Online detection method and device for stator short-circuit faults in motors under transient operating conditions

Abstract

This invention relates to an online detection method and apparatus for stator short-circuit faults in motors under transient operating conditions. The method includes: performing Clark transform on the three-phase currents to obtain α and β two-phase currents; calculating the current amplitude, estimating the fundamental frequency, and determining the target frequency band based on the two-phase currents; constructing two-phase current signals by combining historical data; obtaining new two-phase current signals through fast Fourier transform, Hilbert filtering, bandpass filtering, and inverse fast Fourier transform; extracting the new two-phase currents; calculating the negative sequence current, current amplitude, and short-circuit fault index based on the new two-phase currents; and performing fault early warning processing based on the short-circuit fault index and fault index threshold. This invention can improve the detection sensitivity and reliability of short-circuit fault detection under transient operating conditions.

Description

Technical Field

[0001] This invention relates to the field of motor fault diagnosis technology, and in particular to an online detection method and apparatus for stator short-circuit faults in motors under transient operating conditions. Background Technology

[0002] Stator inter-turn short circuit faults are among the most common early-stage faults in motors. If not detected in time, they can quickly escalate into phase-to-phase short circuits or grounding faults, leading to motor damage. Therefore, online detection of stator short circuits is crucial for ensuring the safe operation of the system. Existing fault detection methods are mainly divided into offline testing and online detection. Offline testing (such as insulation resistance testing and pulse voltage testing) cannot meet real-time requirements and is difficult to use for fault early warning during operation. Online detection methods based on current signal analysis have attracted widespread attention due to their advantages such as low cost, non-invasiveness, and no need for additional sensors.

[0003] Currently, traditional online detection methods include three-phase current imbalance detection, negative sequence current analysis, and harmonic feature extraction. These traditional solutions are generally based on the steady-state assumption—that is, the motor operates under constant speed and load conditions, and the current signal is approximately stable. However, in practical applications, motors often operate under transient conditions, such as variable frequency speed regulation, frequent start-stop, and sudden load changes. Under transient conditions, the current signal exhibits strong non-stationary characteristics, with rapid changes in frequency and amplitude. This leads to the following problems with traditional solutions based on the steady-state assumption: feature extraction based on Fourier transform cannot accurately reflect the frequency components in the transient process, resulting in severe spectral leakage; the amplitude of the negative sequence current is affected by fluctuations in operating conditions, making it difficult to set a fixed threshold; and feature calculation based on integer-cycle integration fails when the frequency changes. These problems reduce the fault detection sensitivity and reliability of traditional solutions.

[0004] To address the above issues, we propose an improved solution: First, we propose a fundamental frequency estimation mechanism based on current amplitude and rotational speed to improve the frequency estimation accuracy under transient conditions. Then, we extract short-time signals using a sliding window and construct an analytic signal in the frequency domain by first eliminating negative frequency components through Hilbert filtering. Next, we filter the harmonics and noise of the analytic signal using a bandpass filter corresponding to the fundamental target frequency band, thereby accurately extracting the instantaneous fundamental component under rapidly changing frequency conditions. Following this, we calculate the negative sequence current amplitude and the fundamental current amplitude based on the extracted instantaneous fundamental component. Finally, we calculate the fault indicators. Under normal circumstances, the negative sequence current amplitude is very small, at which point the index factor... The fault index is a very small value close to 0; when a short-circuit fault occurs, the negative sequence current increases significantly, at which point the index factor... and The significant difference leads to a substantial increase in fault indicators, thus achieving the goal of real-time online detection. The aforementioned improved scheme enhances the accuracy and sensitivity of fundamental current amplitude estimation under transient conditions through an improved fundamental frequency estimation mechanism, and improves the accuracy and sensitivity of online detection under transient conditions through a fault indicator setting mechanism. Based on these improved schemes, the sensitivity and reliability of fault detection under transient conditions can be improved, effectively solving the problem that traditional schemes are not suitable for transient conditions. How to implement these improved schemes is the technical problem that this invention aims to solve. Summary of the Invention

[0005] The purpose of this invention is to address the shortcomings of existing technologies by providing an online detection method and device for short-circuit faults in motor stators under transient operating conditions. This invention performs Clark transform on the real-time three-phase currents of the motor stator to obtain the αβ two-phase currents in a two-phase stationary coordinate system. Based on the αβ two-phase currents, the current amplitude is calculated, and the fundamental frequency is estimated based on the current amplitude and real-time speed. The target frequency band is then determined based on the fundamental frequency. Next, a two-phase current signal is composed of historical and real-time data of the two-phase currents using a sliding window mechanism. This two-phase current signal is then converted into a two-phase frequency domain signal using a Fast Fourier Transform (FFT). Subsequently, a Hilbert filter is constructed to perform negative frequency filtering on the frequency domain signal, a bandpass filter constructed from the target frequency band is used to perform non-target frequency filtering on the Hilbert-filtered signal, and finally, an Inverse Fast Fourier Transform (FFT) is performed. The IFFT (Instantaneous Interpretive Transform) method converts the filtered two-phase frequency domain signal into a two-phase current signal represented in complex form, which can then be approximated as the fundamental frequency signal. The current at the last moment of the two-phase fundamental frequency signal is then considered as the instantaneous fundamental frequency component. Based on this instantaneous fundamental frequency component, the negative-sequence current amplitude and the fundamental frequency current amplitude are calculated, and a short-circuit fault index is calculated based on the negative-sequence / fundamental frequency current amplitude. Finally, a fault warning is performed based on the real-time short-circuit fault index using a threshold method. This invention can improve the detection sensitivity and reliability of short-circuit fault detection under transient operating conditions.

[0006] To achieve the above objectives, a first aspect of the present invention provides an online detection method for stator short-circuit faults in motors under transient operating conditions, the method comprising: Based on the preset sampling frequency f s The three-phase currents (a, b, c) of the motor stator and the real-time speed are continuously collected, and the three-phase currents and real-time speed collected at time t are recorded as i. a(t) i b(t) i c(t) n r(t) Wherein, time zero t=0 is the starting time of the motor in this operating cycle; For three-phase current i a(t) i b(t) i c(t) The Clark transformation is performed to obtain the two-phase current i in the αβ two-phase stationary coordinate system. α(t) i β(t) ; and based on the two-phase current i α(t) i β(t) Calculate the current amplitude I t ; and based on the current amplitude I t Real-time rotational speed n r(t) and the preset rated current frequency f N Rated speed n N Rated current amplitude I N Estimating the fundamental frequency f based on no-load speed n0 and no-load current amplitude I0 1(t) ; and based on the fundamental frequency f 1(t) Determine the target frequency band [f min(t) ,f max(t) ];f min(t) f max(t) These are the lower and upper thresholds of the target frequency band, respectively. The M two-phase currents i from time t-(M-1) to time t α(t) i β(t) The corresponding two-phase current signal S1 is formed α S1 β And the two-phase current signal S1 α S1 β Performing Fast Fourier Transform on each signal yields the corresponding two-phase frequency domain signals X1. α X1 β And construct a Hilbert filter for the two-phase frequency domain signal X1. α X1 β Negative frequency filtering is performed separately to obtain the corresponding two-phase frequency domain signals X2. α X2 β ; and based on the target frequency band [f min(t) ,f max(t) Construct a bandpass filter for the two-phase frequency domain signal X2 α X2 β Non-target frequency filtering is performed to obtain the corresponding two-phase frequency domain signal X3 α X3 β ; and the two-phase frequency domain signal X3 α X3 β Perform inverse fast Fourier transform on each phase to obtain the corresponding two-phase current signals S2. α S2 β ; and from the two-phase current signal S2 α S2 βExtract the two-phase current i corresponding to time t α-com(t) i β-com(t) The total number of sampling points M is a preset positive integer; the two-phase current i α-com(t) i β-com(t) It is a complex current; Based on the two-phase current i α-com(t) i β-com(t) The negative sequence current I is obtained by calculating the negative sequence current amplitude. n(t) And based on the two-phase current i α-com(t) i β-com(t) For the current amplitude I t Perform a reset, based on the negative sequence current I. n(t) and the current amplitude I after reset t Calculate the corresponding short-circuit fault index Z t ; Based on the short-circuit fault index Z t and the preset fault indicator threshold Z hold Implement fault warning and handling.

