A method and device for locating the start and end time of voltage sag waveform

Abstract

The present invention provides a method and device for locating the start and end time of a voltage sag waveform, the method comprising the following steps: step 1, using the maximum stack discrete wavelet transform method to calculate the spectral energy of the voltage, the spectral energy of the voltage including the scale coefficient energy and the wavelet coefficient energy; step 2, using the scale coefficient energy as a calculation index for the start and end time of the voltage sag waveform, when the scale coefficient energy is equal to the set reference energy threshold, determining the sampling point on the waveform corresponding to the scale coefficient energy as the start or end time of the voltage sag. The present invention uses the maximum stack discrete wavelet transform method to calculate the spectral energy of the voltage, and determines the waveform start and end time of the voltage sag event through the change of the spectral energy at different times. This method can solve the problem that the traditional method cannot truly reflect the start and end time of the voltage sag event, and at the same time, the calculation is more accurate and more suitable for real-time applications in engineering practice.

Description

Technical Field

[0001] The invention relates to the technical field of voltage sag, and in particular to a method for locating the start and end times of a voltage sag waveform based on maximum stacked discrete wavelet transform. Background Art

[0002] Voltage sag is one of the main power quality problems. It is usually caused by short-circuit faults, transformer excitation, and large motor starting, which causes a short-term increase in current, resulting in a short-term decrease in the effective value of voltage. Among them, short-circuit faults are the main cause of voltage sags. The amplitude and duration of voltage sags are two important indicators of voltage sags. Accurately estimating the amplitude and duration of voltage sags is very important to help select voltage sag control equipment and improve equipment immunity.

[0003] At present, the conventional method for calculating the amplitude and duration of voltage sag is to analyze the effective value of voltage and locate the start and end time of voltage sag. Although the above method is simple and effective, the method based on the effective value of voltage may make errors in the estimation of certain event characteristics. For example, when a voltage sag occurs, the effective value of voltage will not drop to a lower value immediately, but will continue for a period of time during the transition process, and the effective value of voltage will not recover immediately after the fault, which will lead to the wrong location of the start and recovery time of voltage sag.

[0004] The traditional method for locating the start and end time of voltage sag based on voltage effective value is as follows:

[0005] Step 1: Calculate the voltage RMS value based on the voltage waveform data. The voltage RMS value is usually measured by the RMS value of a complete cycle of the voltage. rms The calculation formula is:

[0006]

[0007] Where v(m) is the instantaneous value of the voltage at the mth point, and △k is the number of samples in a complete voltage cycle;

[0008] Step 2: According to the calculated voltage RMS value v rms Plot the voltage RMS value versus time, such as Figure 1 As shown;

[0009] Step 3: According to the definition of voltage sag duration, the voltage sag threshold is set to 0.9pu. Then, the starting time of the voltage sag corresponding to the voltage amplitude of 0.9pu can be found as k1. rms , the end time is k2 rms .

[0010] Depend on Figure 1From the voltage waveform shown, it can be seen that the actual start time of the voltage sag corresponds to the waveform point k1, and the actual end time corresponds to the waveform point k2, which is obviously different from the k1 calculated above. rms and k2 rms The inconsistency indicates that the traditional calculation method based on voltage effective value cannot truly reflect the start and end time of voltage sag events. In addition, the traditional method needs to calculate the voltage effective value point by point, so the calculation speed is slow and cannot be applied in real time in engineering practice.

[0011] The inventors of the present application have found through research in the process of realizing the present invention that the spectrum energy of the phase voltage can be well used to characterize the voltage sag, and its performance is exactly the same as that of the method based on the voltage effective value. Before the voltage sag and during the duration of the voltage sag, the phase voltage contains a large number of high-frequency components, which are called transients. Transient analysis can accurately identify the corresponding moments on the voltage waveform when the voltage sag starts and recovers.

[0012] Therefore, the present invention proposes a method for locating the start and end time of a voltage sag waveform based on maximum stacked discrete wavelet transform. The voltage spectrum energy is decomposed by the scale coefficient energy and the wavelet coefficient energy, and the scale coefficient energy of the phase voltage is used to characterize the voltage sag, and a method for locating the start and end time of a voltage sag waveform that meets the definition is proposed. Summary of the invention

[0013] The present invention provides a method and device for locating the start and end time of a voltage sag waveform based on maximum stacking discrete wavelet transform. The method adopts the maximum stacking discrete wavelet transform method to calculate the spectral energy of the voltage, and determines the waveform start and end time of the voltage sag event through the changes in the spectral energy at different times. The method can solve the problem that traditional methods cannot truly reflect the start and end time of the voltage sag event. At the same time, the calculation is more accurate and more suitable for real-time applications in engineering practice.

