Short circuit fault control method and device

Abstract

The invention discloses a short circuit fault control method and device. According to the method, the fault estimation value is calculated by acquiring the real-time current and combining the F test algorithm with the real-time and estimated current values, and the fault estimation value is compared with the threshold value to judge the fault. And after the fault is confirmed, determining a short-circuit current zero point according to a zero point prediction algorithm, and cutting off the fault current at the point by using the circuit breaker. According to the invention, the accuracy and speed of fault detection are improved, and the operation opportunity of the circuit breaker is optimized, so that the problem that the short-circuit fault cannot be accurately controlled in the prior art is solved.

Description

Technical Field

[0001] The present invention relates to the field of short circuit fault control, and in particular to a short circuit fault control method and device. Background Art

[0002] As the scale of the power system continues to expand, the short-circuit current of the system continues to rise, the DC component of the short-circuit current increases, and the harm of short-circuit accidents further expands. Therefore, in response to this problem, reducing the time of fault occurrence is one of the measures to solve this problem. At present, domestic and foreign scholars have conducted a lot of research on how to reduce the fault identification time and quickly identify the fault. In the early stage, the idea of ​​fault identification was to form a criterion by integrating multiple characteristic quantities of the post-fault signal, such as the instantaneous value of current, current slope, current curvature, instantaneous value of voltage, power, etc. This type of algorithm can usually accurately determine whether a fault has occurred, but there are problems such as long detection time and sensitivity to noise harmonics. Subsequently, there appeared identification algorithms using phase shifters, graphical criterion algorithms by constructing inequalities, and optimized windowed Fourier transform and short-time Fourier transform algorithms, but these algorithms cannot meet the requirements of fast disconnection for identification time and accuracy. In recent years, artificial neural network algorithms have also begun to be used in the field of fault identification. Although they are significantly better than other algorithms in terms of recognition speed, they require a large amount of data training to improve algorithm performance. Their recognition effect in practical applications has yet to be verified, resulting in the inability to accurately detect and control short-circuit faults in existing technologies. Summary of the invention

[0003] The present invention provides a short circuit fault control method and device to solve the problem that the short circuit fault cannot be accurately controlled in the prior art.

[0004] In a first aspect, the present application provides a short circuit fault control method, comprising:

[0005] Get real-time current data;

[0006] Calculating a current fault estimation value according to an F-test algorithm, the real-time current data and the estimated current value;

[0007] Comparing the current fault estimation value with a preset threshold value to determine whether a fault occurs;

[0008] If a fault occurs, the real-time current data is predicted according to a zero point prediction algorithm to obtain a predicted zero point of the short-circuit fault current;

[0009] According to the predicted zero point and the preset circuit breaker, the short-circuit fault current is interrupted.

[0010] The present application can monitor the current state in real time by acquiring the real-time current data of the power system, and provide an accurate data basis for fault detection. The present application uses the F-test algorithm to compare the real-time current data and the estimated current value, and the calculated current fault estimate can quantify the deviation between the current data and the normal operating state, thereby providing an objective basis for fault judgment. By comparing the current fault estimate with the preset threshold, it is possible to quickly and accurately determine whether there is a short-circuit fault in the system. This method reduces the dependence on the time of the fault occurrence and improves the response speed of fault detection. Once the fault is determined, the zero-point prediction algorithm can predict the zero point of the short-circuit fault current, ensuring that the circuit breaker is disconnected when the current is close to zero, which can minimize the ablation of the circuit breaker contacts, reduce the risk of operating overvoltage, and increase the service life of the circuit breaker. The present application not only improves the accuracy and speed of fault detection, but also optimizes the operation timing of the circuit breaker to solve the problem that the short-circuit fault cannot be accurately controlled in the prior art.

[0011] As a preferred embodiment of the first aspect, the estimated current value is obtained by fitting and calculating according to historical current data, specifically:

[0012] Get historical current data;

[0013] According to a preset algorithm and the historical current data, a historical current expression is obtained by fitting;

[0014] According to the historical current expression, an estimated current value is calculated.

[0015] In this preferred embodiment, the present application first uses a preset algorithm to conduct in-depth analysis and fitting of these data, thereby constructing a mathematical expression of the historical current. This step not only provides an accurate mathematical model for the normal operating state of the power system, but also lays a solid foundation for subsequent fault detection. Using this historical current expression as an estimated current value and comparing it with the current data obtained by real-time sampling of the sensor can effectively identify any current changes that deviate from the normal mode. In addition, through real-time monitoring and intelligent analysis, the present application can quickly respond to and accurately locate faults, reduce the power outage time caused by faults, and improve the stability and reliability of the power system. Finally, the present application significantly reduces the economic losses and safety hazards caused by short-circuit faults in the power system through early prediction and timely response.

[0016] As a preferred embodiment of the first aspect, the current fault estimation value is calculated based on the F-test algorithm, the real-time current data and the estimated current value, specifically:

[0017] Determine the regression sum of squares and the residual sum of squares of the current data according to the real-time current data and the estimated current value;

[0018] According to the expression of the F-test algorithm, the current fault estimation value is calculated;

[0019] Among them, the expression of the F test algorithm is:

[0020]

[0021] In the formula, SS R is the regression sum of squares, SS E is the residual sum of squares, m is the number of unknown variables, n is the number of sample values ​​in the test, p is the number of columns in the coefficient matrix, and F0 is the current fault estimate.

[0022] As a preferred embodiment of the first aspect, comparing the current fault estimation value with a preset threshold to determine whether a fault occurs is specifically:

[0023] If the current fault estimation value is less than a preset threshold, it is determined that a fault occurs.

