Arc fault detection method and device

Abstract

The present invention discloses a method and device for detecting arc faults. By setting up two sensing schemes with different frequency response characteristics as control groups, the output correlation of the control group is strong when the line is normal, with the fundamental current being the primary component. However, after an arc fault occurs, the correlation between the outputs of the two control groups weakens due to significant changes in the high-frequency components of the line. The present invention uses this difference as the primary characteristic of the arc fault. This invention addresses the problem that arc fault detection based on sudden changes in the preceding and following current cycles is susceptible to misjudgment due to load startup and mode switching.

Description

Technical Field

[0001] The present invention relates to the field of electrical safety detection, and in particular to a fault arc detection method and device. Background Art

[0002] Arc faults are a significant cause of electrical fires. When an arc fault occurs in a line, the normal current waveform is distorted. This distortion is random and unstable. In engineering inspections, the difference between the current cycle and the historical cycle is often calculated as an important feature of arc fault detection.

[0003] Most current solutions for detecting arc faults involve continuous sampling of line currents. A current waveform data set is selected at intervals of several cycles or fixed time intervals as a reference data set for normal times. Reference characteristic data is calculated, and then characteristic data for the current cycle is calculated. Certain judgment rules are used to mark whether the current cycle is an arc fault cycle and to count the fault cycles. When the count value reaches a set value, it is determined that an arc fault has occurred in the line.

[0004] However, the above solution is based on a current sampling device, and uses the current cycle and the historical cycle as control groups; this solution is easily affected by equipment startup, equipment operating mode, and power changes, resulting in inaccurate detection results. Summary of the Invention

[0005] In order to overcome the current technical problem of inaccurate arc fault detection, the present invention provides a method and device for accurately detecting arc faults.

[0006] In order to achieve the above technical purpose, the technical solution of the present invention is:

[0007] A method for detecting an arc fault comprises the following steps:

[0008] Step 1: synchronously sampling the voltage and current of the AC line, wherein the current sampling includes two current sampling schemes, a first current sampling scheme and a second current sampling scheme, which have different high-frequency responses, wherein the first current sampling scheme is a sampling scheme that can accurately reflect the magnitude of the line current;

[0009] Step 2: The time between two voltage zero crossing points in the voltage sampling result is taken as a cycle;

[0010] Step 3, obtaining a current sampling sequence I1 of the first current sampling scheme and a current sampling sequence I2 of the second current sampling scheme in the first cycle;

[0011] Step 4, calculate the effective value ratio of I1 to I2 as the proportional registration factor K;

[0012] Step 5: Get the current sampling sequence I1 of the second cycle* and I2 * ;

[0013] Step 6: I1 * The ratio of K is taken as the new current sequence I3;

[0014] Step 7, count I2 * The number Q of the difference sequence between I2 and I3 is greater than the preset value Thr1, and then I2 * and I3 are subjected to fast Fourier transform respectively to obtain the amplitude sequence D of different frequency components. K and D' K , and then D K and D' K The amplitudes of the harmonic components in are differentiated by order, and the sum of the squares of the differences W is obtained;

[0015] Step 8: Determine whether Q is greater than a preset value α, and whether W is greater than a preset value β. If both conditions are met, then the number of fault arc cycles T is increased by 1, otherwise T is decreased by 1.

[0016] Step 9: If the arc fault cycle count T is greater than the preset value T0, it is determined that an arc fault has occurred on the line, and the arc fault cycle is reset to zero, and the loop is returned to step 3; otherwise, the current sampling sequence I1 of the second cycle is reset to zero. * and I2 * As the first cycle sequence I1 and I2; and return to step 4 to execute the loop.

[0017] In the method, in step 1, the first current sampling scheme uses a current sensor that can accurately reflect the magnitude of the line current; the second current sampling scheme uses a current sensor whose high-frequency response is different from that of the first current sampling scheme, serving as a control group for the first current sampling scheme.