[0007] Preferably, the three-phase current i a(t) i b(t) i c(t) and the two-phase current i α(t) i β(t) All are real currents; from the three-phase current i a(t) i b(t) i c(t) to the two-phase current i α(t) i β(t) The Clark transformation method is as follows: , ; Among them, C 3 / 2 This is the Clark transformation matrix; The current amplitude I t The calculation method is as follows: ; The fundamental frequency f 1(t) The estimation method is as follows: ; The target frequency band [f min(t) ,f max(t) The method for determining ] is as follows: , ; Where △f is a preset adjustable frequency band parameter.

[0008] Preferably, the two-phase current signal S1α S1 β Current signal S1 α Specifically expressed as Current signal S1 β Specifically expressed as Two-phase current signal , of , Let be the two-phase real current corresponding to the u-th time sampling point; u=1 corresponds to time t-(M-1), then the current , Let i be the two-phase current at time t-(M-1). α(t-(M-1)) i β(t-(M-1)) ; if u=M corresponds to time t, then the current , Let i be the two-phase current at time t. α(t) i β(t) ; 1 ≤ sampling point index u ≤ M, sampling time interval Δt = 1 / f between every two time sampling points s ; The two-phase frequency domain signal X1 α X1 β frequency domain signal X1 α Specifically expressed as Frequency domain signal X1 β Specifically expressed as Two-phase frequency domain signals , of , The two-phase frequency domain features corresponding to the u-th frequency domain sampling point; two-phase frequency domain features , It is in complex form; The two-phase frequency domain signal X2 α X2 β frequency domain signal X2 α Specifically expressed as Frequency domain signal X2 β Specifically expressed as Two-phase frequency domain signals , of , The two-phase frequency domain features corresponding to the u-th frequency domain sampling point; two-phase frequency domain features , It is in complex form; The two-phase frequency domain signal X3 α X3 β frequency domain signal X3 α Specifically expressed as Frequency domain signal X3 β Specifically expressed as Two-phase frequency domain signals , of , The two-phase frequency domain features corresponding to the u-th frequency domain sampling point; two-phase frequency domain features , It is in complex form; The two-phase current signal S2 α S2 β Current signal S2 α Specifically expressed as Current signal S2 β Specifically expressed as Two-phase current signal , of , Let be the two-phase complex current corresponding to the u-th time sampling point; two-phase complex current , It is in complex form; Complex current Specifically, it can be expressed as follows: , , These correspond to the real and imaginary parts, respectively; complex current. Specifically, it can be expressed as follows: , , These represent the real and imaginary parts, respectively; j is the imaginary unit. The two-phase current i α-com(t) i β-com(t) Specifically, it can be expressed as follows: , ; , They are the currents i α-com(t) The real and imaginary parts, , They are the currents i β-com(t) The real and imaginary parts.

[0009] Preferably, the two-phase current signal S1 α S1 β Performing Fast Fourier Transform on each signal yields the corresponding two-phase frequency domain signals X1. α X1 β Specifically, it includes: Step 41, based on the sampling frequency f s The sampling bandwidth Δf of the Fast Fourier Transform is determined by the total number of sampling points M. FFTand frequency range; and based on the frequency range and the sampling bandwidth Δf FFT Confirm M frequency domain sampling points; Among them, △f FFT =f s / M; the lower limit of the frequency range is 0Hz and the upper limit is f s -△f FFT ; The total number of frequency domain sampling points in the frequency range is consistent with the total number of time domain sampling points of the current signal; Step 42, for the two-phase current signals , Performing Fast Fourier Transform on each signal yields the corresponding two-phase frequency domain signals. , : , ; FFT() is the Fast Fourier Transform function.

[0010] Preferably, the constructed Hilbert filter is applied to the two-phase frequency domain signal X1. α X1 β Negative frequency filtering is performed separately to obtain the corresponding two-phase frequency domain signals X2. α X2 β Specifically, it includes: Step 51, construct the Hilbert filter H u : ; Step 52, based on the Hilbert filter H u For the two-phase frequency domain signal X1 α X1 β Negative frequency filtering is performed separately to obtain the corresponding two-phase frequency domain signals X2. α X2 β : , .

[0011] Preferably, the step based on the target frequency band [f] min(t) ,f max(t) Construct a bandpass filter for the two-phase frequency domain signal X2 α X2 β Non-target frequency filtering is performed to obtain the corresponding two-phase frequency domain signal X3 α X3 β Specifically, it includes: Step 61, based on the target frequency band [f min(t) ,f max(t) and the sampling frequency fs Calculate the corresponding lower bound index u min and upper limit index u max : , ; floor() is the floor function, and ceiling() is the ceiling function. Step 62, based on the lower bound index u min and the upper limit index u max Construct a bandpass filter P u : ; Step 63, based on the bandpass filter P u For the two-phase frequency domain signal X2 α X2 β The corresponding two-phase frequency domain signals X3 are obtained by performing non-target frequency filtering respectively. α X3 β : , .

[0012] Preferably, the two-phase frequency domain signal X3 α X3 β Perform inverse fast Fourier transform on each phase to obtain the corresponding two-phase current signals S2. α S2 β Specifically, it includes: For the two-phase frequency domain signals , Perform inverse fast Fourier transform on each phase to obtain the corresponding two-phase current signals. , : , ; Here, IFFT() is the inverse fast Fourier transform function.

[0013] Preferably, the two-phase current signal S2 α S2 β Extract the two-phase current i corresponding to time t α-com(t) i β-com(t) Specifically, it includes: Current signal The Mth complex current As the corresponding current i α-com(t) ; and the current signal The Mth complex current As the corresponding current i β-com(t) ; in, , , ; , , ; The current i α-com(t) real part virtual part The complex currents are respectively real part virtual part ; The current i β-com(t) real part virtual part The complex currents are respectively real part virtual part .

[0014] Preferably, the two-phase current i α-com(t) i β-com(t) The negative sequence current I is obtained by calculating the negative sequence current amplitude. n(t) And based on the two-phase current i α-com(t) i β-com(t) For the current amplitude I t Perform a reset, based on the negative sequence current I. n(t) and the current amplitude I after reset t Calculate the corresponding short-circuit fault index Z t Specifically, it includes: Step 91, convert the two-phase current i α-com(t) i β-com(t) Substituting the two sets of real and imaginary parts into the formula for calculating the negative sequence current amplitude, the corresponding negative sequence current I is obtained. n(t) : The formula for calculating the negative sequence current amplitude is as follows: ; Step 92, based on the two-phase current i α-com(t) i β-com(t) The real part of the current amplitude I t Reset: ; Step 93, based on the negative sequence current I n(t) and the current amplitude I t Calculate the short-circuit fault index Z t : .