[0024] In this preferred embodiment, the present application accurately calculates the current fault estimate by applying the F-test algorithm to determine whether there is a short-circuit fault in the power system. First, the average value of the real-time current data is obtained to provide a benchmark for subsequent comparisons. It not only provides a quantitative fault detection standard, but also can more accurately identify short-circuit faults by comparing the real-time current data with the current expression of historical fitting. The present application is based on statistical principles to ensure the scientificity and accuracy of fault detection. When the current fault estimate exceeds the preset threshold, it indicates that there is a significant difference between the real-time current data and the estimated current value, thereby judging that a short-circuit fault may have occurred in the system. This method based on statistical analysis improves the sensitivity and reliability of fault detection, helps to take preventive or repair measures in a timely manner, reduces the losses caused by faults in the power system, and improves the stability and safety of the system.

[0025] As a preferred embodiment of the first aspect, predicting the real-time current data according to the zero-point prediction algorithm is specifically as follows:

[0026] The zero point prediction algorithm includes a safe point algorithm, an adaptive self-detection algorithm, an arithmetic recursion method, a weighted minimum mean square error algorithm, an improved half-wave Fourier algorithm, a Prony algorithm or a recursive least squares algorithm.

[0027] In this preferred embodiment, the present application uses a zero-point prediction algorithm to predict real-time current data, which can predict the zero point of the short-circuit fault current, thereby optimizing the operation timing of the circuit breaker. Since the zero-point prediction algorithm includes a variety of advanced algorithms, such as a safe point algorithm, an adaptive self-detection algorithm, an arithmetic recursion method, a weighted minimum mean square error algorithm, an improved half-wave Fourier algorithm, a Prony algorithm or a recursive least squares algorithm, these algorithms are all intended to improve the accuracy and response speed of the prediction. Through these algorithms, the system can analyze the dynamic changes of the current data in real time and quickly identify the position of the zero point in the current waveform. This precise zero-point prediction capability enables the circuit breaker to perform a breaking operation when the current is close to zero, thereby minimizing arc generation and contact erosion, and improving the service life of the circuit breaker and the reliability of the system. In addition, accurate zero-point prediction also helps to reduce the voltage shock caused by current mutations and protect other components of the power system from damage. Therefore, the present method significantly improves the stability and safety of the power system by improving the accuracy and efficiency of fault current interruption.

[0028] In a second aspect, the present application provides a short circuit fault control device. The short circuit fault control device comprises an acquisition module, a calculation module, a comparison module, a prediction module and a disconnection module;

[0029] The acquisition module is used to obtain real-time current data;

[0030] The calculation module is used to calculate the current fault estimation value according to the F-test algorithm, the real-time current data and the estimated current value;

[0031] The comparison module is used to compare the current fault estimation value with a preset threshold value to determine whether a fault occurs;

[0032] The prediction module is used to predict the real-time current data according to the zero point prediction algorithm if a fault occurs, and obtain the predicted zero point of the short-circuit fault current;

[0033] The breaking module is used to break the short-circuit fault current according to the predicted zero point and the preset circuit breaker.

[0034] This device uses five modules to divide the work and coordinate the work to better control the short-circuit fault. This application can monitor the current state in real time by acquiring the real-time current data of the power system, and provide an accurate data basis for fault detection. This application uses the F test algorithm to compare the real-time current data and the estimated current value, and the calculated current fault estimate can quantify the deviation between the current data and the normal operating state, thereby providing an objective basis for fault judgment. By comparing the current fault estimate with the preset threshold, it is possible to quickly and accurately determine whether there is a short-circuit fault in the system. This method reduces the dependence on the time of the fault occurrence and improves the response speed of fault detection. Once the fault is determined, the zero-point prediction algorithm can predict the zero point of the short-circuit fault current to ensure that the circuit breaker is disconnected when the current is close to zero, which can minimize the erosion of the circuit breaker contacts, reduce the risk of operating overvoltage, and increase the service life of the circuit breaker. This application not only improves the accuracy and speed of fault detection, but also optimizes the operation timing of the circuit breaker to solve the problem that the prior art cannot accurately control short-circuit faults.

[0035] As a preferred embodiment of the second aspect, the estimated current value is obtained by fitting and calculating according to historical current data, specifically:

[0036] Get historical current data;

[0037] According to a preset algorithm and the historical current data, a historical current expression is obtained by fitting;

[0038] According to the historical current expression, an estimated current value is calculated.

[0039] In this preferred embodiment, the present application first uses a preset algorithm to conduct in-depth analysis and fitting of these data, thereby constructing a mathematical expression of the historical current. This step not only provides an accurate mathematical model for the normal operating state of the power system, but also lays a solid foundation for subsequent fault detection. Using this historical current expression as an estimated current value and comparing it with the current data obtained by real-time sampling of the sensor can effectively identify any current changes that deviate from the normal mode. In addition, through real-time monitoring and intelligent analysis, the present application can quickly respond to and accurately locate faults, reduce the power outage time caused by faults, and improve the stability and reliability of the power system. Finally, the present application significantly reduces the economic losses and safety hazards caused by short-circuit faults in the power system through early prediction and timely response.

[0040] As a preferred embodiment of the second aspect, the current fault estimation value is calculated based on the F-test algorithm, the real-time current data and the estimated current value, specifically:

[0041] Determine the regression sum of squares and the residual sum of squares of the current data according to the real-time current data and the estimated current value;

[0042] According to the expression of the F-test algorithm, the current fault estimation value is calculated;

[0043] Among them, the expression of the F test algorithm is:

[0044]

[0045] In the formula, SS R is the regression sum of squares, SS E is the residual sum of squares, m is the number of unknown variables, n is the number of sample values ​​in the test, p is the number of columns in the coefficient matrix, and F0 is the current fault estimate.

[0046] As a preferred embodiment of the second aspect, comparing the current fault estimation value with a preset threshold to determine whether a fault occurs is specifically:

[0047] If the current fault estimation value is less than a preset threshold, it is determined that a fault occurs.