[0018] The method, in step 2, determines the period is v n-1 ·v n+1 When <0, then v n is the position of the zero crossing point, and the period between two zero crossing points is considered as a cycle, where v n is the nth voltage sampling point.

[0019] The method, in step 5, when obtaining the current sampling sequence I1 of the second cycle * and I2 * Afterwards, the phase difference between I1 and I2 in the first cycle is used as the phase registration factor θ, and I1 * and I2 * The phase advance is one step back by θ phases.

[0020] The method described above, the step of tracing back θ phases, is that when I1 is ahead of I2, I2 * Take the complete second cycle sampling sequence, and I1 * Then trace back θ phases; when I1 lags behind I2, I1 * Take the complete second cycle sampling sequence, and I2 * Then trace back θ phases.

[0021] The method, in step 7, the D K and D' K In the equation, K = 0, 1, 2, ..., N / 2-1; where N is the number of sampling points;

[0022] A fault arc detection device, used in the aforementioned method, is arranged on an AC line and includes a voltage sampling device and two current sampling devices, wherein one of the two current sampling devices is a high-precision current sensor such as a manganese copper resistor, and the other is a current sensor made of a magnetic core with a magnetic permeability less than 100.

[0023] The technical effect of the present invention lies in that it proposes setting up two sensing schemes with different frequency response characteristics as control groups. These control schemes are compared simultaneously and are not affected by factors such as equipment startup and power changes. Because the fundamental current is the primary component when the line is normal, the output correlation of this control group is strong. However, after a fault arc occurs in the line, the high-frequency components in the line undergo significant changes, and the correlation between the outputs of the two control groups weakens. Therefore, by using this difference as the primary characteristic of the fault arc, the present invention can simplify complex judgment rules and improve detection accuracy.

[0024] The present invention also proposes a phase and amplitude registration scheme, which can compensate for the phase and amplitude differences of the original output sequences of the two sensing schemes and improve the accuracy of the control group difference calculation. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 A schematic diagram of a flow chart of an embodiment of the present invention;

[0026] Figure 2 Schematic diagram of phase registration according to an embodiment of the present invention;

[0027] Figure 3 Schematic diagram of a device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0028] See also Figure 3 In this embodiment, three sampling channels are arranged on the AC line to perform corresponding sampling, wherein the three sampling channels are one voltage sampling channel and two current sampling channels, and the three sampling channels are synchronous sampling.

[0029] Voltage sampling is used to determine the frequency of the AC signal and to perform phase alignment on the two current sampling waveforms.

[0030] The current sensor in the first current sampling link is implemented using a high-precision current sensor, such as a manganese copper resistor, to accurately reflect the magnitude of the line current.

[0031] The second current sampling link is set up to serve as a control group for the first current sampling link. The current sensor used therein can be made of a magnetic core with a magnetic permeability of less than 100, such as iron, iron powder core, sendust, etc. As a result, there are significant differences in the frequency response, especially the high-frequency response, of the sensing schemes of the first current sampling link and the second current sampling link. Because when a fault arc occurs in the line, the high-frequency components in the line will change significantly, which can cause the waveforms of the two current samples to show significant differences. Therefore, when the line is normal, the current waveform obtained by the first current sampling is highly correlated with the current waveform obtained by the second current sampling; when a fault arc exists on the line, the correlation between the current waveform obtained by the first current sampling and the current waveform obtained by the second current sampling is weakened.

[0032] Determine the position of the voltage zero crossing point according to the voltage sampling sequence, and the determination method is v n-1 ·v n+1 <0, then v n The position of the zero crossing point is recorded as a cycle, and the current sampling sequence of two cycles is cached.

[0033] Then, a first current sampling sequence I1 and a second current sampling sequence I2 of the first cycle are obtained in the buffer pool.

