Method for extracting voltage sag characteristic indexes and comprehensively evaluating sag severity

Abstract

The present invention discloses a method for extracting characteristic indicators of voltage sag and comprehensively evaluating the severity of sag. The method mainly includes extracting characteristic indicators of voltage sag and comprehensively evaluating the severity of sag. When extracting the characteristics of voltage sag, on the basis of traditional voltage sag characteristics (mainly including sag depth and sag duration, etc.), eight voltage sag characteristics are extracted considering the impact of voltage sag on sensitive loads on the user side. These eight characteristics are then used as indicators for evaluating the severity of sag, and the thresholds of different severity levels of each indicator are determined according to the mutual information entropy, and the severity is divided. Finally, the eight characteristic indicators are used as the input of DBN, and the model is trained with measured data to improve the clustering ability of DBN, determine the severity of each voltage sag event, and achieve the purpose of comprehensive evaluation.

Description

Technical Field

[0001] The present invention relates to the technical field of power quality, and in particular to a method for extracting voltage sag characteristic indicators and comprehensively evaluating the severity of the sag. Background Art

[0002] With the deepening reform of the power industry since the 21st century, while the capacity of high-power impact and nonlinear loads has grown rapidly, power users have adopted more sophisticated equipment and technologies on a large scale to meet the personalized and diverse product production needs, and have put forward higher requirements for power quality. For industrial power users with sensitive loads, voltage sags will directly affect the production efficiency and product quality of industrial users. When the sag is severe, it will cause the shutdown of automatic production lines, etc., resulting in huge economic losses. Therefore, accurately assessing the comprehensive severity of voltage sags is of great significance for the analysis and treatment of voltage sags.

[0003] A large number of scholars have studied the evaluation of the severity of voltage sags and proposed a variety of methods. Common random prediction methods use random models for evaluation. This method can be divided into the fault point method, the critical distance method, and the analytical method. These methods are scalable, but random parameters such as component failure rates are often set subjectively based on experience, so the accuracy of the evaluation results cannot be guaranteed.

[0004] At present, comprehensive evaluation methods are widely used to combine the weights of multiple indicators to conduct comprehensive evaluation of regions or nodes. The algorithm is simple and the data used is of low dimension. However, most evaluation methods are based on indicators on the grid side and do not fully consider the impact of voltage sag on the user side. In addition, for the voltage sag severity evaluation method that considers both the grid side and the user side, the voltage sag amplitude and duration are considered to be important characteristic quantities. However, these two basic characteristic quantities can only reflect the overall characteristics of the sag event in a statistical sense. The information contained in a single evaluation indicator is limited, and its evaluation result may be inaccurate, especially for sags with non-rectangular waveforms, which will cause over-evaluation. In classification, the relevant standard limit values ​​are generally used as the basis for judging whether the power quality is qualified or not, but there is no unified threshold for the voltage sag indicator.

[0005] Therefore, it is necessary to study a method for extracting characteristic quantities that can more accurately reflect the severity of voltage sag and a method for evaluating the severity of voltage sag that considers both the grid side and the user side. Summary of the invention

[0006] The present invention overcomes the shortcomings of the prior art and provides a method for extracting voltage sag characteristic indicators and comprehensively evaluating the severity of sags. In order to reasonably evaluate the impact of voltage sag events and consider the importance of loads, the present invention extracts 8 types of voltage sag characteristics based on traditional voltage sag indicators, taking into account problems such as three-phase imbalance and phase jump, and uses them as sag evaluation indicators. In addition, in order to reduce the impact of subjective judgment on the evaluation of sag severity, mutual information entropy is introduced to determine the thresholds of different severity levels of each indicator, and the severity is divided.

[0007] The technical solution of the present invention is: a method for voltage sag feature extraction and sag severity comprehensive evaluation, the main contents of which include voltage sag feature extraction and voltage sag severity classification based on mutual information entropy.