[0048] In this preferred embodiment, the present application accurately calculates the current fault estimate by applying the F-test algorithm to determine whether there is a short-circuit fault in the power system. First, the average value of the real-time current data is obtained to provide a benchmark for subsequent comparisons. It not only provides a quantitative fault detection standard, but also can more accurately identify short-circuit faults by comparing the real-time current data with the current expression of historical fitting. The present application is based on statistical principles to ensure the scientificity and accuracy of fault detection. When the current fault estimate exceeds the preset threshold, it indicates that there is a significant difference between the real-time current data and the estimated current value, thereby judging that a short-circuit fault may have occurred in the system. This method based on statistical analysis improves the sensitivity and reliability of fault detection, helps to take preventive or repair measures in a timely manner, reduces the losses caused by faults in the power system, and improves the stability and safety of the system.

[0049] As a preferred embodiment of the second aspect, predicting the real-time current data according to the zero-point prediction algorithm is specifically:

[0050] The zero point prediction algorithm includes a safe point algorithm, an adaptive self-detection algorithm, an arithmetic recursion method, a weighted minimum mean square error algorithm, an improved half-wave Fourier algorithm, a Prony algorithm or a recursive least squares algorithm.

[0051] In this preferred embodiment, the present application uses a zero-point prediction algorithm to predict real-time current data, which can predict the zero point of the short-circuit fault current, thereby optimizing the operation timing of the circuit breaker. Since the zero-point prediction algorithm includes a variety of advanced algorithms, such as a safe point algorithm, an adaptive self-detection algorithm, an arithmetic recursion method, a weighted minimum mean square error algorithm, an improved half-wave Fourier algorithm, a Prony algorithm or a recursive least squares algorithm, these algorithms are all intended to improve the accuracy and response speed of the prediction. Through these algorithms, the system can analyze the dynamic changes of the current data in real time and quickly identify the position of the zero point in the current waveform. This precise zero-point prediction capability enables the circuit breaker to perform a breaking operation when the current is close to zero, thereby minimizing arc generation and contact erosion, and improving the service life of the circuit breaker and the reliability of the system. In addition, accurate zero-point prediction also helps to reduce the voltage shock caused by current mutations and protect other components of the power system from damage. Therefore, the present method significantly improves the stability and safety of the power system by improving the accuracy and efficiency of fault current interruption. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] Figure 1 : A flow chart of an embodiment of a short-circuit fault control method provided by the present application;

[0053] Figure 2 : A structural diagram of an embodiment of a prony algorithm current estimation value fitting waveform provided in the present application;

[0054] Figure 3 : A structural diagram of an embodiment of the prony algorithm current estimation value fitting waveform right-shifted waveform comparison provided by the present application;

[0055] Figure 4 : A structural diagram of an embodiment of the F0 hypothesis test simulation result provided by the present application;

[0056] Figure 5 : A structural schematic diagram of an embodiment of the short-circuit fault control operation provided by the present application;

[0057] Figure 6 : A structural schematic diagram of an embodiment of a short-circuit fault control device provided in the present application. DETAILED DESCRIPTION

[0058] 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.

[0059] Embodiment 1

[0060] Please refer to Figure 1 , which is a short circuit fault control method provided by an embodiment of the present invention.

[0061] In this embodiment, the process of the short circuit fault control method in the present application is described in detail through steps S01-S05.

[0062] S01: Get real-time current data.

[0063] S02: Calculate and obtain a current fault estimation value according to an F-test algorithm, the real-time current data and the estimated current value.

[0064] As a preferred embodiment of the first embodiment, the estimated current value is obtained by fitting and calculating according to historical current data, specifically:

[0065] Collect the current data of historical normal conditions. Since the system is in normal operation most of the time, the estimated current can be obtained through a large number of sampling points.

[0066] The collected normal operating current data is used to fit the short-circuit current expression through the prony algorithm, and the estimated current value is calculated based on the historical current expression.

[0067] More specifically, the fitting process of the prony algorithm is as follows:

[0068] The current waveform expression of normal operation is:

[0069]

[0070] Among them, I0 is the amplitude of the fundamental component of the power frequency, is the fundamental phase.

[0071] The mathematical model of the prony algorithm is:

[0072]

[0073] Among them, A i is the fitting amplitude; α i is the attenuation factor; f i is the fitting frequency; θ i is the fitting phase. p is the fitting order of prony. Combined with Euler's formula, the combination of two complex exponentials can form a sinusoidal signal, and the mathematical model of the prony algorithm is transformed into:

[0074]

[0075] When the attenuation factor α i=0, the sampling signal x(t) has no damping characteristics and can be regarded as a fit of sinusoidal signals of various orders that change periodically; combined with the normal operating current waveform expression, we know that p=1.

[0076] According to the prony algorithm model, when the parameters f, θ, A, and α are obtained, the expression corresponding to the current can be obtained.

[0077] Discretize the continuous-time Prony algorithm to get:

[0078]

[0079] Where n is the signal-time data sequence, n=0, ... N-1, N is the sampling window length used by the prony algorithm, and Δt represents the sampling time interval. It is found that the simplified parameter form is consistent with the form of the homogeneous solution of the linear difference equation with constant coefficients. Therefore, in order to obtain the above characterization parameters, a characteristic polynomial can be defined to obtain the solution of the homogeneous equation as follows:

[0080]

[0081] in,

[0082]

[0083] According to the simplified prony expression, the sampling data at the previous k moments can be expressed as:

[0084]

[0085] Multiplying both sides by αk and summing them according to the signal order, we can get:

[0086]

[0087] After the coefficient transformation on the right side, we have:

[0088]

[0089] Although the specific value of zi is unknown at this time, z=zi is exactly the characteristic root corresponding to this polynomial, so the second summation term on the right side of the formula about αk is 0, then:

[0090]

[0091] After one least squares algorithm operation, we can get α1, α2, ..., α p , so we can get z1, z2...z p , and then converted into a linear equation for parameter b: Zb = X

[0092] According to the above solution, the parameters of each order can be obtained:

[0093]

[0094] Substitute the prony algorithm expression to find the current expression under historical normal operating conditions.