[0015] Preferably, the short-circuit fault index Z is used as the basis for... t and the preset fault indicator threshold Z hold Fault early warning processing includes: For the short-circuit fault index Z t Is it greater than the fault indicator threshold Z? hold Perform identification; if yes, set the corresponding warning identification state to yes; if no, set the corresponding warning identification state to no; and when the warning identification state is yes, use time t and its corresponding three-phase current (i a(t) i b(t) i c(t) The real-time rotational speed n r(t) The negative sequence current I n(t) The current amplitude I t The short-circuit fault index Z t and the two-phase current signal S1 α S1 β The corresponding early warning data groups are formed and sent to the preset online early warning interface for short circuit faults.

[0016] A second aspect of the present invention provides an apparatus for implementing the online detection method for short-circuit faults of motor stator under transient operating conditions as described in the first aspect above. The apparatus includes: a data acquisition module, a fundamental target frequency band processing module, an improved Hilbert transform module, a negative sequence current and short-circuit fault index processing module, and a fault early warning module. The data acquisition module is based on a preset sampling frequency f s The three-phase currents (a, b, c) of the motor stator and the real-time speed are continuously collected, and the three-phase currents and real-time speed collected at time t are recorded as i. a(t) i b(t) i c(t) n r(t) Wherein, time zero t=0 is the starting time of the motor in this operating cycle; The fundamental target frequency band processing module is used to process the three-phase current i a(t) i b(t) i c(t) The Clark transformation is performed to obtain the two-phase current i in the αβ two-phase stationary coordinate system. α(t) i β(t) ; and based on the two-phase current i α(t) i β(t) Calculate the current amplitude I t ; and based on the current amplitude I t Real-time rotational speed n r(t) and the preset rated current frequency fN Rated speed n N Rated current amplitude I N Estimating the fundamental frequency f based on no-load speed n0 and no-load current amplitude I0 1(t) ; and based on the fundamental frequency f 1(t) Determine the target frequency band [f min(t) ,f max(t) ];f min(t) f max(t) These are the lower and upper thresholds of the target frequency band, respectively. The improved Hilbert transform module is used to process M two-phase currents i from time t-(M-1) to time t. α(t) i β(t) The corresponding two-phase current signal S1 is formed α S1 β And the two-phase current signal S1 α S1 β Performing Fast Fourier Transform on each signal yields the corresponding two-phase frequency domain signals X1. α X1 β And construct a Hilbert filter for the two-phase frequency domain signal X1. α X1 β Negative frequency filtering is performed separately to obtain the corresponding two-phase frequency domain signals X2. α X2 β ; and based on the target frequency band [f min(t) ,f max(t) Construct a bandpass filter for the two-phase frequency domain signal X2 α X2 β Non-target frequency filtering is performed to obtain the corresponding two-phase frequency domain signal X3 α X3 β ; and the two-phase frequency domain signal X3 α X3 β Perform inverse fast Fourier transform on each phase to obtain the corresponding two-phase current signals S2. α S2 β ; and from the two-phase current signal S2 α S2 β Extract the two-phase current i corresponding to time t α-com(t) i β-com(t) The total number of sampling points M is a preset positive integer; the two-phase current i α-com(t) i β-com(t) It is a complex current; The negative sequence current and short-circuit fault index processing module is based on the two-phase current i α-com(t) i β-com(t) The negative sequence current I is obtained by calculating the negative sequence current amplitude. n(t) And based on the two-phase current iα-com(t) i β-com(t) For the current amplitude I t Perform a reset, based on the negative sequence current I. n(t) and the current amplitude I after reset t Calculate the corresponding short-circuit fault index Z t ; The fault early warning module is based on the short-circuit fault index Z. t and the preset fault indicator threshold Z hold Implement fault warning and handling.

[0017] This invention provides an online detection method and apparatus for short-circuit faults in the stator of a motor under transient operating conditions. As described above, this invention performs Clark transform on the real-time three-phase current of the motor stator to obtain the αβ two-phase current in a two-phase stationary coordinate system. Based on the αβ two-phase current, the current amplitude is calculated, and the fundamental frequency is estimated based on the current amplitude and real-time speed. The target frequency band is then determined based on the fundamental frequency. Next, a two-phase current signal is composed of historical and real-time data of the two-phase current using a sliding window mechanism. This two-phase current signal is then converted into a two-phase frequency domain signal using a Fast Fourier Transform. Finally, a Hilbert filter is constructed to perform negative frequency filtering on the frequency domain signal. The Hilbert-filtered signal is filtered at a non-target frequency using a bandpass filter constructed from the target frequency band. The filtered two-phase frequency domain signal is then converted into a two-phase current signal represented in complex form using an inverse fast Fourier transform. This two-phase current signal can be approximated as the fundamental wave signal. The current at the last moment of the two-phase fundamental wave signal is then considered as the instantaneous fundamental wave component. Based on the instantaneous fundamental wave component, the negative-sequence current amplitude and the fundamental current amplitude are calculated, and a short-circuit fault index is calculated based on the negative-sequence / fundamental current amplitude. Finally, a fault warning is performed based on the real-time short-circuit fault index using a threshold method. This embodiment of the invention improves the detection sensitivity and reliability of short-circuit fault detection under transient operating conditions. Attached Figure Description

[0018] Figure 1 This is a schematic diagram of an online detection method for stator short-circuit faults in a motor under transient operating conditions, provided in Embodiment 1 of the present invention. Figure 2 This is a module structure diagram of an online detection device for stator short-circuit faults in a motor under transient operating conditions, provided in Embodiment 2 of the present invention. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are merely some embodiments of this invention, and not all embodiments. Based on the embodiments of this invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this invention.

[0020] Embodiment 1 of the present invention provides an online detection method for stator short-circuit faults in motors under transient operating conditions, such as... Figure 1 The schematic diagram of an online detection method for stator short-circuit faults in a motor under transient operating conditions provided in Embodiment 1 of the present invention shows that the method mainly includes the following steps: Step 1, based on the preset sampling frequency f s The three-phase currents (a, b, c) of the motor stator and the real-time speed are continuously collected, and the three-phase currents and real-time speed collected at time t are recorded as i. a(t) i b(t) i c(t) n r(t) .

[0021] Here, the sampling frequency f in this embodiment of the invention s This is a pre-set time and frequency parameter, which can be customized based on application requirements. In this embodiment of the invention, time zero (t=0) is the starting time of the motor's current operating cycle. The three-phase current i in this embodiment of the invention... a(t) i b(t) i c(t) The current is a real number.

[0022] Step 2, for the three-phase current i a(t) i b(t) i c(t) The Clark transformation is performed to obtain the two-phase current i in the αβ two-phase stationary coordinate system. α(t) i β(t) ; and based on the two-phase current i α(t) i β(t) Calculate the current amplitude I t And based on the current amplitude I t Real-time rotational speed n r(t) and the preset rated current frequency f N Rated speed n N Rated current amplitude I N Estimating the fundamental frequency f based on no-load speed n0 and no-load current amplitude I0 1(t) Based on the fundamental frequency f 1(t) Determine the target frequency band [f min(t) ,f max(t) ].

