Motor three-phase unbalance analysis method based on instantaneous current information

Abstract

A kind of motor three-phase unbalance analysis method based on current instantaneous information, first obtain three-phase current signal, pass through vector demodulation to construct phase shift signal to obtain current signal instantaneous frequency and instantaneous amplitude, calculate its variation coefficient, compare and analyze the instantaneous information characteristic coefficient performance under three-phase unbalance;Actual experiment analysis under different rotating speed, compare and analyze the variation coefficient characteristics of instantaneous frequency and instantaneous amplitude of normal motor and fault motor after direct calculation and VMD extraction trend term, it can be found that the instantaneous frequency variation coefficient after VMD extraction trend term can obviously distinguish fault motor, the present application can be used as effective means for analyzing motor three-phase unbalance fault, timely find motor fault, to reduce production, time loss.

Description

Technical Field

[0001] The present invention belongs to the technical field of motor fault diagnosis, and in particular relates to a motor three-phase imbalance analysis method based on current instantaneous information. Background Art

[0002] Electric motors are widely used in industry and can be divided into DC motors and AC motors according to the type of power supply. Three-phase induction motors are mainly used in CNC machine tools, medical equipment, computer peripherals and other fields. For power grid systems, three-phase balance mainly refers to the equality of the voltage vectors of the three phases. Three-phase imbalance refers to the inconsistency of the three-phase current / voltage amplitudes in the power system, the phase difference is not 120°, and the difference value exceeds the specified range. Three-phase imbalance is a common fault caused by various reasons, such as broken wire faults and ground faults. Excessive imbalance can lead to operational instability of electrical equipment, resulting in abnormal speed, heat generation and other faults. Research on three-phase imbalance detection methods is of great significance to improving the operational reliability of industrial equipment.

[0003] Common methods for analyzing three-phase motor imbalance include fast Fourier transform (FFT) analysis and symmetrical component analysis. FFT is suitable for steady-state, periodic signals, but using it for non-steady-state, non-periodic signals can result in spectral aliasing and leakage. The symmetrical component method decomposes the three-phase voltage / current into symmetrical positive-sequence, negative-sequence, and zero-sequence components. Imbalance is calculated by vector calculation using the negative-sequence and zero-sequence components. Due to the superposition principle, the symmetrical component method is only applicable to circuits with linear parameters. Summary of the Invention

[0004] In order to overcome the above-mentioned shortcomings of the prior art, the purpose of the present invention is to provide a motor three-phase imbalance analysis method based on current instantaneous information, which can detect motor faults in time to reduce production and time losses.

[0005] To achieve the above objectives, the technical solution adopted by the present invention is:

[0006] A method for analyzing three-phase imbalance of a motor based on instantaneous current information comprises the following steps:

[0007] 1) Extract the instantaneous frequency and instantaneous amplitude of the three-phase current through Park vector demodulation;

[0008] 2) Extract the trend terms of instantaneous frequency and instantaneous amplitude through VMD;

[0009] 3) Calculate the coefficient of variation of instantaneous frequency and instantaneous amplitude, compare and analyze the characteristics of the coefficient of variation of instantaneous frequency and instantaneous amplitude of normal motor and faulty motor directly calculated and VMD extracted trend items, and perform motor fault diagnosis.

[0010] In the step 1) Park vector demodulation, the speed calculation method of the three-phase asynchronous motor is shown in the following formula:

[0011]

[0012] Where n r is the speed (r / min), s is the slip ratio, p is the number of pole pairs, and f is the rotation frequency;

[0013] Signal demodulation is accomplished by using the Park transform to create a 90° phase-shifted signal, as shown below:

[0014]

[0015] Where I α and I β is the stator coordinate system current, I u , I v and I w They are the three-phase currents of the motor U, V, and W respectively;

[0016] The three-phase current is expressed as:

[0017]

[0018] Where f M is the signal fundamental frequency, a is the signal amplitude; combining equations (2) and (3), we can get:

[0019]