[0014] A method for locating the start and end time of a voltage sag waveform based on maximum stacking discrete wavelet transform comprises the following steps:

[0015] Step 1: Calculate the spectrum energy of voltage by using the maximum stacking discrete wavelet transform method, wherein the spectrum energy of voltage includes the scale coefficient energy and the wavelet coefficient energy;

[0016] Step 2: Use the scale factor energy as a calculation index for the start and end time of the voltage sag waveform. When the scale factor energy is equal to the set reference energy threshold, determine that the sampling point on the waveform corresponding to the scale factor energy is the start or end time of the voltage sag.

[0017] Furthermore, the step 1 specifically includes:

[0018] The first step is to calculate the spectrum energy ε(k) of a window with a sliding length of △k in the real-time voltage:

[0019]

[0020] Where △k=f s / f is the number of sampling points in one cycle. The corresponding spectrum energy can be calculated according to the phase voltage waveform v. v(m) represents the instantaneous value of the phase voltage v at the mth sampling point.

[0021] In the second step, in the real-time voltage waveform, the spectrum energy of the sliding window is decomposed into the scale factor energy ε s (k) and wavelet coefficient energy ε w (k) as follows:

[0022] ε(k)=ε s (k)+ε w (k)

[0023] Scaling coefficient energy ε s (k) and wavelet coefficient energy ε w (k) consists of two parts, as follows:

[0024] ε s (k) = ε sa (k)+ε sb (k)

[0025] ε w (k) = ε wa (k)+ε wb (k)

[0026] Step 3: Component ε sa and ε wa It is caused by the scale coefficient and wavelet coefficient of the sample boundary effect within the signal sliding window, as shown below:

[0027]

[0028]

[0029] The pyramid algorithm is used to perform the maximum stacking discrete wavelet transform on the voltage waveform with the deepest sag amplitude, and the scale coefficient is calculated. and wavelet coefficients as follows:

[0030]

[0031]

[0032] Where n = 1, 2, ..., L-1, v(k-L+3),...,v(k),v(k-△k+1),...,v(k-△k+L-1)}, is the sequence of the last L-1 and first L-1 samples of the signal sliding window, and They are scale filter and wavelet filter respectively, and L is the number of half-cycle sampling points of the original voltage instantaneous value;

[0033] Component ε sb and ε wb The real-time scale coefficient and wavelet coefficient are calculated as follows:

[0034]

[0035]

[0036] Without considering the boundary effect of the sliding window, the scale coefficient energy and wavelet coefficient energy are defined as follows:

[0037]

[0038]

[0039] Furthermore, the step 2 specifically includes:

[0040] The voltage waveform before the voltage sag event is used to calculate the reference energy E during normal operation. s , as shown below:

[0041]

[0042] Among them, E s is the reference energy; ε(m) represents the spectrum energy calculated at the phase voltage sampling point m; k i represents the phase voltage sampling point within the sliding window;

[0043] By converting the phase voltage scaling factor energy ε s With reference energy E s By comparing the threshold with the voltage drop threshold, the start and end time of the voltage sag can be detected.

[0044] Furthermore, the scale factor energy ε of the phase voltage is s With reference energy E s The start and end time of the voltage sag can be detected by comparing the threshold with the voltage drop, including:

[0045] When the scale factor energy ε s Continues to decrease and is equal to the first reference energy threshold E2E sWhen the sampling point k1 on the corresponding waveform s That is the voltage sag start time T start ; When the scale coefficient energy ε s Continues to increase and is equal to the second reference energy threshold E1E s When the sampling point k2 on the corresponding waveform s That is the voltage sag end time T end , thus we can get the start and end time of voltage sag waveform,

[0046] The details are as follows:

[0047]

[0048]

[0049] Where E1 = V1 2 =0.01, V1=0.1, E2=V2 2 =0.81, V2=0.9, which are the voltage thresholds of voltage sag respectively.

