-axis voltage to current loop current transfer function is H Uq_Iq (s), the q-axis voltage to speed transfer function is H Uq_ω (s), gradually establish a complete system dynamic characteristic model.

[0049] The sixth step involves utilizing the dynamic characteristics of the FOC control system and combining the obtained transfer function matrices of each loop to derive the load torque Tr to the relevant speed ω of the transmission system. tr Transfer function:

[0050]

[0051] The characteristics of the target harmonic multiple are determined so that the transfer function describes the dynamic behavior of a specific harmonic.

[0052] Step 3: Based on the collected motor current id, iq and rotor angle θ r The electromagnetic torque response signal Te is calculated by inputting it into the motor magnetic field model. Then, using the known electromagnetic torque Te and motor speed ω, the resistance torque signal Tr on the motor shaft is derived.

[0053] Step 4: Input the calculated resistance torque signal Tr into the derived closed-loop transfer function H. T_ω In (s), the signal component ω that indicates the influence of the transmission system on the motor speed is obtained. tr This signal component reflects the operating state and fault characteristics of the transmission system. That is, ω tr (s)=H T_ω (s)·T r(s).

[0054] Step 5: Subtract the transmission system-related speed signal ω from the actual measured motor speed signal ω. tr This yields the speed signal ω, which is only related to the motor. m This separation process eliminates the coupling between the motor and drive system signals, achieving signal separation between the two.

[0055] Step Six: Separate the signal ω tr and ω m Frequency domain analysis was performed to observe the amplitude differences of signals with the same frequency and the spectral characteristics of signals with different frequencies. The analysis results can verify the effectiveness of signal separation and provide a basis for identifying fault characteristics of transmission systems and motors.

[0056] Step 7: Based on the transmission system signal ω tr and motor signal ω m The method diagnoses potential faults. Experiments verify the accuracy and practicality of the separation algorithm, and further improve the diagnostic effect by optimizing model parameters, ensuring the reliability and robustness of the method in real-world applications.

[0057] This method separates the signals from the motor and transmission system, and combines this with harmonic frequency analysis to accurately determine the source of faults in the motor and transmission system, thus replacing traditional fault diagnosis methods based on vibration sensors. In contrast, this method eliminates the need for expensive vibration sensors, reducing system costs, while improving the accuracy and efficiency of fault diagnosis. It is applicable to fields such as industrial automation, intelligent manufacturing, and high-precision mechanical control.

[0058] While preferred embodiments of the present invention have been described above in conjunction with the accompanying drawings, the present invention should not be limited to structures and operations that are exactly the same as those described above and shown in the drawings. Those skilled in the art can make many equivalent improvements and variations to the above embodiments through logical analysis, reasoning, or limited experiments without departing from the concept and scope of the present invention, but all such improvements and variations should fall within the scope of protection claimed by the present invention.