[0020] byI α and I β Construct the analytical signal as shown in formula (5):

[0021] I z =I α +I β ·i (5)

[0022] Where I z is the constructed complex signal;

[0023] Calculate the modulus and phase angle of the constructed complex signal to get the instantaneous amplitude and instantaneous phase. z Finding the modulus and argument yields:

[0024]

[0025] Where a(t) is the instantaneous amplitude, is the instantaneous phase;

[0026] The instantaneous frequency is the inverse of the instantaneous phase, that is:

[0027]

[0028] Where f(t) is the instantaneous frequency;

[0029] The instantaneous amplitude and instantaneous frequency are the instantaneous information of the three-phase current signal multi-sensor information fusion, reflecting the comprehensive characteristics of the three-phase current.

[0030] In step 2) the VMD method is used to obtain the modal bandwidth, and the steps are as follows:

[0031] The related analytical signal is calculated by Hilbert transform to obtain a one-sided spectrum;

[0032] For each mode, the spectrum of the mode is shifted to “baseband” by exponential mixing tuned to the corresponding estimated center frequency;

[0033] The bandwidth is estimated by h Gaussian smoothing of the gradient, and the resulting constrained variational problem is as follows;

[0034]

[0035] In the formula, {u k} is a set of k intrinsic mode functions, {ω k} is a set of k center frequencies, δ(t) is the impulse function, H is the original signal, t is the time scale, and j is the imaginary unit;

[0036] Through the Lagrange multiplier and the quadratic penalty term, the constrained objective function is transformed into an unconstrained objective function, and the augmented Lagrangian function is obtained, that is:

[0037]

[0038] Where α is the balance coefficient, λ is the Lagrange multiplier, {u k} is a set of k intrinsic mode functions, {ω k} is a set of k center frequencies, δ(t) is the impulse function, H(t) is the original signal, t is time, and j is the imaginary unit;

[0039] The VMD method is used to extract the imfs of multiple characteristic frequencies, and the trend term is obtained by calculating the maximum mean term.

[0040] The coefficient of variation in step 3) is calculated as:

[0041] The coefficient of variation is the ratio of the standard deviation to the mean, which is used to reflect the degree of variation of each observation value and is a dimensionless statistic.

[0042] The mean μ is expressed as follows:

[0043]

[0044] Where xi is the i-th data, n is the data length;

[0045] The sample standard deviation σ is expressed as follows:

[0046]

[0047] According to the definition of coefficient of variation, combining equations (10) and (11), we can get the coefficient of variation c v ;

[0048]

[0049] The coefficient of variation is only defined when the mean is not zero and is used when the mean is greater than zero. The coefficient of variation measures the fluctuation of the overall data and analyzes its fluctuation state by calculating the coefficient of variation of the instantaneous frequency and instantaneous amplitude.

[0050] The beneficial effects of the present invention are:

[0051] Because the present invention converts three-phase signals to obtain instantaneous information features, it enables a comprehensive and intuitive analysis of imbalance conditions. The difference in the coefficient of variation is used to characterize the degree of three-phase imbalance between a normal motor and a faulty motor. The coefficient of variation can also be used to measure overall data fluctuations. By calculating the coefficient of variation of the instantaneous frequency and amplitude, their fluctuations can be analyzed.

[0052] To overcome the problem of similar coefficients of variation of the instantaneous frequencies of faulty and normal motors, the present invention uses VMD decomposition to extract the coefficient of variation calculated from the trend term. The characteristic value of the faulty motor is higher than that of the normal motor. Compared to the undecomposed signal, the instantaneous speed can be more clearly distinguished. The instantaneous frequency coefficient of variation of the present invention is considered as a characteristic coefficient for determining three-phase imbalance, enabling timely detection of motor faults and reducing production and time losses. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 It is a flow chart of the present invention.