[0023] Here, the rated current frequency f of the embodiment of the present invention N Rated speed n N Rated current amplitude I N The no-load speed n0 and no-load current amplitude I0 are known parameters from the motor manufacturer and are matched with the specific motor type and model.

[0024] Specifically, it includes: Step 21, for the three-phase current i a(t) i b(t) i c(t) The Clark transformation is performed to obtain the two-phase current i in the αβ two-phase stationary coordinate system. α(t) i β(t) .

[0025] Here, the two-phase current i in this embodiment of the invention α(t) i β(t) All are real currents.

[0026] From the three-phase current i a(t) i b(t) i c(t) to two-phase current i α(t) i β(t) The Clark transformation method is as follows: , ; Among them, C 3 / 2 This is the Clark transformation matrix.

[0027] Step 22, based on the two-phase current i α(t) i β(t) Calculate the current amplitude I t .

[0028] Here, the current amplitude I in the current step 22 t The calculation method is as follows: .

[0029] The current amplitude I calculated in step 22 t Its physical meaning is the real-time current amplitude at time t.

[0030] Step 23, and based on the current amplitude I t Real-time rotational speed n r(t) and the preset rated current frequency f N Rated speed n N Rated current amplitude I N Estimating the fundamental frequency f based on no-load speed n0 and no-load current amplitude I0 1(t) .

[0031] Here, the fundamental frequency f of this embodiment of the invention 1(t) The estimation method is as follows: .

[0032] It should be noted that the no-load speed n0 is greater than the rated speed n N .

[0033] It should also be noted that the conventional method for estimating the fundamental frequency is as follows: This estimation method assumes the real-time fundamental frequency f. 1(t) With real-time rotational speed n r(t) The ratio is fixed at f N / n N However, in actual transient operating conditions, because the mechanical inertia speed cannot instantly follow the load change, while the current can instantaneously reflect the load, the conventional fixed-proportion estimation method will lead to frequency estimation lag. To eliminate the problem of decreased estimation accuracy caused by frequency estimation lag under transient operating conditions, this embodiment of the invention improves the estimation method of the fundamental frequency by adding a dynamic correction term to the denominator. This correction term is essentially based on the real-time current amplitude I. t It instantly predicts the impact of load changes on the speed reference, thereby improving the accuracy of the estimation.

[0034] The effectiveness of the improved estimation mechanism in this invention embodiment is verified: when the load is the rated load, I t =I N I N -I t =0, correction term is 0, denominator is n N The formula estimation result is consistent with the conventional estimation method; when the load is less than the rated load, I t <I N I N -I t >0, correction term greater than 0, estimated real-time frequency decreases; when the load exceeds the rated load, I t >I N I N -I t <0, correction term less than 0, estimated real-time frequency increases. These states all conform to the frequency change trend under actual transient operating conditions. Therefore, the improved estimation mechanism of this invention is effective and can achieve the goal of improving estimation accuracy.

[0035] Step 24, and based on the fundamental frequency f 1(t) Determine the target frequency band [f min(t) ,f max(t) ].

[0036] Here, f in the embodiment of the present inventionmin(t) f max(t) These are the lower and upper thresholds of the target frequency band, respectively; target frequency band [f min(t) ,f max(t) The method for determining ] is as follows: , ; Here, △f is a preset adjustable frequency band parameter. This adjustable frequency band parameter △f is a pre-set frequency parameter that can be customized based on application implementation requirements.

[0037] It should be noted that the target frequency band [f] in this embodiment of the invention min(t) ,f max(t) In reality, it is a narrow frequency band near the fundamental frequency, which can also be called the fundamental frequency band.

[0038] Step 3, the M two-phase currents i from time t-(M-1) to time t. α(t) i β(t) The corresponding two-phase current signal S1 is formed α S1 β And for the two-phase current signal S1 α S1 β Performing Fast Fourier Transform on each signal yields the corresponding two-phase frequency domain signals X1. α X1 β And construct a Hilbert filter for the two-phase frequency domain signal X1. α X1 β Negative frequency filtering is performed separately to obtain the corresponding two-phase frequency domain signals X2. α X2 β ; and based on the target frequency band [f min(t) ,f max(t) Construct a bandpass filter for the two-phase frequency domain signal X2 α X2 β Non-target frequency filtering is performed to obtain the corresponding two-phase frequency domain signal X3 α X3 β And for the two-phase frequency domain signal X3 α X3 β Perform inverse fast Fourier transform on each phase to obtain the corresponding two-phase current signals S2. α S2 β ; and from the two-phase current signal S2 α S2 β Extract the two-phase current i corresponding to time t α-com(t) i β-com(t) .

[0039] Specifically, it includes: Step 31, the M two-phase currents i from time t-(M-1) to time t.α(t) i β(t) The corresponding two-phase current signal S1 is formed α S1 β .

[0040] Here, the total number of sampling points M in this embodiment of the invention is a preset positive integer.

[0041] Two-phase current signal S1 in this embodiment of the invention α S1 β Current signal S1 α Specifically expressed as Current signal S1 β Specifically expressed as .

[0042] 1 ≤ sampling point index u ≤ M, sampling time interval Δt = 1 / f between every two time sampling points s .

[0043] Two-phase current signal , of , Let be the two-phase real current corresponding to the u-th time sampling point. Simply put: u=1 corresponds to time t-(M-1), then the current... , Let i be the two-phase current at time t-(M-1). α(t-(M-1)) i β(t-(M-1)) ; if u=M corresponds to time t, then the current , Let i be the two-phase current at time t. α(t) i β(t) .

[0044] Step 32, and process the two-phase current signal S1 α S1 β Performing Fast Fourier Transform on each signal yields the corresponding two-phase frequency domain signals X1. α X1 β .

[0045] The two-phase frequency domain signal X1 of this invention embodiment α X1 β frequency domain signal X1 α Specifically expressed as Frequency domain signal X1 β Specifically expressed as Two-phase frequency domain signals , of , The two-phase frequency domain features corresponding to the u-th frequency domain sampling point; two-phase frequency domain features , It is in complex form.

[0046] The current step 32 specifically includes: Step 321, based on the sampling frequency f s The sampling bandwidth Δf of the Fast Fourier Transform is determined by the total number of sampling points M. FFT and frequency range; and based on the frequency range and sampling bandwidth Δf FFT Confirm M frequency domain sampling points.

[0047] Here, △f FFT =f s / M; The lower limit of the frequency range is 0Hz, and the upper limit is f. s -△f FFT The total number of frequency domain sampling points in the frequency range is consistent with the total number of time domain sampling points of the current signal.

[0048] Step 322, for the two-phase current signal , Performing Fast Fourier Transform on each signal yields the corresponding two-phase frequency domain signals. , : , ; FFT() is the Fast Fourier Transform function.

[0049] Step 33, and construct a Hilbert filter for the two-phase frequency domain signal X1. α X1 β Negative frequency filtering is performed separately to obtain the corresponding two-phase frequency domain signals X2. α X2 β .

[0050] Here, the two-phase frequency domain signal X2 in this embodiment of the invention α X2 β frequency domain signal X2 α Specifically expressed as Frequency domain signal X2 β Specifically expressed as Two-phase frequency domain signals , of , The two-phase frequency domain features corresponding to the u-th frequency domain sampling point; two-phase frequency domain features , It is in complex form.