[0008] To achieve the above object, the technical solution adopted by the present invention is: a method for extracting voltage sag characteristic indicators and comprehensively evaluating the severity of voltage sag, characterized in that it includes the following steps:

[0009] S1. When extracting the characteristics of voltage sag, the impact of voltage sag on sensitive loads on the user side is considered, and 8 characteristic indicators of each sag event are extracted, namely, minimum amplitude, average effective value, sag duration, sag amplitude integral, voltage change rate, average equipment failure rate, three-phase unbalance coefficient, and phase offset;

[0010] S2, taking the eight characteristic indicators extracted in S1 as indicators for evaluating the severity of temporary drop, determining the thresholds of different severity levels of each indicator based on mutual information entropy, and dividing the severity levels;

[0011] S3. Using eight characteristic indicators as the input of DBN, the model is trained with measured data to improve the clustering ability of DBN, determine the severity of each voltage sag event, and achieve the purpose of comprehensive evaluation.

[0012] In a preferred embodiment of the present invention, the sag amplitude integral is the average value of the three-phase sag values ​​integrated over time during the sag duration, using the formula:

[0013] Where C is the integral of the voltage sag amplitude, V(t) is the voltage value of the voltage sag RMS waveform at time t, V N is the standard voltage, t1 and t2 are the start and end time of the sag respectively.

[0014] In a preferred embodiment of the present invention, the voltage change rate is the change steepness of the three-phase voltage amplitude during the transition period, using the formula: Where V(t) is the voltage value of the voltage sag RMS waveform at time t, and t1 and t'1 are the start and end times of the transition period, respectively.

[0015] In a preferred embodiment of the present invention, the average failure rate of the equipment is the average failure rate of PC, PLC, and ASD, using the formula:

[0016] In a preferred embodiment of the present invention, the three-phase unbalance coefficient is the ratio of the maximum difference between the single-phase voltage and the average voltage to the average voltage, using the formula: Among them U i is the integral of the I phase voltage during the sag, For U a , U b and U c The voltage amplitude imbalance VUB represents the ratio of the maximum difference between the single-phase voltage and the average voltage to the average voltage.

[0017] In a preferred embodiment of the present invention, the phase shift degree is the maximum phase angle between the three-phase phase shift and the three-phase symmetry during the sag process.

[0018] In a preferred embodiment of the present invention, it is assumed that the probability distribution of a discrete random variable (such as X) is: p(x)=P(X=x), x∈X, and the entropy of X is:

[0019] Mutual information entropy is a dimensionless statistic used to measure the information that a random variable X can provide about the change of another random variable Y, that is, the degree to which the random variable X can reduce the uncertainty of the random variable Y. The mutual information entropy formula is:

[0020] In order to eliminate the influence of dimension, the mutual information needs to be standardized to obtain the standardized mutual information, using the formula:

[0021] In a preferred embodiment of the present invention, after the critical values ​​of all temporary sag characteristic indicators are divided, the corresponding mutual information entropy can be used as a weight to temporarily determine the severity of each temporary sag event, using the formula:

[0022] Where n is the number of sag events; i represents the i-th characteristic indicator; I is the total number of characteristic indicator types; N(X i ; L i ) is the mutual information entropy of the i-th feature index; L i (n) represents the severity of indicator i of event n.

[0023] In a preferred embodiment of the present invention, the start time and the end time of the dip are determined by finding the distortion point using a wavelet transform method.

[0024] In a preferred embodiment of the present invention, the severity is normalized and the critical range of the comprehensive severity level is set to 5 levels, wherein level 1 is 0-0.2; level 2 is 0.2-0.4; level 3 is 0.4-0.6; level 4 is 0.6-0.8; and level 5 is 0.8-1.