[0095] More specifically, Figure 2 As shown in the figure, according to the parameters obtained by the prony algorithm, the fitting amplitude, attenuation factor, frequency and phase are restored to obtain the current waveform expression under normal operation. Figure 3 This is a waveform comparison diagram of the current estimation value fitting waveform of the prony algorithm in the embodiment of the present application shifted to the right by 1ms. The current estimation value waveform obtained by the prony algorithm overlaps with the original waveform. Figure 3 The current estimation waveform obtained by the prony algorithm is shifted to the right by 1ms, while the original current waveform remains unchanged.

[0096] In this preferred embodiment, the present application first uses a preset algorithm to conduct in-depth analysis and fitting of these data, thereby constructing a mathematical expression of the historical current. This step not only provides an accurate mathematical model for the normal operating state of the power system, but also lays a solid foundation for subsequent fault detection. Using this historical current expression as an estimated current value and comparing it with the current data obtained by real-time sampling of the sensor can effectively identify any current changes that deviate from the normal mode. In addition, through real-time monitoring and intelligent analysis, the present application can quickly respond to and accurately locate faults, reduce the power outage time caused by faults, and improve the stability and reliability of the power system. Finally, the present application significantly reduces the economic losses and safety hazards caused by short-circuit faults in the power system through early prediction and timely response.

[0097] As a preferred embodiment of the first embodiment, the current fault estimation value is calculated based on the F-test algorithm, the real-time current data and the estimated current value, specifically:

[0098] The "F0 hypothesis test" or "significance test of the regression equation" is defined as follows:

[0099]

[0100] in,

[0101]

[0102] Among them, SS R is the regression sum of squares, SS E is the residual sum of squares, m is the number of unknown variables, n is the number of sample values ​​in the test, p is the number of columns in the coefficient matrix, Estimate the current value for the model, is the sampling mean, y i is the current sampling value.

[0103] More specifically, Figure 4 As shown, according to the "F0 hypothesis test", the current estimation value, the sampled current data and the sampled current average value obtained in advance by the prony algorithm are substituted into the formula of the "F0 hypothesis test" and compared with the set threshold.

[0104] S03: Compare the current fault estimation value with a preset threshold to determine whether a fault occurs.

[0105] As a preferred embodiment of the first embodiment, the current fault estimation value is compared with a preset threshold value to determine whether a fault occurs, specifically:

[0106] According to the expression of F0 hypothesis test, it can be seen that when no fault occurs, the current sampling data and the current estimated value data are equal, and F0 should be infinite; when a fault occurs, the current sampling data and the current estimated value data are not equal, and F0 will be a relatively small number. By setting the threshold of F0, the occurrence of the fault can be judged. If the obtained F0 value is less than the set threshold, it is judged that a fault has occurred. Otherwise, it returns to the "F0 hypothesis test" detection part and obtains F0 again for comparison.

[0107] More specifically, Figure 5 As shown, it is the algorithm flow of fault identification of the present invention. The current estimation value is first obtained by fitting the prony algorithm, and the F0 value is tested by the "F0 hypothesis test". The current estimation value obtained by fitting the prony algorithm: the power system is in normal operation most of the time, and the fault occupies a relatively short period of time in the entire system operation. First, the current data under normal conditions is collected, and the waveform expression of the current estimation value is obtained in advance. The "F0 hypothesis test" tests the F0 value part: the current estimation value, the sampled current data and the sampled current average value obtained in advance by the prony algorithm are substituted into the formula of the "F0 hypothesis test", and compared with the set threshold value. If the obtained F0 value is less than the set threshold value, it is judged that a fault has occurred. Otherwise, it returns to the "F0 hypothesis test" detection part and obtains F0 again for comparison.

[0108] Since the F0 value is generally large, f0 = log(F0) is set as the criterion. When the system is operating normally, if the noise is not considered, the current estimation value and the sampled current value are equal, and f0 tends to infinity. When the signal-to-noise ratio is 20dB and the sampling frequency is 4kHz, within one power frequency cycle, n = 80, p = 2, m = 2, and then it can be calculated that f0 ≈ 3.591 06. In practical applications, in order to ensure the sensitivity of the algorithm and the convenience of calculation, the f0 threshold is set to 3.45.

[0109] Figure 5 The hardware part of the controller (phase control device) in the control panel mainly includes the control panel core, signal acquisition part, opening and closing and charging control part, power supply part, and communication part. It is mainly responsible for real-time monitoring of the voltage and current status of the line, and performing phase control opening operation at the appropriate time. The main function of the control panel part is to process the collected signals, obtain various parameters by prony algorithm, and then calculate according to the "F0 hypothesis test". If the obtained F0 value is less than the set threshold, it is judged that a fault has occurred. Otherwise, it returns to the "F0 hypothesis test" detection part and obtains F0 again for comparison. In addition, the parameters can be set by pressing buttons, and the current information of the system can be displayed in real time through the display screen. By setting buttons, commands from the outside world can be received to realize human-computer interaction functions. The RS485 serial port is used to realize communication with the host computer, which is used to realize remote control of the vacuum switch, transmission of current recording data, debugging and other functions. The storage structure is set to store information such as the set F0 threshold and judgment results. The power supply part adopts AC220V for power supply, which is converted into DC to power the controller after lightning protection and surge protection, and obtains different levels of device working voltage after various levels of step-down conversion. The control board is equipped with an isolated AD sampling and conditioning block, whose main function is to condition the electrical signal collected by the signal acquisition module, and isolate the signal acquisition module from the core control board. An optical fiber driver board is provided, whose function is to convert the power on and off signal transmitted by the core control board into an optical signal and transmit it to the operating mechanism and drive circuit of the circuit breaker. The signal acquisition part mainly collects the grid voltage and current signals of the voltage transformer PT and the current transformer CT, the opening and closing position signals of the circuit breaker mechanism, and the charging voltage signal of the opening and closing drive circuit, that is, the charging voltage of the energy storage capacitor, and transmits it to the isolated AD sampling and conditioning module of the control board. The signal acquisition part collects the grid voltage and current signals of the voltage transformer PT and the current transformer CT, the opening and closing position signals of the circuit breaker mechanism, and the charging voltage signal of the opening and closing drive circuit, and transmits the collected signals to the isolated AD sampling and conditioning module of the control board. The sampling and conditioning module performs mathematical transformation and other operations on the sampled signals and isolates the signal acquisition part from the core control board. The conditioned signal is transmitted to the core control board through the isolated AD sampling and conditioning module. The core control board judges the fault based on the collected grid voltage and current signals.