[0051] The current step 33 specifically includes: Step 331, construct the Hilbert filter H u : .

[0052] Step 332, based on Hilbert filter H u For the two-phase frequency domain signal X1 α X1 β Negative frequency filtering is performed separately to obtain the corresponding two-phase frequency domain signals X2. α X2 β : , .

[0053] Step 34, and based on the target frequency band [f min(t) ,f max(t) Construct a bandpass filter for the two-phase frequency domain signal X2 α X2 β Non-target frequency filtering is performed to obtain the corresponding two-phase frequency domain signal X3 α X3 β .

[0054] Here, the two-phase frequency domain signal X3 in this embodiment of the invention α X3 β frequency domain signal X3 α Specifically expressed as Frequency domain signal X3 β Specifically expressed as Two-phase frequency domain signals , of , The two-phase frequency domain features corresponding to the u-th frequency domain sampling point; two-phase frequency domain features , It is in complex form.

[0055] The current step 34 specifically includes: Step 341, based on the target frequency band [f min(t) ,f max(t) and sampling frequency f s Calculate the corresponding lower bound index u min and upper limit index u max : , .

[0056] Among them, floor() is the floor function and ceiling() is the floor function.

[0057] Step 342, based on the lower bound index u min and upper limit index u max Construct a bandpass filter P u : .

[0058] Step 343, based on bandpass filter P u For the two-phase frequency domain signal X2 α X2 β The corresponding two-phase frequency domain signals X3 are obtained by performing non-target frequency filtering respectively. α X3 β : , .

[0059] Step 35, and process the two-phase frequency domain signal X3 α X3 β Perform inverse fast Fourier transform on each phase to obtain the corresponding two-phase current signals S2. α S2 β .

[0060] Here, the two-phase current signal S2 in this embodiment of the invention. α S2 β Current signal S2 α Specifically expressed as Current signal S2 β Specifically expressed as Two-phase current signal , of , Let be the two-phase complex current corresponding to the u-th time sampling point; two-phase complex current , It is in complex form.

[0061] Complex current in embodiments of the present invention Specifically, it can be expressed as follows: , in, , These are the corresponding real and imaginary parts, respectively; j is the imaginary unit.

[0062] Complex current in embodiments of the present invention Specifically, it can be expressed as follows: , in, , These are the corresponding real and imaginary parts, respectively.

[0063] The current step 35 specifically includes: processing the two-phase frequency domain signals. , Perform inverse fast Fourier transform on each phase to obtain the corresponding two-phase current signals. , : , ; Here, IFFT() is the inverse fast Fourier transform function.

[0064] It should be noted that, in this embodiment of the invention, the transformation process in steps 32-35 is considered as an improved Hilbert transform process. The two-phase current signal S2 obtained in this embodiment of the invention... α S2 β These are two extracted fundamental current signals. The improved Hilbert transform in steps 32-35 of this embodiment can improve the extraction accuracy of the fundamental current signal under transient operating conditions.

[0065] Step 36, and from the two-phase current signal S2 α S2 β Extract the two-phase current i corresponding to time t α-com(t) i β-com(t) .

[0066] Here, the two-phase current i in this embodiment of the invention α-com(t) i β-com(t) It is a complex current.

[0067] Two-phase current i α-com(t) i β-com(t) Specifically, it can be expressed as follows: , ; in, , They are the currents i α-com(t) The real and imaginary parts, , They are the currents i β-com(t) The real and imaginary parts.

[0068] The current step 36 specifically includes: converting the current signal The Mth complex current As the corresponding current i α-com(t) ; and the current signal The Mth complex current As the corresponding current i β-com(t) .

[0069] in: , , ; , , .

[0070] Current i α-com(t) real part virtual part These are complex currents. real part virtual part Current i β-com(t) real part virtual part These are complex currents. real part virtual part .

[0071] Step 4, based on the two-phase current i α-com(t) i β-com(t) The negative sequence current I is obtained by calculating the negative sequence current amplitude. n(t) And based on the two-phase current i α-com(t) i β-com(t) For current amplitude I t Perform a reset, based on the negative sequence current I. n(t) and the current amplitude I after reset t Calculate the corresponding short-circuit fault index Z t .

[0072] Specifically, it includes: Step 41, convert the two-phase current i α-com(t) i β-com(t) Substituting the two sets of real and imaginary parts into the formula for calculating the negative sequence current amplitude, the corresponding negative sequence current I is obtained. n(t) .

[0073] Here, the formula for calculating the negative sequence current amplitude in this embodiment of the invention is: .

[0074] Step 42, based on the two-phase current i α-com(t) i β-com(t) The real part of the current amplitude I t Reset.

[0075] Here, in the current step 42, the current amplitude I is... t The reset method is as follows: .

[0076] It should be noted that the current amplitude I calculated in step 42 is... t Its physical meaning is the amplitude of the fundamental current at time t.

[0077] Step 43, based on negative sequence current I n(t) and current amplitude I t Calculate the short-circuit fault index Z t : .

[0078] Under normal circumstances, the negative sequence current amplitude is very small, at which point the index factor... Short circuit fault index Z t It is a minimum value close to 0; when a short-circuit fault occurs, the negative sequence current increases significantly, at which point the index factor... and The difference is significant, resulting in a decrease in the short-circuit fault index Z. t The size increases significantly, thus enabling real-time online detection.

[0079] Step 5, based on the short-circuit fault index Z t and the preset fault indicator threshold Z hold Implement fault warning and handling.

[0080] Specifically, this includes: the short-circuit fault index Z. t Is it greater than the fault indicator threshold Z? hold Perform identification; if yes, set the corresponding warning identification state to yes; if no, set the corresponding warning identification state to no; and when the warning identification state is yes, use time t and its corresponding three-phase current (i a(t) i b(t) i c(t) Real-time rotational speed n r(t) Negative sequence current I n(t) Current amplitude I t Short circuit fault index Z t and two-phase current signal S1 α S1 β The corresponding early warning data groups are formed and sent to the preset online early warning interface for short circuit faults.

[0081] Here, the fault index threshold Z in this embodiment of the invention hold The threshold parameter is a pre-set value and can be customized based on application implementation requirements. The online short-circuit fault early warning interface in this embodiment of the invention is a pre-set early warning processing interface that can be customized based on application implementation requirements; for example: an early warning data storage interface for a motor monitoring platform; an early warning data analysis interface for a motor monitoring platform; an audible and visual alarm interface for motor monitoring facilities; or an instant message, voice notification, and SMS notification interface for motor maintenance personnel, etc.

[0082] It should also be noted that the following is a brief explanation of the derivation process of the formula for calculating the negative sequence current amplitude given in step 41 in this embodiment of the invention.

[0083] First, due to the change in stator winding impedance parameters after an inter-turn short circuit, a three-phase asymmetry will occur in the three-phase current. According to the symmetrical component method, the asymmetrical three-phase current can be decomposed into positive-sequence I... p Negative order I n There are three components: I0, I0, and I0. However, as long as the fault has not yet spread to a ground short circuit, according to Kirchhoff's current law, the sum of the three-phase currents is 0, so there is no zero-sequence current. At this time, the three-phase current i0... a(t) i b(t) i c(t) It can be expressed as the sum of the positive and negative components as follows: .