Claims

1. A method for signal separation and fault diagnosis of a permanent magnet synchronous motor and its transmission system, characterized in that... The original speed signal is separated into a transmission system speed signal and a motor-related speed signal. These two signals eliminate coupling information, thus further narrowing down the fault range. This method includes the following steps: Step 1: After assembling the motor and transmission system, build an experimental platform, including a permanent magnet synchronous motor, transmission components, and measuring devices. Use high-precision sensors to collect signals generated during the operation of the motor and transmission system in real time, mainly including motor speed ω, dq-axis currents id and iq, and rotor angle θ. r These signals serve as input for subsequent analysis, providing fundamental data for signal separation and fault diagnosis. Step Two: First, the harmonic transmission mechanism of the electromechanical system is analyzed, and the coupling relationship and frequency characteristics between the speed signal and the resistance torque signal are studied. For components in the electromechanical system that can be described by theoretical formulas, the Laplace transform method is used to derive the component transfer function. For complex nonlinear components, the known harmonic excitation signal is input, the system output response is measured, and the component transfer function is calculated by combining the frequency characteristics. Finally, the closed-loop transfer function H from the motor shaft torque T to the speed is derived. T_ω (s); Step 3: Based on the collected motor current id, iq and rotor angle θ r The electromagnetic torque response signal Te is calculated by inputting it into the motor magnetic field model. Then, combining this signal with the motor dynamics equations, the resistance torque signal Tr on the motor shaft is derived from the known electromagnetic torque Te and motor speed ω. Step 4: Input the calculated resistance torque signal Tr into the derived closed-loop transfer function H. T_ω In (s), the signal component ω that indicates the influence of the transmission system on the motor speed is obtained. tr This signal component reflects the operating state and fault characteristics of the transmission system, i.e., ω tr (s)=H T_ω (s)·T r (s); Step 5: Subtract the transmission system-related speed signal ω from the actual measured motor speed signal ω. tr This yields the speed signal ω, which is only related to the motor. m This separation process eliminates the coupling between the signals of the motor and the transmission system, thus achieving the separation of their signals; Step Six: Separate the signal ω tr and ω m Frequency domain analysis is performed to observe the amplitude differences of signals with the same frequency and the spectral characteristics of signals with different frequencies. The analysis results can verify the effectiveness of signal separation and provide a basis for identifying fault characteristics of transmission systems and motors. Step 7: Based on the transmission system signal ω tr and motor signal ω m The method diagnoses potential faults, verifies the accuracy and practicality of the separation algorithm through experiments, and further improves the diagnostic effect by optimizing model parameters, ensuring the reliability and robustness of the method in practical applications.

2. In the method for signal separation and fault diagnosis of a permanent magnet synchronous motor and transmission system according to claim 1, step two will use a high-frequency harmonic injection algorithm to obtain the corresponding transfer function, the principle of which is as follows: First, let the system fundamental frequency angle be θ. base The target harmonic frequency is a multiple of the fundamental frequency, k·θ base The target harmonic frequency θ is extracted by using angular domain single-point Fourier transform (SPFT). inj =k·θ base A pure real harmonic signal I is injected into the input of the target loop. inj,real =Acos(k·θ) base ), where A is the amplitude of the injected signal, the response of the system output signal is recorded, the real and imaginary responses of the target frequency are extracted by SPFT and marked as Re0 and Im0 respectively, in order to measure the dynamic response of the system to the real disturbance, the signal components, the real and imaginary parts are recorded and marked as Re0 and Im0 respectively, these initial values ​​are the reference for subsequent injected signals; The second step is to inject a pure real harmonic signal I into the input of the target loop. inj,real =Acos(k·θ) base ), where A is the amplitude of the injected signal, the response of the system output signal is recorded, and the real and imaginary responses of the target frequency are extracted by SPFT and labeled as Re1 and Im1, respectively, to measure the dynamic response of the system to the real disturbance; The third step is to inject a pure imaginary harmonic signal I into the input of the target loop. inj,real =Asin(k·θ) base ), where A is the amplitude of the injected signal, the response of the system output signal is recorded, and the real and imaginary responses of the target frequency are extracted by SPFT and labeled as Re3 and Im3, respectively, to measure the dynamic response of the system to the imaginary disturbance; The fourth step is to calculate the transfer function matrix H of the target loop based on the initial signal (Re0, Im0), the response (Re1, Im1) and (Re3, Im3) after injecting the real and imaginary parts of the signal, respectively. Each element of the matrix represents the gain of the real and imaginary parts of the input signal on the real and imaginary parts of the output signal. Step 5: Following the above method, conduct experiments on the target loops, including speed loop current to current loop current, speed loop current to speed, q-axis voltage to current loop current, and q-axis voltage to speed, and obtain the corresponding transfer function matrices. The transfer function from speed loop current to current loop current is H. Iq2_Iq (s), the transfer function from the speed loop current to the speed is H. Iq2_ω (s), the q-axis voltage to current loop current transfer function is H Uq_Iq (s), the q-axis voltage to speed transfer function is H Uq_ω (s), gradually establish a complete system dynamic characteristic model; The sixth step involves utilizing the dynamic characteristics of the FOC control system and combining the obtained transfer function matrices of each loop to derive the load torque Tr to the relevant speed ω of the transmission system. tr Transfer function: The characteristics of the target harmonic multiple are determined so that the transfer function describes the dynamic behavior of a specific harmonic.

Images