[0110] In this preferred embodiment, the present application accurately calculates the current fault estimate by applying the F-test algorithm to determine whether there is a short-circuit fault in the power system. First, the average value of the real-time current data is obtained to provide a benchmark for subsequent comparisons. It not only provides a quantitative fault detection standard, but also can more accurately identify short-circuit faults by comparing the real-time current data with the current expression of historical fitting. The present application is based on statistical principles to ensure the scientificity and accuracy of fault detection. When the current fault estimate exceeds the preset threshold, it indicates that there is a significant difference between the real-time current data and the estimated current value, thereby judging that a short-circuit fault may have occurred in the system. This method based on statistical analysis improves the sensitivity and reliability of fault detection, helps to take preventive or repair measures in a timely manner, reduces the losses caused by faults in the power system, and improves the stability and safety of the system.

[0111] S04: If a fault occurs, the real-time current data is predicted according to a zero point prediction algorithm to obtain a predicted zero point of the short-circuit fault current.

[0112] As a preferred embodiment of the first embodiment, the real-time current data is predicted according to the zero point prediction algorithm to obtain the predicted zero point of the short-circuit fault current, specifically:

[0113] When a fault is detected, the controller collects the voltage and current waveform signals of the circuit through the voltage transformer and the current transformer respectively. The signals are input into the phase control device after isolation and conditioning. After the signals are digitally processed by DSP, the current zero point is calculated by the fault current expression obtained by the prony algorithm, and a tripping optical signal is sent to the circuit breaker through the optical fiber driver board to realize phase-controlled breaking.

[0114] Among them, the zero point prediction algorithm includes safe point algorithm, adaptive self-detection algorithm, arithmetic recursion method, weighted minimum mean square error algorithm, improved half-wave Fourier algorithm, Prony algorithm, recursive least squares algorithm and other methods. The sampling and calculation time of the fitting of the normal current expression at this stage is not included in the fault detection time, so the current fitting at normal time can be calculated through a large number of sampling points. The zero point prediction algorithm needs to consider the sampling time and minimize the sampling time while ensuring accuracy.

[0115] In this preferred embodiment, the present application uses a zero-point prediction algorithm to predict real-time current data, which can predict the zero point of the short-circuit fault current, thereby optimizing the operation timing of the circuit breaker. Since the zero-point prediction algorithm includes a variety of advanced algorithms, such as a safe point algorithm, an adaptive self-detection algorithm, an arithmetic recursion method, a weighted minimum mean square error algorithm, an improved half-wave Fourier algorithm, a Prony algorithm or a recursive least squares algorithm, these algorithms are all intended to improve the accuracy and response speed of the prediction. Through these algorithms, the system can analyze the dynamic changes of the current data in real time and quickly identify the position of the zero point in the current waveform. This precise zero-point prediction capability enables the circuit breaker to perform a breaking operation when the current is close to zero, thereby minimizing arc generation and contact erosion, and improving the service life of the circuit breaker and the reliability of the system. In addition, accurate zero-point prediction also helps to reduce the voltage shock caused by current mutations and protect other components of the power system from damage. Therefore, the present method significantly improves the stability and safety of the power system by improving the accuracy and efficiency of fault current interruption.

[0116] S05: interrupting the short-circuit fault current according to the predicted zero point and the preset circuit breaker.

[0117] The present application can monitor the current state in real time by acquiring the real-time current data of the power system, and provide an accurate data basis for fault detection. The present application uses the F-test algorithm to compare the real-time current data and the estimated current value, and the calculated current fault estimate can quantify the deviation between the current data and the normal operating state, thereby providing an objective basis for fault judgment. By comparing the current fault estimate with the preset threshold, it is possible to quickly and accurately determine whether there is a short-circuit fault in the system. This method reduces the dependence on the time of the fault occurrence and improves the response speed of fault detection. Once the fault is determined, the zero-point prediction algorithm can predict the zero point of the short-circuit fault current, ensuring that the circuit breaker is disconnected when the current is close to zero, which can minimize the ablation of the circuit breaker contacts, reduce the risk of operating overvoltage, and increase the service life of the circuit breaker. The present application not only improves the accuracy and speed of fault detection, but also optimizes the operation timing of the circuit breaker to solve the problem that the short-circuit fault cannot be accurately controlled in the prior art.

[0118] Embodiment 2

[0119] Please refer to Figure 6 , is a short circuit fault control device provided in an embodiment of the present application.

[0120] In this embodiment, the short circuit fault control device includes an acquisition module 10 , a calculation module 20 , a comparison module 30 , a prediction module 40 and a disconnection module 50 .

[0121] The acquisition module 10 is used to acquire real-time current data.

[0122] The calculation module 20 is used to calculate and obtain the current fault estimation value according to the F-test algorithm, the real-time current data and the estimated current value.

[0123] As a preferred embodiment of the second embodiment, the estimated current value is obtained by fitting and calculating according to historical current data, specifically:

[0124] Collect the current data of historical normal conditions. Since the system is in normal operation most of the time, the estimated current can be obtained through a large number of sampling points.