[0084] in, , where represents the corresponding fundamental angular frequency. Positive sequence component current The corresponding initial phase angle, Negative sequence component current The corresponding initial phase angle.

[0085] The two-phase current i obtained by performing Clark transformation on the above three-phase currents α(t) i β(t) That is: .

[0086] The two-phase current signal S2 obtained in step 3 above α S2 β To obtain an analytical signal that retains only the fundamental component, considering the imaginary part lags the real part by π / 2, the corresponding two-phase currents i α-com(t) i β-com(t) The real and imaginary parts can be expressed as: , .

[0087] Therefore, we can obtain the negative order I. n α and β components i nα(t) i nβ(t) expression: .

[0088] The magnitude of the negative sequence current I is also known. n(t) With α and β components i nα(t) i nβ(t) The calculation formula is: .

[0089] Then, the component i nα(t) i nβ(t)Substituting the expression into the above calculation formula, we can obtain: .

[0090] Figure 2 This is a module structure diagram of an online detection device for stator short-circuit faults in a motor under transient operating conditions, provided in Embodiment 2 of the present invention. This device can be a terminal device or server implementing the aforementioned method embodiments, or it can be a device that enables the aforementioned terminal device or server to implement the aforementioned method embodiments. For example, the device can be a device or chip system of the aforementioned terminal device or server. Figure 2 As shown, the device includes: a data acquisition module 201, a fundamental target frequency band processing module 202, an improved Hilbert transform module 203, a negative sequence current and short-circuit fault index processing module 204, and a fault early warning module 205.

[0091] Data acquisition module 201 is based on a preset sampling frequency f s The three-phase currents (a, b, c) of the motor stator and the real-time speed are continuously collected, and the three-phase currents and real-time speed collected at time t are recorded as i. a(t) i b(t) i c(t) n r(t) ; where time zero t=0 is the start time of the motor's current operating cycle.

[0092] The fundamental frequency band processing module 202 is used to process the three-phase current i a(t) i b(t) i c(t) The Clark transformation is performed to obtain the two-phase current i in the αβ two-phase stationary coordinate system. α(t) i β(t) ; and based on the two-phase current i α(t) i β(t) Calculate the current amplitude I t And based on the current amplitude I t Real-time rotational speed n r(t) and the preset rated current frequency f N Rated speed n N Rated current amplitude I N Estimating the fundamental frequency f based on no-load speed n0 and no-load current amplitude I0 1(t) Based on the fundamental frequency f 1(t) Determine the target frequency band [f min(t) ,f max(t) ];f min(t) f max(t) These are the lower and upper thresholds of the target frequency band, respectively.

[0093] The improved Hilbert transform module 203 is used for M two-phase currents i from time t-(M-1) to time t.α(t) i β(t) The corresponding two-phase current signal S1 is formed α S1 β And for the two-phase current signal S1 α S1 β Performing Fast Fourier Transform on each signal yields the corresponding two-phase frequency domain signals X1. α X1 β And construct a Hilbert filter for the two-phase frequency domain signal X1. α X1 β Negative frequency filtering is performed separately to obtain the corresponding two-phase frequency domain signals X2. α X2 β ; and based on the target frequency band [f min(t) ,f max(t) Construct a bandpass filter for the two-phase frequency domain signal X2 α X2 β Non-target frequency filtering is performed to obtain the corresponding two-phase frequency domain signal X3 α X3 β And for the two-phase frequency domain signal X3 α X3 β Perform inverse fast Fourier transform on each phase to obtain the corresponding two-phase current signals S2. α S2 β ; and from the two-phase current signal S2 α S2 β Extract the two-phase current i corresponding to time t α-com(t) i β-com(t) The total number of sampling points M is a preset positive integer; the two-phase current i α-com(t) i β-com(t) It is a complex current.

[0094] Negative sequence current and short-circuit fault index processing module 204 is based on two-phase current i α-com(t) i β-com(t) The negative sequence current I is obtained by calculating the negative sequence current amplitude. n(t) And based on the two-phase current i α-com(t) i β-com(t) For current amplitude I t Perform a reset, based on the negative sequence current I. n(t) and the current amplitude I after reset t Calculate the corresponding short-circuit fault index Z t .

[0095] The fault early warning module 205 is based on the short-circuit fault index Z. t and the preset fault indicator threshold Z hold Implement fault warning and handling.

[0096] The present invention provides an online detection device for stator short-circuit faults of motors under transient operating conditions, which can execute the method steps in the above method embodiments. Its implementation principle and technical effect are similar, and will not be repeated here.

[0097] It should be noted that the division of the various modules in the above device is merely a logical functional division. In actual implementation, they can be fully or partially integrated into a single physical entity, or they can be physically separated. Furthermore, these modules can be implemented entirely in software via processing element calls; they can be fully implemented in hardware; or some modules can be implemented by processing element calls to software, while others are implemented in hardware. For example, the data acquisition module can be a separate processing element, or it can be integrated into a chip in the above device. Alternatively, it can be stored as program code in the memory of the above device, and called and executed by a processing element of the device. The implementation of other modules is similar. Moreover, these modules can be fully or partially integrated together, or they can be implemented independently. The processing element described here can be an integrated circuit with signal processing capabilities. In the implementation process, each step of the above method or each of the above modules can be completed through integrated logic circuits in the hardware of the processor element or through software instructions.

[0098] For example, these modules can be one or more integrated circuits configured to implement the above methods, such as one or more Application Specific Integrated Circuits (ASICs), one or more Digital Signal Processors (DSPs), or one or more Field Programmable Gate Arrays (FPGAs). As another example, when a module is implemented using processing element scheduler code, the processing element can be a general-purpose processor, such as a Central Processing Unit (CPU) or other processor capable of calling program code. Furthermore, these modules can be integrated together as a System-on-a-Chip (SOC).

[0099] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented, in whole or in part, as a computer program product. This computer program product includes one or more computer instructions. When these computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the foregoing method embodiments are generated. The computer described above can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The aforementioned computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the aforementioned computer instructions can be transmitted from one website, computer, server, or data center to another via wired (e.g., coaxial cable, fiber optic, Digital Subscriber Line (DSL)) or wireless (e.g., infrared, wireless, Bluetooth, microwave, etc.) means. The aforementioned computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that integrates one or more available media. The aforementioned available media can be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., DVDs), or semiconductor media (e.g., solid-state disks (SSDs)).

[0100] This invention provides an online detection method and apparatus for short-circuit faults in the stator of a motor under transient operating conditions. As described above, this invention performs Clark transform on the real-time three-phase current of the motor stator to obtain the αβ two-phase current in a two-phase stationary coordinate system. Based on the αβ two-phase current, the current amplitude is calculated, and the fundamental frequency is estimated based on the current amplitude and real-time speed. The target frequency band is then determined based on the fundamental frequency. Next, a two-phase current signal is composed of historical and real-time data of the two-phase current using a sliding window mechanism. This two-phase current signal is then converted into a two-phase frequency domain signal using a Fast Fourier Transform. Finally, a Hilbert filter is constructed to perform negative frequency filtering on the frequency domain signal. The Hilbert-filtered signal is filtered at a non-target frequency using a bandpass filter constructed from the target frequency band. The filtered two-phase frequency domain signal is then converted into a two-phase current signal represented in complex form using an inverse fast Fourier transform. This two-phase current signal can be approximated as the fundamental wave signal. The current at the last moment of the two-phase fundamental wave signal is then considered as the instantaneous fundamental wave component. Based on the instantaneous fundamental wave component, the negative-sequence current amplitude and the fundamental current amplitude are calculated, and a short-circuit fault index is calculated based on the negative-sequence / fundamental current amplitude. Finally, a fault warning is performed based on the real-time short-circuit fault index using a threshold method. This embodiment of the invention improves the detection sensitivity and reliability of short-circuit fault detection under transient operating conditions.