[0050] A device for locating the start and end time of a voltage sag waveform based on maximum stacking discrete wavelet transform, comprising:

[0051] A spectrum energy calculation module, used to calculate the spectrum energy of voltage by using a maximum stacking discrete wavelet transform method, wherein the spectrum energy of voltage includes scale coefficient energy and wavelet coefficient energy;

[0052] The start and end time determination module is used to use the scale factor energy as the calculation index of the start and end time of the voltage sag waveform. When the scale factor energy is equal to the set reference energy threshold, the sampling point on the waveform corresponding to the scale factor energy is determined to be the start time or end time of the voltage sag.

[0053] The present invention utilizes the spectral energy of phase voltage to characterize the characteristics of voltage sag. The maximum stacking discrete wavelet transform method is adopted to calculate the spectral energy of voltage. The waveform start and end times of the voltage sag event are determined by the change of the spectral energy at different times. The waveform points at the start and end times of the voltage waveform of the voltage sag event can be located. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 When the traditional voltage sag start and end time positioning method based on voltage effective value is used, the voltage root mean square value v is calculated. rms Draw a waveform graph of the voltage effective value changing over time;

[0055] Figure 2 This is a waveform diagram of a primary voltage sag monitored by a substation in Wuhan City in an embodiment of the present invention;

[0056] Figure 3 for Figure 2 The calculated voltage effective value;

[0057] Figure 4 The voltage scale coefficient energy calculated by the embodiment of the present invention. DETAILED DESCRIPTION

[0058] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0059] The embodiment of the present invention provides a method for locating the start and end time of a voltage sag waveform based on maximum stacking discrete wavelet transform, comprising the following steps:

[0060] Step 1: Calculate the spectral energy of voltage by using the maximum stacking discrete wavelet transform method. The spectral energy of voltage includes the scale coefficient energy and the wavelet coefficient energy.

[0061] The maximum stacking discrete wavelet transform is an improvement of the discrete wavelet transform. By avoiding downsampling, the calculation time is reduced while ensuring accuracy.

[0062] The maximum stack discrete wavelet transform divides the voltage into scale coefficients (s) and wavelet coefficients (w) through a scale filter and a wavelet filter. By using the maximum stack discrete wavelet transform algorithm, all signal samples to be analyzed (main window) need to calculate the scale coefficients and wavelet coefficients. The scale coefficient is mainly affected by the minimum frequency component of the original voltage, that is, the frequency from the DC component to f s / 4, where f s is the sampling frequency; the wavelet coefficient is mainly affected by the highest frequency component of the original voltage, that is, the frequency from f s / 4 to f s / 2. The wavelet coefficients can be used to detect the starting point and recovery point of the transient change of the original signal.

[0063] The step 1 specifically includes the following steps:

[0064] The first step is to calculate the spectrum energy ε(k) of a window with a sliding length of △k in the real-time voltage (signal sliding window):

[0065]

[0066] Where △k=f s / f is the number of sampling points in one cycle. The corresponding spectrum energy can be calculated based on the phase voltage waveform v. v(m) represents the instantaneous value of the phase voltage v at the mth sampling point.

[0067] In the second step, according to Parseval's theorem, the energy of the signal can be decomposed into the scale coefficient energy and the wavelet coefficient energy. Therefore, in the real-time voltage waveform, the spectrum energy of the sliding window can also be decomposed into the scale coefficient energy ε s (k) and wavelet coefficient energy ε w (k) as follows:

[0068] ε(k)=ε s (k)+ε w (k)

[0069] Scaling coefficient energy ε s (k) and wavelet coefficient energy ε w (k) consists of two parts, as follows:

[0070] ε s (k) = ε sa (k)+ε sb (k)

[0071] ε w (k) = ε wa (k)+ε wb (k)

[0072] Step 3: Component ε sa and ε wa It is caused by the scale coefficient and wavelet coefficient of the sample boundary effect within the signal sliding window, as shown below:

[0073]

[0074]

[0075] The pyramid algorithm is used to perform the maximum stacking discrete wavelet transform on the voltage waveform with the deepest sag amplitude, and the scale coefficient is calculated. and wavelet coefficients as follows:

[0076]

[0077]

[0078] Where n = 1, 2, ..., L-1, v(k-L+3),...,v(k),v(k-△k+1),...,v(k-△k+L-1)}, is the sequence of the last L-1 and first L-1 samples of the signal sliding window, and They are scale filter and wavelet filter respectively, and L is the number of half-cycle sampling points of the original voltage instantaneous value;

[0079] Component ε sb and ε wb It can be calculated based on the real-time scale coefficient and wavelet coefficient as follows:

[0080]

[0081]

[0082] Therefore, without considering the boundary effect of the sliding window, the scale coefficient energy and wavelet coefficient energy are defined as follows:

[0083]

[0084]

[0085] Considering the frequency response of the scaling filter and wavelet filter and the energy decomposition theorem, ε s It is mainly related to the energy of the fundamental frequency component of the signal, while ε w It is related to the energy of the high-frequency component of the signal. The energy of the fundamental frequency component is usually much higher than the energy of the transient high-frequency component. Therefore, the spectrum energy ε is approximately equal to the scale factor energy ε s .