[0125] The collected normal operating current data is used to fit the short-circuit current expression through the prony algorithm, and the estimated current value is calculated based on the historical current expression.

[0126] More specifically, the fitting process of the prony algorithm is as follows:

[0127] The current waveform expression of normal operation is:

[0128]

[0129] Among them, I0 is the amplitude of the fundamental component of the power frequency, is the fundamental phase.

[0130] The mathematical model of the prony algorithm is:

[0131]

[0132] Among them, A i is the fitting amplitude; α i is the attenuation factor; f i is the fitting frequency; θ i is the fitting phase. p is the fitting order of prony. Combined with Euler's formula, the combination of two complex exponentials can form a sinusoidal signal, and the mathematical model of the prony algorithm is transformed into:

[0133]

[0134] When the attenuation factor α i =0, the sampling signal x(t) has no damping characteristics and can be regarded as a fit of sinusoidal signals of various orders that change periodically; combined with the normal operating current waveform expression, we know that p=1.

[0135] According to the prony algorithm model, when the parameters f, θ, A, and α are obtained, the expression corresponding to the current can be obtained.

[0136] Discretize the continuous-time Prony algorithm to get:

[0137]

[0138] Where n is the signal-time data sequence, n=0, ... N-1, N is the sampling window length used by the prony algorithm, and Δt represents the sampling time interval. It is found that the simplified parameter form is consistent with the form of the homogeneous solution of the linear difference equation with constant coefficients. Therefore, in order to obtain the above characterization parameters, a characteristic polynomial can be defined to obtain the solution of the homogeneous equation as follows:

[0139]

[0140] in,

[0141]

[0142] According to the simplified prony expression, the sampling data at the previous k moments can be expressed as:

[0143]

[0144] Multiplying both sides by αk and summing them according to the signal order, we can get:

[0145]

[0146] After the coefficient transformation on the right side, we have:

[0147]

[0148] Although the specific value of zi is unknown at this time, z=zi is exactly the characteristic root corresponding to this polynomial, so the second summation term on the right side of the formula about αk is 0, then:

[0149]

[0150] After one least squares algorithm operation, we can get α1, α2, ..., α p , so we can get z1, z2...z p , and then converted into a linear equation for parameter b: Zb = X

[0151] According to the above solution, the parameters of each order can be obtained:

[0152]

[0153] Substitute the prony algorithm expression to find the current expression under historical normal operating conditions.

[0154] More specifically, Figure 2As shown in the figure, according to the parameters obtained by the prony algorithm, the fitting amplitude, attenuation factor, frequency and phase are restored to obtain the current waveform expression under normal operation. Figure 3 This is a waveform comparison diagram of the current estimation value fitting waveform of the prony algorithm in the embodiment of the present application shifted to the right by 1ms. The current estimation value waveform obtained by the prony algorithm overlaps with the original waveform. Figure 3 The current estimation waveform obtained by the prony algorithm is shifted to the right by 1ms, while the original current waveform remains unchanged.

[0155] In this preferred embodiment, the present application first uses a preset algorithm to conduct in-depth analysis and fitting of these data, thereby constructing a mathematical expression of the historical current. This step not only provides an accurate mathematical model for the normal operating state of the power system, but also lays a solid foundation for subsequent fault detection. Using this historical current expression as an estimated current value and comparing it with the current data obtained by real-time sampling of the sensor can effectively identify any current changes that deviate from the normal mode. In addition, through real-time monitoring and intelligent analysis, the present application can quickly respond to and accurately locate faults, reduce the power outage time caused by faults, and improve the stability and reliability of the power system. Finally, the present application significantly reduces the economic losses and safety hazards caused by short-circuit faults in the power system through early prediction and timely response.

[0156] As a preferred embodiment of the second embodiment, the current fault estimation value is calculated based on the F-test algorithm, the real-time current data and the estimated current value, specifically:

[0157] The "F0 hypothesis test" or "significance test of the regression equation" is defined as follows:

[0158]

[0159] in,

[0160]

[0161] Among them, SS R is the regression sum of squares, SS E is the residual sum of squares, m is the number of unknown variables, n is the number of sample values ​​in the test, p is the number of columns in the coefficient matrix, Estimate the current value for the model, is the sampling mean, y i is the current sampling value.

[0162] More specifically, Figure 4As shown, according to the "F0 hypothesis test", the current estimation value, the sampled current data and the sampled current average value obtained in advance by the prony algorithm are substituted into the formula of the "F0 hypothesis test" and compared with the set threshold.

[0163] The comparison module 30 is used to compare the current fault estimation value with a preset threshold value to determine whether a fault occurs.

[0164] As a preferred embodiment of the second embodiment, the current fault estimation value is compared with a preset threshold value to determine whether a fault occurs, specifically:

[0165] According to the expression of F0 hypothesis test, it can be seen that when no fault occurs, the current sampling data and the current estimated value data are equal, and F0 should be infinite; when a fault occurs, the current sampling data and the current estimated value data are not equal, and F0 will be a relatively small number. By setting the threshold of F0, the occurrence of the fault can be judged. If the obtained F0 value is less than the set threshold, it is judged that a fault has occurred. Otherwise, it returns to the "F0 hypothesis test" detection part and obtains F0 again for comparison.

[0166] More specifically, Figure 5 As shown, it is the algorithm flow of fault identification of the present invention. The current estimation value is first obtained by fitting the prony algorithm, and the F0 value is tested by the "F0 hypothesis test". The current estimation value obtained by fitting the prony algorithm: the power system is in normal operation most of the time, and the fault occupies a relatively short period of time in the entire system operation. First, the current data under normal conditions is collected, and the waveform expression of the current estimation value is obtained in advance. The "F0 hypothesis test" tests the F0 value part: the current estimation value, the sampled current data and the sampled current average value obtained in advance by the prony algorithm are substituted into the formula of the "F0 hypothesis test", and compared with the set threshold value. If the obtained F0 value is less than the set threshold value, it is judged that a fault has occurred. Otherwise, it returns to the "F0 hypothesis test" detection part and obtains F0 again for comparison.