[0101] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein can be implemented in hardware, a software module executed by a processor, or a combination of both. The software module can be located in random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art.

[0102] The specific embodiments described above further illustrate the purpose, technical solution, and beneficial effects of the present invention. It should be understood that the above description is only a specific embodiment of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for online detection of stator short-circuit faults in motors under transient operating conditions, characterized in that, The method includes: Based on the preset sampling frequency f s The three-phase currents (a, b, c) of the motor stator and the real-time speed are continuously collected, and the three-phase currents and real-time speed collected at time t are recorded as i. a(t) i b(t) i c(t) n r(t) Wherein, time zero t=0 is the starting time of the motor in this operating cycle; For three-phase current i a(t) i b(t) i c(t) The Clark transformation is performed to obtain the two-phase current i in the αβ two-phase stationary coordinate system. α(t) i β(t) ; and based on the two-phase current i α(t) i β(t) Calculate the current amplitude I t ; and based on the current amplitude I t Real-time rotational speed n r(t) and the preset rated current frequency f N Rated speed n N Rated current amplitude I N Estimating the fundamental frequency f based on no-load speed n0 and no-load current amplitude I0 1(t) ; and based on the fundamental frequency f 1(t) Determine the target frequency band [f min(t) ,f max(t) ];f min(t) f max(t) These are the lower and upper thresholds of the target frequency band, respectively. The M two-phase currents i from time t-(M-1) to time t α(t) i β(t) The corresponding two-phase current signal S1 is formed α S1 β ; and the two-phase current signal S1 α S1 β Performing Fast Fourier Transform on each signal yields the corresponding two-phase frequency domain signals X1. α X1 β And construct a Hilbert filter for the two-phase frequency domain signal X1. α X1 β Negative frequency filtering is performed separately to obtain the corresponding two-phase frequency domain signals X2. α X2 β ; and based on the target frequency band [f min(t) ,f max(t) Construct a bandpass filter for the two-phase frequency domain signal X2 α X2 β Non-target frequency filtering is performed to obtain the corresponding two-phase frequency domain signal X3 α X3 β ; and the two-phase frequency domain signal X3 α X3 β Perform inverse fast Fourier transform on each phase to obtain the corresponding two-phase current signals S2. α S2 β ; and from the two-phase current signal S2 α S2 β Extract the two-phase current i corresponding to time t α-com(t) i β-com(t) The total number of sampling points M is a preset positive integer; the two-phase current i α-com(t) i β-com(t) It is a complex current; Based on the two-phase current i α-com(t) i β-com(t) The negative sequence current I is obtained by calculating the negative sequence current amplitude. n(t) And based on the two-phase current i α-com(t) i β-com(t) For the current amplitude I t Perform a reset, based on the negative sequence current I. n(t) and the current amplitude I after reset t Calculate the corresponding short-circuit fault index Z t ; Based on the short-circuit fault index Z t and the preset fault indicator threshold Z hold Implement fault warning and handling.

2. The online detection method for stator short-circuit faults in motors under transient operating conditions according to claim 1, characterized in that, The three-phase current i a(t) i b(t) i c(t) and the two-phase current i α(t) i β(t) All are real currents; from the three-phase current i a(t) i b(t) i c(t) to the two-phase current i α(t) i β(t) The Clark transformation method is as follows: ， ； Among them, C 3 / 2 This is the Clark transformation matrix; The current amplitude I t The calculation method is as follows: ； The fundamental frequency f 1(t) The estimation method is as follows: ； The target frequency band [f min(t) ,f max(t) The method for determining ] is as follows: ， ； Where △f is a preset adjustable frequency band parameter.

3. The online detection method for stator short-circuit faults in motors under transient operating conditions according to claim 1, characterized in that, The two-phase current signal S1 α S1 β Current signal S1 α Specifically expressed as Current signal S1 β Specifically expressed as Two-phase current signal , of , Let be the two-phase real current corresponding to the u-th time sampling point; u=1 corresponds to time t-(M-1), then the current , Let i be the two-phase current at time t-(M-1). α(t-(M-1)) i β(t-(M-1)) ; if u=M corresponds to time t, then the current , Let i be the two-phase current at time t. α(t) i β(t) ; 1 ≤ sampling point index u ≤ M, sampling time interval Δt = 1 / f between every two time sampling points s ; The two-phase frequency domain signal X1 α X1 β frequency domain signal X1 α Specifically expressed as Frequency domain signal X1 β Specifically expressed as Two-phase frequency domain signals , of , The two-phase frequency domain characteristics corresponding to the u-th frequency domain sampling point; Two-phase frequency domain characteristics , It is in complex form; The two-phase frequency domain signal X2 α X2 β frequency domain signal X2 α Specifically expressed as Frequency domain signal X2 β Specifically expressed as Two-phase frequency domain signals , of , The two-phase frequency domain characteristics corresponding to the u-th frequency domain sampling point; Two-phase frequency domain characteristics , It is in complex form; The two-phase frequency domain signal X3 α X3 β frequency domain signal X3 α Specifically expressed as Frequency domain signal X3 β Specifically expressed as Two-phase frequency domain signals , of , The two-phase frequency domain characteristics corresponding to the u-th frequency domain sampling point; Two-phase frequency domain characteristics , It is in complex form; The two-phase current signal S2 α S2 β Current signal S2 α Specifically expressed as Current signal S2 β Specifically expressed as Two-phase current signal , of , Let be the two-phase complex current corresponding to the u-th time sampling point; two-phase complex current , It is in complex form; Complex current Specifically, it can be expressed as follows: , , These correspond to the real and imaginary parts, respectively; complex current. Specifically, it can be expressed as follows: , , These represent the real and imaginary parts, respectively; j is the imaginary unit. The two-phase current i α-com(t) i β-com(t) Specifically, it can be expressed as follows: ， ； , They are the currents i α-com(t) The real and imaginary parts, , They are the currents i β-com(t) The real and imaginary parts.

4. The online detection method for stator short-circuit faults in motors under transient operating conditions according to claim 3, characterized in that, The two-phase current signal S1 α S1 β Performing Fast Fourier Transform on each signal yields the corresponding two-phase frequency domain signals X1. α X1 β Specifically, it includes: Step 41, based on the sampling frequency f s The sampling bandwidth Δf of the Fast Fourier Transform is determined by the total number of sampling points M. FFT and frequency range; and based on the frequency range and the sampling bandwidth Δf FFT Confirm M frequency domain sampling points; Among them, △f FFT =f s / M; the lower limit of the frequency range is 0Hz and the upper limit is f s -△f FFT ; The total number of frequency domain sampling points in the frequency range is consistent with the total number of time domain sampling points of the current signal; Step 42, for the two-phase current signals , Performing Fast Fourier Transform on each signal yields the corresponding two-phase frequency domain signals. , : ， ； FFT() is the Fast Fourier Transform function.