[0086] Step 2: Use the scale factor energy ε s As the calculation index of the start and end time of the voltage sag waveform, when the scale factor energy ε s is equal to the reference energy threshold, then the scale coefficient energy ε s The corresponding sampling point on the waveform is the starting or ending time of the voltage sag.

[0087] First, the voltage waveform before the voltage sag event is used to calculate the reference energy E during normal operation. s , as shown below:

[0088]

[0089] Among them, E s is the reference energy; ε(m) represents the spectrum energy calculated at the phase voltage sampling point m; k i Represents the phase voltage sampling points within the sliding window.

[0090] By converting the phase voltage scaling factor energy ε s With reference energy E sThe threshold value is compared to detect the start and end time of voltage sag. s Continues to decrease and is equal to the first reference energy threshold E2E s When the sampling point k1 on the corresponding waveform s That is the voltage sag start time T start ; When the scale coefficient energy ε s Continues to increase and is equal to the second reference energy threshold E1E s When the sampling point k2 on the corresponding waveform s That is the voltage sag end time T end , thus the start and end time of the voltage sag waveform is obtained. The details are as follows:

[0091]

[0092]

[0093] Where E1 = V1 2 =0.01, V1=0.1, E2=V2 2 =0.81, V2=0.9, which are the voltage thresholds of voltage sag respectively.

[0094] The embodiment of the present invention further provides a device for locating the start and end time of a voltage sag waveform based on maximum stacked discrete wavelet transform, comprising:

[0095] A spectrum energy calculation module, used to calculate the spectrum energy of voltage by using a maximum stacking discrete wavelet transform method, wherein the spectrum energy of voltage includes scale coefficient energy and wavelet coefficient energy;

[0096] The start and end time determination module is used to use the scale factor energy as the calculation index of the start and end time of the voltage sag waveform. When the scale factor energy is equal to the set reference energy threshold, the sampling point on the waveform corresponding to the scale factor energy is determined to be the start time or end time of the voltage sag.

[0097] Case Analysis

[0098] Taking the waveform of a voltage sag monitored at a substation in Wuhan as an example, the waveform of the phase with the deepest voltage sag is selected for analysis. Figure 2 As shown in the waveform, the voltage sag starts at t1 and ends at t2, and t1 is 0.089 seconds and t2 is 0.167 seconds. Figure 3 for Figure 2 The calculated voltage effective value, assuming that the voltage sag threshold is 0.9pu, then the voltage sag start time of the traditional method is t I,1 , the end time is t I,2 , and t I,1 is 0.098 seconds, tI,2 It is 0.188 seconds.

[0099] The voltage scale coefficient energy is calculated by the method proposed in the present invention, and the result is as follows: Figure 4 As shown, the voltage sag starts at t II,1 , the end time is t II,2 , and t II,1 is 0.088 seconds, t II,2 The result comparison is shown in Table 1:

[0100] Table 1

[0101]

[0102] As can be seen from the above table, the errors of the method proposed in the present invention are smaller than those of the traditional method for the positioning of the start and end time of the voltage sag, as well as the calculation of the duration. For the calculation results of the three (start time, end time and duration), the errors of the method proposed in the present invention are reduced by 8.99%, 10.77% and 10.25% respectively, indicating that the present invention can effectively improve the accuracy of the positioning of the start and end time of the voltage sag, and has good engineering application value.

[0103] like Figure 1 As shown in the figure, the traditional method for locating the start and end time of voltage sag is to calculate and judge the voltage effective value. However, this method needs to calculate the root mean square value of a complete cycle of voltage to obtain a voltage effective value. Therefore, when the voltage effective value is equal to 0.9pu, it is not the time when the voltage sag starts and ends. The present invention can effectively solve this problem by calculating the spectrum energy of the voltage waveform to directly locate the waveform points at the time when the voltage sag starts and ends, thereby accurately calculating the start and end time of the voltage sag, so the accuracy is higher.