[0034] In this embodiment, considering that the two current sampling sensing schemes are different, especially the second current sampling sensing scheme adopts a special design, it is very likely that there will be a phase difference between the two sampling sequences. In order to eliminate the impact of the phase difference on subsequent calculations, this embodiment adopts a phase alignment method, that is, calculating the phase difference of I1 relative to I2, recording it as the phase alignment factor θ, and eliminating the phase difference based on θ. The specific method of determining the phase difference in this embodiment is: the method of determining the reference voltage zero crossing point, that is, I n-1 I n+1 <0, then I n The zero crossing point is used as the basis for determining the phase difference.

[0035] Then obtain the current sampling sequence of the second cycle, which is recorded as I1 * and I2 * , here, since there may be a phase difference between the two sequences, that is, the phase registration factor θ calculated above. Then Figure 2 As shown, when I1 is ahead of I2, I2 * Take the complete second cycle sampling sequence, and I1 * Then trace back θ phases. Similarly, when I1 lags behind I2, I1 * Take the complete second cycle sampling sequence, and I2 * Then trace back θ phases.

[0036] At the same time, due to the special design of the second current sampling sensing scheme and the different scaling factors of the conditioning circuit, the amplitudes of the two sequences will be different, but they are generally linear. The ratio of the effective value of I1 to I2 is calculated and recorded as the ratio registration factor K.

[0037] When the current sampling sequence of the second cycle is obtained, it is recorded as I1 * and I2 * Afterwards, according to I1 * Calculate the new current sequence

[0038] The obtained I2 * The difference between the phase and I3 is Δ i =|(I2 i *-I3 i )|, record the difference sequence Δ i The number Q of the median values ​​being greater than the preset value Thr1.

[0039] I2 * Perform fast Fourier transform to obtain the amplitude sequence D of different frequency components k (k=0,1,2…,N / 2-1); the amplitude of the DC component is D0, and the amplitude of the fundamental wave is D1. Perform fast Fourier transform on I3 to obtain the amplitude sequence D' of different frequency components k (k=0,1,2…,N / 2-1); D K and D' K The amplitudes of the 3rd to 40th harmonic components are subtracted by order, and the sum of the squares of the differences is obtained.

[0040] Determine whether Q is greater than the preset value α, and whether W is greater than the preset value β. If both conditions are met, the number of fault arc cycles T = T+1, otherwise T = T-1, T≥0. In other words, Q represents the difference in the time domain between the current sequences obtained by the two sensing schemes with different frequency response characteristics after a fault arc occurs in the line, and W represents the difference in the frequency domain between the two sensing schemes with different frequency response characteristics after a fault arc occurs in the line. At the same time, T here is a count value, which means that there is an arc in this current cycle. If the T value exceeds the threshold, it means that the arc duration exceeds the threshold, and the energy released by the arc accumulates and will cause harm. If the fault arc cycle count T is greater than the preset value T0, it is determined that a fault arc has occurred on the line, and the fault arc cycle is cleared. If the arc cycle count T is not greater than the preset value T0, the current sampling sequence I1 of the current second cycle is reset. * and I2 * Update to the first periodic sequences I1 and I2; and repeat the above process.

[0041] This invention eliminates the existing practice of installing a single current sensor on a phase (L or N) of an active power line, continuously sampling the line current and using the difference between the current and historical cyclic currents as a key characteristic of a fault arc. Instead, it proposes setting up two sensing schemes with different frequency response characteristics as control groups. When the line is normal, the fundamental current is the primary component, and the output correlation of this control group is strong. However, after a fault arc occurs, the correlation between the outputs of the two control groups weakens due to significant changes in the high-frequency components of the line. The present invention uses this difference as the primary characteristic of the fault arc.