[0054] Figure 2 This is a schematic diagram of a comprehensive gearbox dynamics test bench constructed according to an embodiment of the present invention, wherein (a) is a diagram of the test bench structure, (b) is a diagram of the actual current sensor clamping device, and (c) is a schematic diagram of the number of gear teeth in the gearbox.

[0055] Figure 3 These are three-phase current signal diagrams of a faulty motor with a duration of 1s at different rotational frequencies according to an embodiment of the present invention, wherein (a) is a three-phase current signal diagram of a faulty motor with a duration of 1s at a rotational frequency of 5Hz, (b) is a three-phase current signal diagram of a faulty motor with a duration of 1s at a rotational frequency of 10Hz; (c) is a three-phase current signal diagram of a faulty motor with a duration of 1s at a rotational frequency of 15Hz; and (d) is a three-phase current signal diagram of a faulty motor with a duration of 1s at a rotational frequency of 20Hz.

[0056] Figure 4 1s long three-phase current signal diagram of a normal motor at different rotational frequencies according to an embodiment of the present invention, wherein (a) is a diagram of a normal motor at a rotational frequency of 5Hz with a length of 1s; (b) is a diagram of a normal motor at a rotational frequency of 10Hz with a length of 1s; (c) is a diagram of a normal motor at a rotational frequency of 15Hz with a length of 1s; and (d) is a diagram of a normal motor at a rotational frequency of 20Hz with a length of 1s.

[0057] Figure 5 This is a graph of the coefficient of variation of the instantaneous frequency and instantaneous amplitude of the three-phase current directly calculated according to an embodiment of the present invention.

[0058] Figure 6 This is a diagram of the VMD decomposition results of the instantaneous frequency when the rotation frequency is 10 Hz according to an embodiment of the present invention.

[0059] Figure 7 3 is a graph of the coefficient of variation of the instantaneous frequency and instantaneous amplitude of the current signal after trend items are extracted by VMD according to an embodiment of the present invention. DETAILED DESCRIPTION

[0060] The technical solution of the present invention is further described in detail below in conjunction with the accompanying drawings and examples. The following examples are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

[0061] Reference Figure 1 A method for analyzing three-phase imbalance of a motor based on instantaneous current information comprises the following steps:

[0062] 1) Extract the instantaneous frequency and instantaneous amplitude of the three-phase current through Park vector demodulation;

[0063] Park vector demodulation, the speed calculation method of the three-phase asynchronous motor is shown in the following formula:

[0064]

[0065] Where n r is the speed (r / min), s is the slip ratio, p is the number of pole pairs, and f is the rotation frequency;

[0066] The three-phase current signal is a modulated signal with a 120° phase difference between any two phases. By using Park transform, a 90° phase-shifted signal can be created to complete the signal demodulation, as shown below:

[0067]

[0068] Where I α and I β is the stator coordinate system current, I u , I vand I w They are the three-phase currents of the motor U, V, and W respectively;

[0069] Then the three-phase current can be expressed as:

[0070]

[0071] Where f M is the signal fundamental frequency, a is the signal amplitude; combining equations (2) and (3), we can get:

[0072]

[0073] byI α and I β Construct the analytical signal as shown in formula (5):

[0074] I z =I α +I β ·i (5)

[0075] Where I z is the constructed complex signal;

[0076] Calculating the modulus and phase angle of the constructed complex signal can obtain the instantaneous amplitude and instantaneous phase. z Finding the modulus and argument yields:

[0077]

[0078] Where a(t) is the instantaneous amplitude, is the instantaneous phase;

[0079] The instantaneous frequency is the inverse of the instantaneous phase, that is:

[0080]

[0081] Where f(t) is the instantaneous frequency;

[0082] The instantaneous amplitude and instantaneous frequency are the instantaneous information of the three-phase current signal multi-sensor information fusion, reflecting the comprehensive characteristics of the three-phase current;

[0083] 2) Extract the trend terms of instantaneous frequency and instantaneous amplitude through VMD;

[0084] The VMD method obtains the modal bandwidth. The steps are as follows:

[0085] The related analytical signal is calculated by Hilbert transform to obtain a one-sided spectrum;