[0025] The present invention solves the defects existing in the background technology and has the following beneficial effects:

[0026] (1) The present invention provides a method for extracting voltage sag characteristic indicators and comprehensively evaluating the severity of sag. The present invention adds a new sag characteristic indicator, sag amplitude integral, to the traditional sag characteristic indicators (sag depth and sag duration). The traditional characteristic indicators are described as follows: Figure 1 In the case of the two types of severe sags shown in the figure, the sag depth and duration are the same, but their sag severity is obviously not the same. The use of the sag amplitude integral can clearly distinguish the severity of the two types of sags and solve the problem of over-evaluation of non-rectangular sags.

[0027] (2) The present invention comprehensively considers the adverse effects of voltage sag on the grid side and the user side, newly adds the average equipment failure rate characteristic index, and presents the relationship between the voltage tolerance characteristics of sensitive loads on the user side and the sag event itself.

[0028] (3) The present invention introduces mutual information entropy to classify the severity of voltage sags, and mines the probability density correlation between each sag characteristic feature index and the severity in the massive monitoring data. The optimal solution of each level of critical index is sought, and the severity is divided into five categories, which solves the problem that there is no universal standard in the field of voltage sag assessment and the critical values ​​of each severity are determined by subjective experience. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art are briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative work.

[0030] Figure 1 1. The non-rectangular sag and rectangular sag RMS waveform diagrams of the preferred embodiment of the present invention;

[0031] Figure 2 This is a diagram of the wavelet transform result of the dip waveform of the preferred embodiment of the present invention;

[0032] Figure 3 is a sensitive load voltage tolerance curve diagram of a preferred embodiment of the present invention;

[0033] Figure 4 It is a schematic diagram of various types of temporary drops in a preferred embodiment of the present invention;

[0034] Figure 5 is an RMS waveform diagram of a sag event in a preferred embodiment of the present invention;

[0035] Figure 6 is a three-phase voltage phase shift diagram of a preferred embodiment of the present invention;

[0036] Figure 7 4 is a distribution diagram of the severity of the post-PCA dip in a preferred embodiment of the present invention. DETAILED DESCRIPTION

[0037] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only 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.

[0038] Based on the traditional voltage sag characteristic indicators, the present invention extracts 8 types of voltage sag characteristic indicators, as shown in Table 1 below.

[0039] Table 1 Characteristic indicators for voltage sag severity assessment

[0040]

[0041]

[0042] Among them, the minimum amplitude, average effective value and sag duration in Table 1 are traditional features. The extraction methods of these three features belong to the prior art. Here, the five new features proposed in the present invention are described.

[0043] 1. Temporary drop amplitude integral

[0044] In order to more accurately describe the non-rectangular sag, this paper combines the two single variables of depth and duration as the index of the sag amplitude integral according to the energy accumulation principle, as shown in (1):

[0045]

[0046] Where C is the integration result, V(t) is the voltage value of the voltage sag RMS waveform at time t, V N is the standard voltage, t1 and t2 are the start and end time of the sag respectively. Figure 1 The two types of sag waveforms are shown. Figure 1The waveforms are non-rectangular sag and rectangular sag RMS.

[0047] Figure 1 The integral of the temporary drop amplitude in (a) is 0.1449. Figure 1 The integral of the sag amplitude in (b) is 0.3549. This new feature can effectively represent the difference between the overall severity of the two events. In addition, the voltage used in the present invention is changed to an integral form, as shown in (2):

[0048]

[0049] U is the integral of the voltage RMS waveform during the sag period, V(t) is the voltage value of the voltage RMS waveform at time t, V N is the standard voltage, t1 and t2 are the start and end time of the sag respectively.

[0050] 2. Voltage change rate

[0051] Due to the occurrence of a fault or the instantaneous action of the equipment, the system quickly transitions from one stable state to another. This transition period is very short, so the voltage amplitude changes and fluctuates very quickly. The maximum change steepness of the effective value of the three-phase voltage during the transition period is defined as a new feature, and the voltage change rate is calculated according to formula (3):

[0052]

[0053] V(t) is the voltage value of the voltage sag RMS waveform at time t, and t1 and t'1 are the start and end times of the transition period, respectively. Figure 2 The figure shows the wavelet transform result of the dip waveform. The transition period t1 and t'1 of the dip waveform can be obtained by wavelet transform.