[0167] Since the F0 value is generally large, f0 = log(F0) is set as the criterion. When the system is operating normally, if the noise is not considered, the current estimation value and the sampled current value are equal, and f0 tends to infinity. When the signal-to-noise ratio is 20dB and the sampling frequency is 4kHz, within one power frequency cycle, n = 80, p = 2, m = 2, and then it can be calculated that f0 ≈ 3.591 06. In practical applications, in order to ensure the sensitivity of the algorithm and the convenience of calculation, the f0 threshold is set to 3.45.

[0168] Figure 5The hardware part of the controller (phase control device) in the control panel mainly includes the control panel core, signal acquisition part, opening and closing and charging control part, power supply part, and communication part. It is mainly responsible for real-time monitoring of the voltage and current status of the line, and performing phase control opening operation at the appropriate time. The main function of the control panel part is to process the collected signals, obtain various parameters by prony algorithm, and then calculate according to the "F0 hypothesis test". If the obtained F0 value is less than the set threshold, it is judged that a fault has occurred. Otherwise, it returns to the "F0 hypothesis test" detection part and obtains F0 again for comparison. In addition, the parameters can be set by pressing buttons, and the current information of the system can be displayed in real time through the display screen. By setting buttons, commands from the outside world can be received to realize human-computer interaction functions. The RS485 serial port is used to realize communication with the host computer, which is used to realize remote control of the vacuum switch, transmission of current recording data, debugging and other functions. The storage structure is set to store information such as the set F0 threshold and judgment results. The power supply part adopts AC220V for power supply, which is converted into DC to power the controller after lightning protection and surge protection, and obtains different levels of device working voltage after various levels of step-down conversion. The control board is equipped with an isolated AD sampling and conditioning block, whose main function is to condition the electrical signal collected by the signal acquisition module, and isolate the signal acquisition module from the core control board. An optical fiber driver board is provided, whose function is to convert the power on and off signal transmitted by the core control board into an optical signal and transmit it to the operating mechanism and drive circuit of the circuit breaker. The signal acquisition part mainly collects the grid voltage and current signals of the voltage transformer PT and the current transformer CT, the opening and closing position signals of the circuit breaker mechanism, and the charging voltage signal of the opening and closing drive circuit, that is, the charging voltage of the energy storage capacitor, and transmits it to the isolated AD sampling and conditioning module of the control board. The signal acquisition part collects the grid voltage and current signals of the voltage transformer PT and the current transformer CT, the opening and closing position signals of the circuit breaker mechanism, and the charging voltage signal of the opening and closing drive circuit, and transmits the collected signals to the isolated AD sampling and conditioning module of the control board. The sampling and conditioning module performs mathematical transformation and other operations on the sampled signals and isolates the signal acquisition part from the core control board. The conditioned signal is transmitted to the core control board through the isolated AD sampling and conditioning module. The core control board judges the fault based on the collected grid voltage and current signals.

[0169] In this preferred embodiment, the present application accurately calculates the current fault estimate by applying the F-test algorithm to determine whether there is a short-circuit fault in the power system. First, the average value of the real-time current data is obtained to provide a benchmark for subsequent comparisons. It not only provides a quantitative fault detection standard, but also can more accurately identify short-circuit faults by comparing the real-time current data with the current expression of historical fitting. The present application is based on statistical principles to ensure the scientificity and accuracy of fault detection. When the current fault estimate exceeds the preset threshold, it indicates that there is a significant difference between the real-time current data and the estimated current value, thereby judging that a short-circuit fault may have occurred in the system. This method based on statistical analysis improves the sensitivity and reliability of fault detection, helps to take preventive or repair measures in a timely manner, reduces the losses caused by faults in the power system, and improves the stability and safety of the system.

[0170] The prediction module 40 is used to: if a fault occurs, predict the real-time current data according to a zero point prediction algorithm to obtain a predicted zero point of the short-circuit fault current.

[0171] As a preferred embodiment of the second embodiment, the real-time current data is predicted according to the zero point prediction algorithm to obtain the predicted zero point of the short-circuit fault current, specifically:

[0172] When a fault is detected, the controller collects the voltage and current waveform signals of the circuit through the voltage transformer and the current transformer respectively. The signals are input into the phase control device after isolation and conditioning. After the signals are digitally processed by DSP, the current zero point is calculated by the fault current expression obtained by the prony algorithm, and a tripping optical signal is sent to the circuit breaker through the optical fiber driver board to realize phase-controlled breaking.

[0173] Among them, the zero point prediction algorithm includes safe point algorithm, adaptive self-detection algorithm, arithmetic recursion method, weighted minimum mean square error algorithm, improved half-wave Fourier algorithm, Prony algorithm, recursive least squares algorithm and other methods. The sampling and calculation time of the fitting of the normal current expression at this stage is not included in the fault detection time, so the current fitting at normal time can be calculated through a large number of sampling points. The zero point prediction algorithm needs to consider the sampling time and minimize the sampling time while ensuring accuracy.

[0174] In this preferred embodiment, the present application uses a zero-point prediction algorithm to predict real-time current data, which can predict the zero point of the short-circuit fault current, thereby optimizing the operation timing of the circuit breaker. Since the zero-point prediction algorithm includes a variety of advanced algorithms, such as a safe point algorithm, an adaptive self-detection algorithm, an arithmetic recursion method, a weighted minimum mean square error algorithm, an improved half-wave Fourier algorithm, a Prony algorithm or a recursive least squares algorithm, these algorithms are all intended to improve the accuracy and response speed of the prediction. Through these algorithms, the system can analyze the dynamic changes of the current data in real time and quickly identify the position of the zero point in the current waveform. This precise zero-point prediction capability enables the circuit breaker to perform a breaking operation when the current is close to zero, thereby minimizing arc generation and contact erosion, and improving the service life of the circuit breaker and the reliability of the system. In addition, accurate zero-point prediction also helps to reduce the voltage shock caused by current mutations and protect other components of the power system from damage. Therefore, the present method significantly improves the stability and safety of the power system by improving the accuracy and efficiency of fault current interruption.