[0086] For each mode, the spectrum of the mode is shifted to “baseband” by exponential mixing tuned to the corresponding estimated center frequency;

[0087] The bandwidth is estimated by h Gaussian smoothing of the gradient. The resulting constrained variational problem is as follows;

[0088]

[0089] In the formula, {u k} is a set of k intrinsic mode functions, {ω k} is a set of k center frequencies, δ(t) is the impulse function, H is the original signal, t is time, and j is the imaginary unit;

[0090] Through the Lagrange multiplier and the quadratic penalty term, the constrained objective function is transformed into an unconstrained objective function. The intervention of the Lagrange multiplier makes the condition of the constrained variation more stringent, and due to the influence of the limited penalty weight, the convergence of the quadratic penalty term is improved, and finally the augmented Lagrangian function can be obtained, that is:

[0091]

[0092] Where α is the balance coefficient, λ is the Lagrange multiplier, {u k} is a set of k intrinsic mode functions, {ω k} is a set of k center frequencies, δ(t) is the impulse function, H(t) is the original signal, t is time, and j is the imaginary unit;

[0093] For the current, vibration and other data of rotating mechanical equipment, by analyzing the characteristics of the equipment components, it can be obtained that the signal data mainly includes the main shaft rotation frequency and meshing frequency of the gear, which corresponds to the value of K;

[0094] Through the VMD method, the intrinsic mode functions (IMFs) of multiple characteristic frequencies can be extracted, and the trend term can be obtained by calculating the maximum mean value.

[0095] 3) Calculate the coefficient of variation of instantaneous frequency and instantaneous amplitude, compare and analyze the characteristics of the coefficient of variation of instantaneous frequency and instantaneous amplitude of normal motors and faulty motors directly calculated and extracted by VMD trend items, and perform motor fault diagnosis;

[0096] Calculation of coefficient of variation: The coefficient of variation is the ratio of the standard deviation to the mean, which is used to reflect the degree of variation of each observation value and is a dimensionless statistic;

[0097] The mean μ is expressed as follows:

[0098]

[0099] Where x i is the i-th data, n is the data length;

[0100] The sample standard deviation σ is shown as follows

[0101]

[0102] According to the definition of coefficient of variation, combining equations (10) and (11), we can get the coefficient of variation c v ;

[0103]

[0104] The coefficient of variation is only defined when the mean is not zero and is usually used when the mean is greater than zero. The coefficient of variation measures the fluctuation of the overall data. By calculating the coefficient of variation of the instantaneous frequency and instantaneous amplitude, its fluctuation state can be analyzed.

[0105] This embodiment builds a comprehensive gearbox dynamics test bench. Figure 2 As shown in the figure, (a) is the experimental bench structure diagram, (b) is the actual diagram of the current sensor clamping; (c) is a schematic diagram of the number of gear teeth in the gearbox; the three-phase current signals of the faulty motor and the normal motor with a rotation frequency of 5Hz, 10Hz, 15Hz and 20Hz are intercepted for 1s, as shown in Figure 3-Figure 4 shown.

[0106] Reference Figure 5 、 Figure 6 ,The difference in the coefficient of variation reflects the degree of three-phase imbalance between the ,normal motor and the faulty motor. When the trend item is not ,extracted, the coefficient of variation of the instantaneous frequency of the ,faulty motor and the normal motor are similar, and the coefficient of variation of the instantaneous ,amplitude is difficult to distinguish, and the motor fault ,diagnosis cannot be performed.

[0107] Reference Figure 7 ,After VMD decomposition and extraction of the trend item and ,the coefficient of variation calculated, the eigenvalue of the faulty motor is still higher than the ,eigenvalue of the normal motor. Compared with the undecomposed signal, ,the instantaneous amplitude is still not well distinguished, but the instantaneous frequency can be ,distinguished more clearly. The instantaneous frequency variation coefficient can ,be used as a characteristic coefficient for judging three-phase imbalance and ,performing motor fault diagnosis.