[0104] At the same time, in engineering practice, it is usually necessary to clearly know the start and end times of the voltage sag, so as to assist the work of relay protection devices and voltage sag control devices. The traditional calculation method cannot meet this requirement. The method proposed in the present invention can effectively solve this problem.

[0105] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed by the present invention should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A method for locating the start and end time of a voltage sag waveform, characterized in that The steps include: Step 1: Calculate the spectrum energy of voltage by using the maximum stacking discrete wavelet transform method, wherein the spectrum energy of voltage includes the scale coefficient energy and the wavelet coefficient energy; Step 2: using the scale factor energy as a calculation index for the start and end time of the voltage sag waveform; when the scale factor energy is equal to the set reference energy threshold, determining the sampling point on the waveform corresponding to the scale factor energy as the start or end time of the voltage sag; The step 1 specifically includes: The first step is to calculate the spectrum energy ε(k) of a window with a sliding length of △k in the real-time voltage: Where △k=f s / f is the number of sampling points in one cycle. The corresponding spectrum energy is calculated according to the phase voltage waveform v. v(m) represents the instantaneous value of the phase voltage v at the mth sampling point. In the second step, in the real-time voltage waveform, the spectrum energy of the sliding window is decomposed into the scale factor energy ε s (k) and wavelet coefficient energy ε w (k), as follows: e(k)=e s (k)+e w (k) Scaling coefficient energy ε s (k) and wavelet coefficient energy ε w (k) consists of two parts, as follows: e s (k)=e sa (k)+e sb (k) e w (k)=e wa (k)+e wb (k) Step 3: Component ε sa and ε wa It is caused by the scale coefficient and wavelet coefficient of the sample boundary effect within the signal sliding window, as shown below: The pyramid algorithm is used to perform the maximum stacking discrete wavelet transform on the voltage waveform with the deepest sag amplitude, and the scale coefficient is calculated. and wavelet coefficients as follows: Where n = 1, 2, ..., L-1, v(k-L+3),...,v(k),v(k-△k+1),...,v(k-△k+L-1)}, is the sequence of the last L-1 and first L-1 samples of the signal sliding window, and They are scale filter and wavelet filter respectively, and L is the number of half-cycle sampling points of the original voltage instantaneous value; Component ε sb and ε wb The real-time scale coefficient and wavelet coefficient are calculated as follows: Without considering the boundary effect of the sliding window, the scale coefficient energy and wavelet coefficient energy are defined as follows: The step 2 specifically includes: The voltage waveform before the voltage sag event is used to calculate the reference energy E during normal operation. s , as shown below: Among them, E s is the reference energy; ε(m) represents the spectrum energy calculated at the phase voltage sampling point m; k i represents the phase voltage sampling point within the sliding window; By converting the phase voltage scaling factor energy ε s With reference energy E s The threshold is compared with the start and end time of the voltage sag to detect the start and end time of the voltage sag; The scale factor energy ε of the phase voltage s With reference energy E s The threshold is compared with the start and end time of the voltage sag, and the specific time includes: When the scale factor energy ε s Continues to decrease and is equal to the first reference energy threshold E2E s When the sampling point k1 on the corresponding waveform s That is the voltage sag start time T start ; When the scale coefficient energy ε s Continues to increase and is equal to the second reference energy threshold E1E s When the sampling point k2 on the corresponding waveform s That is the voltage sag end time T end , thus we can get the start and end time of voltage sag waveform, The details are as follows: Where, E1=V1 2 =0.01, V1=0.1, E2=V2 2 =0.81, V2=0.9, which are the voltage thresholds of voltage sag respectively.

2. A device for locating the start and end time of a voltage sag waveform, characterized in that The method for locating the start and end time of a voltage sag waveform according to claim 1, wherein the device comprises: A spectrum energy calculation module, used to calculate the spectrum energy of voltage by using a maximum stacking discrete wavelet transform method, wherein the spectrum energy of voltage includes scale coefficient energy and wavelet coefficient energy; The start and end time determination module is used to use the scale factor energy as the calculation index of the start and end time of the voltage sag waveform. When the scale factor energy is equal to the set reference energy threshold, the sampling point on the waveform corresponding to the scale factor energy is determined to be the start time or end time of the voltage sag.

Images