[0054] 3. Average equipment failure rate

[0055] The impact of voltage sag on sensitive loads can be described by the failure rate of sensitive equipment. The general voltage tolerance curve (VTC) of sensitive loads is generally rectangular. The VCT of personal computers (PCs), programmable logic controllers (PLCs), and adjustable speed drives (ASDs) is as follows: Figure 3 shown.

[0056] The normal working area is outside curve 1, the equipment failure area is inside curve 2, and the uncertainty area is between curve 1 and curve 2. The failure rate of sensitive equipment on the user side is defined as:

[0057]

[0058] 4. Three-phase unbalance coefficient

[0059] The voltage sag waveform caused by different types of short-circuit faults is different. Different load wiring methods and different transformer wiring methods on the equipment side will also change the amplitude. Taking phase A as the reference phase, the calculation example of the voltage amplitude of each phase of various types of voltage sags caused by faults is as follows: Figure 4 shown.

[0060] The present invention extracts the feature of the unbalanced degree of the temporary sag value as one of the indicators for evaluating the severity of the temporary sag, and the calculation formula is shown in formula (5).

[0061]

[0062] Among them U i is the integral of the I phase voltage during the sag, For U a , U b and U c The voltage amplitude unbalance VUB represents the ratio of the maximum difference between the single-phase voltage and the average voltage to the average voltage.

[0063] 5. Phase deviation

[0064] from Figure 4 It can be seen that there are phase jumps in C, D, F and G type depressions. The phase change can be positive or negative, and its value is mostly close to zero phase. The absolute value of the maximum phase angle between the three-phase phase deviation and the three-phase symmetry during the sag is defined as the new index phase deviation.

[0065] Voltage sag severity classification based on mutual information entropy: The present invention continuously adjusts the value of the critical characteristic indicator for each sag characteristic indicator, takes the maximum mutual information entropy of the characteristic indicator distribution and its severity classification as the goal, and finds the optimal solution for the critical value of the sag characteristic indicator.

[0066] Information entropy is a measure of the uncertainty of a random variable X in statistics. The higher the uncertainty of X, the greater the entropy. Assuming that the probability distribution of a discrete random variable (such as X) is: p(x) = P(X = x), x∈X, the entropy of X is:

[0067]

[0068] The mutual information entropy is a dimensionless statistic used to measure the information that a random variable X can provide about the change of another random variable Y, that is, the degree to which the random variable X can reduce the uncertainty of the random variable Y. The mutual information entropy formula is:

[0069]

[0070] p(x,y) is the joint probability density of x and y. In order to eliminate the influence of dimension, the mutual information needs to be standardized to obtain the standardized mutual information:

[0071]

[0072] After the critical values ​​of all temporary sag characteristic indicators are divided, the corresponding mutual information entropy can be used as a weight to temporarily determine the severity of each temporary sag event:

[0073]

[0074] (9) where n is the number of sag events; i is the i-th characteristic indicator; I is the total number of characteristic indicator types; N(X i ; L i ) is the mutual information entropy of the i-th feature index; L i (n) represents the severity of indicator i of event n.

[0075] Embodiment 1

[0076] The voltage sag feature extraction and sag severity comprehensive evaluation method first calculates the RMS waveform of the three-phase voltage data, and then calculates the eight characteristic indicators of each sag event. Figure 5 The calculation process of the indicator is explained by taking the temporary drop event shown in the figure as an example.

[0077] Characteristic index 1 (minimum amplitude) and characteristic index 3 (sag duration) can be directly expressed from the image. The start and end time of the sag can be determined by finding the distortion point using the wavelet transform method.