[0108] The above embodiments are only used to illustrate the present invention, rather than to limit the present invention. Ordinary technicians in the relevant technical field can make various changes and modifications without departing from the spirit and scope of the present invention. Therefore, all equivalent technical solutions also fall within the scope of the present invention. The scope of patent protection of the present invention should be defined by the claims.

Claims

1. A motor three-phase imbalance analysis method based on current instantaneous information, characterized in that: The following steps are involved: 1) Extract the instantaneous frequency and instantaneous amplitude of the three-phase current through Park vector demodulation; 2) Extract the trend terms of instantaneous frequency and instantaneous amplitude through VMD; The VMD method obtains the modal bandwidth. The steps are as follows: The related analytical signal is calculated by Hilbert transform to obtain a one-sided spectrum; For each mode, the spectrum of the mode is shifted to "baseband" by exponential mixing tuned to the corresponding estimated center frequency; The bandwidth is estimated by Gaussian smoothing of the gradient h. The resulting constrained variational problem is as follows; In the formula, {u k } is a set of k intrinsic mode functions, {ω k } is a set of k center frequencies, δ(t) is the impulse function, H is the original signal, t is time, and j is the imaginary unit; Through the Lagrange multiplier and the quadratic penalty term, the constrained objective function is transformed into an unconstrained objective function, and the augmented Lagrangian function is obtained, that is: Where α is the balance coefficient, λ is the Lagrange multiplier, {u k } is a set of k intrinsic mode functions, {ω k ) is a set of k center frequencies, δ(t) is the impulse function, H(t) is the original signal, t is time, and j is the imaginary unit; The VMD method is used to extract the imf of multiple characteristic frequencies, and the trend term is obtained by calculating the maximum mean term. 3) Calculate the coefficient of variation of instantaneous frequency and instantaneous amplitude, compare and analyze the characteristics of the coefficient of variation of instantaneous frequency and instantaneous amplitude of normal motors and faulty motors directly calculated and extracted by VMD trend items, and perform motor fault diagnosis; The coefficient of variation is calculated as: The coefficient of variation is the ratio of the standard deviation to the mean, which is used to reflect the degree of variation of each observation value and is a dimensionless statistic. The mean μ is expressed as follows: Where x i is the i-th data, n is the data length; The sample standard deviation σ is expressed as follows: According to the definition of coefficient of variation, combining equations (10) and (11), we can get the coefficient of variation c v ; The coefficient of variation is only defined when the mean is not zero and is used when the mean is greater than zero. The coefficient of variation measures the fluctuation of the overall data and analyzes its fluctuation state by calculating the coefficient of variation of the instantaneous frequency and instantaneous amplitude.

2. The method according to claim 1, wherein: In the step 1) Park vector demodulation, the speed calculation method of the three-phase asynchronous motor is shown in the following formula: Where n r is the speed in r / min, s is the slip rate, p is the number of pole pairs, and f is the rotation frequency; Signal demodulation is accomplished by using the Park transform to create a 90° phase-shifted signal, as shown below: Where I α and I β is the stator coordinate system current, I u , I v and I w They are the three-phase currents of the motor U, V, and W respectively; The three-phase current is expressed as: Where f M is the signal fundamental frequency, a is the signal amplitude; combining equations (2) and (3), we can get: byI α and I β Construct the analytical signal as shown in formula (5): I z =I α +I β ·i (5) Where I z is the constructed complex signal; Calculate the modulus and phase angle of the constructed complex signal to get the instantaneous amplitude and instantaneous phase. z Finding the modulus and argument yields: Where a(t) is the instantaneous amplitude, is the instantaneous phase; The instantaneous frequency is the inverse of the instantaneous phase, that is: Where f(t) is the instantaneous frequency; The instantaneous amplitude and instantaneous frequency are the instantaneous information of the three-phase current signal multi-sensor information fusion, reflecting the comprehensive characteristics of the three-phase current.

Images