5. The online detection method for stator short-circuit faults in motors under transient operating conditions according to claim 3, characterized in that, The constructed Hilbert filter is applied to the two-phase frequency domain signal X1 α X1 β Negative frequency filtering is performed separately to obtain the corresponding two-phase frequency domain signals X2. α X2 β Specifically, it includes: Step 51, construct the Hilbert filter H u : ； Step 52, based on the Hilbert filter H u For the two-phase frequency domain signal X1 α X1 β Negative frequency filtering is performed separately to obtain the corresponding two-phase frequency domain signals X2. α X2 β : ， 。 6. The online detection method for stator short-circuit faults in motors under transient operating conditions according to claim 3, characterized in that, The target frequency band [f] min(t) ,f max(t) Construct a bandpass filter for the two-phase frequency domain signal X2 α X2 β Non-target frequency filtering is performed to obtain the corresponding two-phase frequency domain signal X3 α X3 β Specifically, it includes: Step 61, based on the target frequency band [f min(t) ,f max(t) and the sampling frequency f s Calculate the corresponding lower bound index u min and upper limit index u max : ， ； floor() is the floor function, and ceiling() is the ceiling function. Step 62, based on the lower bound index u min and the upper limit index u max Construct a bandpass filter P u : ； Step 63, based on the bandpass filter P u For the two-phase frequency domain signal X2 α X2 β The corresponding two-phase frequency domain signals X3 are obtained by performing non-target frequency filtering respectively. α X3 β : ， 。 7. The online detection method for stator short-circuit faults in motors under transient operating conditions according to claim 3, characterized in that, The two-phase frequency domain signal X3 α X3 β Perform inverse fast Fourier transform on each phase to obtain the corresponding two-phase current signals S2. α S2 β Specifically, it includes: For the two-phase frequency domain signals , Perform inverse fast Fourier transform on each phase to obtain the corresponding two-phase current signals. , : ， ； IFFT() is the inverse fast Fourier transform function.

8. The online detection method for stator short-circuit faults in motors under transient operating conditions according to claim 3, characterized in that, The two-phase current signal S2 α S2 β Extract the two-phase current i corresponding to time t α-com(t) i β-com(t) Specifically, it includes: Current signal The Mth complex current As the corresponding current i α-com(t) ; and the current signal The Mth complex current As the corresponding current i β-com(t) ; in, ， ， ； ， ， ； The current i α-com(t) real part virtual part The complex currents are respectively real part virtual part ; The current i β-com(t) real part virtual part The complex currents are respectively real part virtual part .

9. The online detection method for stator short-circuit faults in motors under transient operating conditions according to claim 3, characterized in that, The basis of the two-phase current i α-com(t) i β-com(t) The negative sequence current I is obtained by calculating the negative sequence current amplitude. n(t) And based on the two-phase current i α-com(t) i β-com(t) For the current amplitude I t Perform a reset, based on the negative sequence current I. n(t) and the current amplitude I after reset t Calculate the corresponding short-circuit fault index Z t Specifically, it includes: Step 91, convert the two-phase current i α-com(t) i β-com(t) Substituting the two sets of real and imaginary parts into the formula for calculating the negative sequence current amplitude, the corresponding negative sequence current I is obtained. n(t) : The formula for calculating the negative sequence current amplitude is as follows: ； Step 92, based on the two-phase current i α-com(t) i β-com(t) The real part of the current amplitude I t Reset: ； Step 93, based on the negative sequence current I n(t) and the current amplitude I t Calculate the short-circuit fault index Z t : 。 10. The online detection method for stator short-circuit faults in motors under transient operating conditions according to claim 3, characterized in that, The short-circuit fault index Z t and the preset fault indicator threshold Z hold Fault early warning processing includes: For the short-circuit fault index Z t Is it greater than the fault indicator threshold Z? hold Perform identification; if yes, set the corresponding warning identification state to yes; if no, set the corresponding warning identification state to no; and when the warning identification state is yes, use time t and its corresponding three-phase current (i a(t) i b(t) i c(t) The real-time rotational speed n r(t) The negative sequence current I n(t) The current amplitude I t The short-circuit fault index Z t and the two-phase current signal S1 α S1 β The corresponding early warning data groups are formed and sent to the preset online early warning interface for short circuit faults.

11. An apparatus for performing the online detection method for stator short-circuit faults of a motor under transient operating conditions as described in any one of claims 1-10, characterized in that, The device includes: a data acquisition module, a fundamental target frequency band processing module, an improved Hilbert transform module, a negative sequence current and short-circuit fault index processing module, and a fault early warning module. The data acquisition module is based on a preset sampling frequency f s The three-phase currents (a, b, c) of the motor stator and the real-time speed are continuously collected, and the three-phase currents and real-time speed collected at time t are recorded as i. a(t) i b(t) i c(t) n r(t) Wherein, time zero t=0 is the starting time of the motor in this operating cycle; The fundamental target frequency band processing module is used to process the three-phase current i a(t) i b(t) i c(t) The Clark transformation is performed to obtain the two-phase current i in the αβ two-phase stationary coordinate system. α(t) i β(t) ; and based on the two-phase current i α(t) i β(t) Calculate the current amplitude I t ; and based on the current amplitude I t Real-time rotational speed n r(t) and the preset rated current frequency f N Rated speed n N Rated current amplitude I N Estimating the fundamental frequency f based on no-load speed n0 and no-load current amplitude I0 1(t) ; and based on the fundamental frequency f 1(t) Determine the target frequency band [f min(t) ,f max(t) ];f min(t) f max(t) These are the lower and upper thresholds of the target frequency band, respectively. The improved Hilbert transform module is used to process M two-phase currents i from time t-(M-1) to time t. α(t) i β(t) The corresponding two-phase current signal S1 is formed α S1 β ; and the two-phase current signal S1 α S1 β Performing Fast Fourier Transform on each signal yields the corresponding two-phase frequency domain signals X1. α X1 β And construct a Hilbert filter for the two-phase frequency domain signal X1. α X1 β Negative frequency filtering is performed separately to obtain the corresponding two-phase frequency domain signals X2. α X2 β ; and based on the target frequency band [f min(t) ,f max(t) Construct a bandpass filter for the two-phase frequency domain signal X2 α X2 β Non-target frequency filtering is performed to obtain the corresponding two-phase frequency domain signal X3 α X3 β ; and the two-phase frequency domain signal X3 α X3 β Perform inverse fast Fourier transform on each phase to obtain the corresponding two-phase current signals S2. α S2 β ; and from the two-phase current signal S2 α S2 β Extract the two-phase current i corresponding to time t α-com(t) i β-com(t) The total number of sampling points M is a preset positive integer; the two-phase current i α-com(t) i β-com(t) It is a complex current; The negative sequence current and short-circuit fault index processing module is based on the two-phase current i α-com(t) i β-com(t) The negative sequence current I is obtained by calculating the negative sequence current amplitude. n(t) And based on the two-phase current i α-com(t) i β-com(t) For the current amplitude I t Perform a reset, based on the negative sequence current I. n(t) and the current amplitude I after reset t Calculate the corresponding short-circuit fault index Z t ; The fault early warning module is based on the short-circuit fault index Z. t and the preset fault indicator threshold Z hold Implement fault warning and handling.

Images