[0078] According to the three-phase waveform and the duration of the drop, characteristic index 2 (average effective value) and characteristic index 4 (temporary drop amplitude integral) can be calculated. According to the results of wavelet transform, according to the start and end time t1-t'1 and t2-t'2 of the transition section, the maximum change rate of the three-phase voltage is calculated according to formula (3), and characteristic index 5 (three-phase voltage change rate) can be extracted. Characteristic index 6 (average equipment failure rate) can be obtained according to formula (4) and Figure 3 , calculated from the average effective value and the duration of the sag. Figure 3 It can be seen that this temporary drop event has no impact on PC and PLC devices, and the failure rate of ASD is 0.0764. Characteristic index 7 (three-phase unbalance coefficient) can be calculated according to formula (5). The curve of three-phase phase offset changing with time is shown in Figure 6 As shown, characteristic index 8 is the maximum absolute value of the three-phase voltage offset phase.

[0079] The calculation results of the eight characteristic indicators of this temporary drop event are shown in Table 2. The larger the characteristic indicators 1 and 2, the smaller the severity, and the greater the severity of the remaining characteristic indicators.

[0080] Table 2 Calculation results of eight characteristic indicators of voltage sag events

[0081] Indicator 1 0.7238 Indicator 2 0.8463 Indicator 3 0.3448 Indicator 4 0.0536 Indicator 5 13.3888 Indicator 6 0.0255 Indicator 7 0.1858 Indicator 8 0.1960

[0082] Then, the severity level and the maximum mutual information entropy of each indicator are used as the objective function to determine the critical value of each level indicator, as shown in Table 3. The higher the level, the more serious the level.

[0083] Table 3 Critical values ​​of indicators at each level

[0084]

[0085]

[0086] Then, the mutual information entropy is used as the weight of each characteristic index. The mutual information entropy can measure the correlation between different characteristic indexes and the severity of the temporary sag. The larger the entropy value, the stronger the correlation between the characteristic index and the severity of the temporary sag, and the greater the importance of the power sector in the evaluation. The calculation results of the mutual information entropy of each characteristic index are shown in Table 4.

[0087] Table 4 Mutual information entropy values ​​of each indicator

[0088] index Mutual Information Entropy Indicator 1 0.6476 Indicator 2 0.6965 Indicator 3 0.6583 Indicator 4 0.7481 Indicator 5 0.5966 Indicator 6 0.6007 Indicator 7 0.5598 Indicator 8 0.5125

[0089] Finally, according to formula (9), the comprehensive severity of each voltage sag event can be calculated: the severity of the characteristic indicators in the example is index 1 (level 2), index 2 (level 5), index 3 (level 2), index 4 (level 5), index 5 (level 3), index 6 (level 3), index 7 (level 4), index 8 (level 1), the comprehensive severity is 3.1247, the maximum value is 5, and the severity is standardized. The critical range of the comprehensive severity level is specified as level 1 (0-0.2), level 2 (0.2-0.4), level 3 (0.4-0.6), level 4 (0.6-0.8) and level 5 (0.8-1). The comprehensive severity of 3.2227 in the example is normalized to 0.6445, which belongs to level 4.

[0090] According to the above steps, the data of 600 voltage sag events were analyzed. The principal component analysis algorithm (PCA) was used to reduce the 8-dimensional characteristic indicators of the 600 voltage sag events to a two-dimensional plane for observation. Figure 7 The color bar on the right shows the color corresponding to the normalized comprehensive severity, and the figure on the left shows the classification results of the severity of 600 temporary drops. It can be observed that mild temporary drops are mostly distributed on the left, and severe temporary drops are distributed in the middle and right. This proves the consistency and effectiveness of the severity classification method based on mutual information entropy.

[0091] The above is based on the ideal embodiment of the present invention. Through the above description, relevant personnel can make various changes and modifications without departing from the technical concept of the present invention. The technical scope of the present invention is not limited to the content in the specification, and the technical scope must be determined according to the scope of the claims.