[0175] The breaking module 50 is used to break the short-circuit fault current according to the predicted zero point and the preset circuit breaker.

[0176] This device uses five modules to divide the work and coordinate the work to better control the short-circuit fault. This application can monitor the current state in real time by acquiring the real-time current data of the power system, and provide an accurate data basis for fault detection. This application uses the F test algorithm to compare the real-time current data and the estimated current value, and the calculated current fault estimate can quantify the deviation between the current data and the normal operating state, thereby providing an objective basis for fault judgment. By comparing the current fault estimate with the preset threshold, it is possible to quickly and accurately determine whether there is a short-circuit fault in the system. This method reduces the dependence on the time of the fault occurrence and improves the response speed of fault detection. Once the fault is determined, the zero-point prediction algorithm can predict the zero point of the short-circuit fault current to ensure that the circuit breaker is disconnected when the current is close to zero, which can minimize the erosion of the circuit breaker contacts, reduce the risk of operating overvoltage, and increase the service life of the circuit breaker. This application not only improves the accuracy and speed of fault detection, but also optimizes the operation timing of the circuit breaker to solve the problem that the prior art cannot accurately control short-circuit faults.

[0177] The specific embodiments described above further illustrate the purpose, technical solutions 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. It is particularly pointed out that for those skilled in the art, any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the scope of protection of the present invention.

Claims

1. A short circuit fault control method, characterized in that: include: Get real-time current data; Calculating a current fault estimation value according to an F-test algorithm, the real-time current data and the estimated current value; Comparing the current fault estimation value with a preset threshold value to determine whether a fault occurs; If a fault occurs, the real-time current data is predicted according to a zero point prediction algorithm to obtain a predicted zero point of the short-circuit fault current; According to the predicted zero point and the preset circuit breaker, the short-circuit fault current is interrupted.

2. The short circuit fault control method according to claim 1, characterized in that: The estimated current value is obtained by fitting and calculating based on historical current data, specifically: Get historical current data; According to a preset algorithm and the historical current data, a historical current expression is obtained by fitting; According to the historical current expression, an estimated current value is calculated.

3. The short circuit fault control method according to claim 1, characterized in that: The current fault estimation value is calculated based on the F-test algorithm, the real-time current data and the estimated current value, specifically: Determine the regression sum of squares and the residual sum of squares of the current data according to the real-time current data and the estimated current value; According to the expression of the F-test algorithm, the current fault estimation value is calculated; Among them, the expression of the F test algorithm is: In the formula, SS R is the regression sum of squares, SS E is the residual sum of squares, m is the number of unknown variables, n is the number of sample values ​​in the test, p is the number of columns in the coefficient matrix, and F0 is the current fault estimate.

4. The short circuit fault control method according to claim 1, characterized in that: The comparing the current fault estimation value with a preset threshold value to determine whether a fault occurs is specifically as follows: If the current fault estimation value is less than a preset threshold, it is determined that a fault occurs.

5. The short circuit fault control method according to claim 1, characterized in that: The real-time current data is predicted according to the zero-point prediction algorithm, specifically: The zero point prediction algorithm includes a safe point algorithm, an adaptive self-detection algorithm, an arithmetic recursion method, a weighted minimum mean square error algorithm, an improved half-wave Fourier algorithm, a Prony algorithm or a recursive least squares algorithm.

6. A short circuit fault control device, characterized in that: It includes an acquisition module, a calculation module, a comparison module, a prediction module and a disconnection module; The acquisition module is used to obtain real-time current data; The calculation module is used to calculate the current fault estimation value according to the F-test algorithm, the real-time current data and the estimated current value; The comparison module is used to compare the current fault estimation value with a preset threshold value to determine whether a fault occurs; The prediction module is used to predict the real-time current data according to the zero point prediction algorithm if a fault occurs, and obtain the predicted zero point of the short-circuit fault current; The breaking module is used to break the short-circuit fault current according to the predicted zero point and the preset circuit breaker.

7. The short circuit fault control device according to claim 6, characterized in that: The estimated current value is obtained by fitting and calculating based on historical current data, specifically: Get historical current data; According to a preset algorithm and the historical current data, a historical current expression is obtained by fitting; According to the historical current expression, an estimated current value is calculated.

8. The short circuit fault control device according to claim 6, characterized in that: The current fault estimation value is calculated based on the F-test algorithm, the real-time current data and the estimated current value, specifically: Determine the regression sum of squares and the residual sum of squares of the current data according to the real-time current data and the estimated current value; According to the expression of the F-test algorithm, the current fault estimation value is calculated; Among them, the expression of the F test algorithm is: In the formula, SS R is the regression sum of squares, SS E is the residual sum of squares, m is the number of unknown variables, n is the number of sample values ​​in the test, p is the number of columns in the coefficient matrix, and F0 is the current fault estimate.

9. The short circuit fault control device according to claim 6, characterized in that: The comparing the current fault estimation value with a preset threshold value to determine whether a fault occurs is specifically as follows: If the current fault estimation value is less than a preset threshold, it is determined that a fault occurs.

10. The short circuit fault control device according to claim 6, characterized in that: The real-time current data is predicted according to the zero-point prediction algorithm, specifically: The zero point prediction algorithm includes a safe point algorithm, an adaptive self-detection algorithm, an arithmetic recursion method, a weighted minimum mean square error algorithm, an improved half-wave Fourier algorithm, a Prony algorithm or a recursive least squares algorithm.

Images