[0042] Because the two sensing schemes have different formulations, that is, the original sampling sequences will have phase and amplitude differences, this patent proposes to determine the cycle start and end positions by using the AC voltage zero crossing point, and calculate the phase difference between the two first-cycle current sequences I1 and I2 to obtain the phase alignment factor; determine the proportional alignment factor by the ratio of the current cycle effective values; use these two alignment factors to transform the obtained second-cycle sequence, and then calculate the time domain and frequency domain differences of the transformed current sequence. Finally, the fault arc period is determined based on the differences in the time domain and frequency domain.

[0043] After determining that the cycle is a fault arc, the current sequence I1 of the current second cycle is * and I2 *The first cycle sequence is updated and the detection process is repeated. This approach has the advantage that, when the line is normal, repeated registration can gradually eliminate the errors between the sensing schemes in the control group, lowering the thresholds of the preset values ​​α and β. However, after a fault arc occurs in the line, since the fault arc is a strongly random process and the frequency components in the line are unstable, the registration process will not be effective and may even exacerbate the differences between the control groups, making the comparison more obvious and easier to identify.

Claims

1. A method for detecting arc faults, characterized in that: The following steps are involved: Step 1: synchronously sampling the voltage and current of the AC line, wherein the current sampling includes two current sampling schemes, a first current sampling scheme and a second current sampling scheme, which have different high-frequency responses, wherein the first current sampling scheme is a sampling scheme that can accurately reflect the magnitude of the line current; The first current sampling scheme uses a current sensor that can accurately reflect the magnitude of the line current; the second current sampling scheme uses a current sensor that has a different high-frequency response from the first current sampling scheme, serving as a control group for the first current sampling scheme; Step 2: The time between two voltage zero crossing points in the voltage sampling result is taken as a cycle; Step 3, obtaining a current sampling sequence I1 of the first current sampling scheme and a current sampling sequence I2 of the second current sampling scheme in the first cycle; Step 4, calculate the effective value ratio of I1 to I2 as the proportional registration factor K; Step 5: Get the current sampling sequence of the second cycle and ; Among them, when obtaining the current sampling sequence of the second cycle and Afterwards, the phase difference between I1 and I2 in the first cycle is used as the phase registration factor θ, and and The phase advance is a step back by θ phases; Step 6, The ratio of K is taken as the new current sequence I3; Step 7, statistics The number Q of the difference sequence between I and I3 that is greater than the preset value Thr1, and then and I3 are subjected to fast Fourier transform respectively to obtain the amplitude sequence D of different frequency components. K and D' K , and then D K and D' K The amplitudes of the harmonic components in are differentiated by order, and the sum of the squares of the differences W is obtained; Step 8: Determine whether Q is greater than a preset value α, and whether W is greater than a preset value β. If both conditions are met, then the number of fault arc cycles T is increased by 1, otherwise T is decreased by 1. Step 9: If the arc fault cycle count T is greater than the preset value T0, it is determined that an arc fault has occurred on the line, and the arc fault cycle is reset to zero, and the loop is returned to step 3; otherwise, the current sampling sequence of the second cycle is reset to zero. and As the first cycle sequence I1 and I2; and return to step 4 to execute the loop.

2. The method according to claim 1, characterized in that In step 2, the period is determined based on v n-1 ·v n+1 When <0, then v n is the position of the zero crossing point, and the period between two zero crossing points is considered as a cycle, where v n is the nth voltage sampling point.

3. The method according to claim 1, characterized in that The step of θ phases is to look back when I1 is ahead of I2. Take the complete second cycle sampling sequence, and Then trace back θ phases; when I1 lags behind I2, Take the complete second cycle sampling sequence, and Then go back phases.

4. The method according to claim 1, wherein In the step 7, the D K and D' K In the equation, K=0,1,2,...,N / 2-1; where N is the number of sampling points; .

5. A fault arc detection device, used in the method according to any one of claims 1 to 4, characterized in that: A voltage sampling device and two current sampling devices are arranged on the AC line. Among the two current sampling devices, one is a high-precision current sensor such as a manganese copper resistor, and the other is a current sensor made of a magnetic core with a magnetic permeability less than 100.

Images