Claims

1. A method for extracting voltage sag characteristic indicators and comprehensively evaluating the severity of voltage sag, characterized in that: The following steps are involved: S1. When extracting the characteristics of voltage sag, the impact of voltage sag on sensitive loads on the user side is considered, and 8 characteristic indicators of each sag event are extracted, namely, minimum amplitude, average effective value, sag duration, sag amplitude integral, voltage change rate, average equipment failure rate, three-phase unbalance coefficient, and phase offset; S2, taking the eight characteristic indicators extracted in S1 as indicators for evaluating the severity of temporary drop, determining the thresholds of different severity levels of each indicator based on mutual information entropy, and dividing the severity levels; S3, using 8 characteristic indicators as DBN input, using measured data to train the model to improve the clustering ability of DBN, determine the severity of each voltage sag event, and achieve the purpose of comprehensive evaluation; Among them, the sag amplitude integral is the average value of the three-phase sag values ​​integrated over time during the sag duration; the voltage change rate is the steepness of the change of the voltage amplitude during the transition period; the average equipment failure rate is the average failure rate of PC, PLC, and ASD; the three-phase unbalance coefficient is the ratio of the maximum difference between the single-phase voltage and the average voltage to the average voltage; the phase shift degree is the maximum phase angle between the three-phase phase shift and the three-phase symmetry during the sag process.

2. The method for extracting voltage sag characteristic indicators and comprehensively evaluating the severity of voltage sag according to claim 1, characterized in that: The integral of the sag amplitude is expressed by the formula: , where C is the integral of the sag amplitude, V(t) is the voltage value of the voltage sag RMS waveform at time t, is the standard voltage, and The start and end times of the slump respectively.

3. The method for extracting voltage sag characteristic indicators and comprehensively evaluating the severity of voltage sag according to claim 1, characterized in that: The voltage change rate is expressed as: , where V(t) is the voltage value of the voltage sag RMS waveform at time t, and The start and end times of the transition period respectively.

4. The method for extracting voltage sag characteristic indicators and comprehensively evaluating the severity of voltage sag according to claim 1, characterized in that: The average failure rate of the equipment is expressed as: 。 5. The method for extracting voltage sag characteristic indicators and comprehensively evaluating the severity of voltage sag according to claim 1, characterized in that: The three-phase unbalance coefficient is expressed by the formula: ;in is the integral of the I phase voltage during the sag, for , and The voltage amplitude unbalance VUB represents the ratio of the maximum difference between the single-phase voltage and the average voltage to the average voltage.

6. The method for extracting voltage sag characteristic indicators and comprehensively evaluating the severity of voltage sag according to claim 1, characterized in that: Assume that the probability distribution of discrete random variable X is: p(x)=P(X=x), x∈X, and the entropy of X is: ； Mutual information entropy is a dimensionless statistic used to measure the information that a random variable X can provide about the change of another random variable Y, that is, the degree to which the random variable X can reduce the uncertainty of the random variable Y. The mutual information entropy formula is: ； In order to eliminate the influence of dimension, the mutual information needs to be standardized to obtain the standardized mutual information, using the formula: 。 7. The method for extracting voltage sag characteristic indicators and comprehensively evaluating the severity of voltage sag according to claim 1, characterized in that: After the critical values ​​of all temporary sag characteristic indicators are divided, the corresponding mutual information entropy can be used as the weight to temporarily determine the severity of each temporary sag event, using the formula: ; Where n is the number of sag events; i represents the i-th characteristic indicator; I is the total number of characteristic indicator types; is the mutual information entropy of the i-th feature index; Indicates the severity of indicator i of event n.

8. The method for extracting voltage sag characteristic indicators and comprehensively evaluating the severity of voltage sag according to claim 1, characterized in that: The start and end time of the sag are determined by finding the distortion point using the wavelet transform method.

9. The method for extracting voltage sag characteristic indicators and comprehensively evaluating the severity of voltage sag according to claim 1, characterized in that: After normalizing the severity, the critical range of the comprehensive severity level is set at 5 levels; among them, level 1 is 0-0.2; level 2 is 0.2-0.4; level 3 is 0.4-0.6; level 4 is 0.6-0.8; level 5 is 0.8-1.

Images