Power grid diagnosis method and system based on discrete median filter and mathematical morphology

Abstract

The invention discloses a power grid diagnosis method based on a discrete median filter and mathematical morphology, and the method comprises the following steps: 1), collecting a current signal from a measurement unit of a power grid protection relay, and carrying out the preprocessing; step 2), fault features in the preprocessed signals are extracted through a mathematical morphology algorithm, and direct current residual errors are calculated; 3) performing cross matching on the direct current residual error and a preset threshold value to judge whether a fault exists or not, and if the threshold value is exceeded, triggering fault isolation logic; and 4) calculating an energy value based on a mathematical morphology algorithm, judging the type of the fault according to the polarity of the energy value, and positioning the specific position of the fault. The beneficial effects are that the discrete median filter and the mathematical morphology are combined, the discrete median filter can effectively eliminate noise interference, the mathematical morphology algorithm can effectively extract fault features, and the combination of the discrete median filter and the mathematical morphology can improve the accuracy and robustness of fault diagnosis.

Description

Technical Field

[0001] The present invention relates to the field of power grid fault diagnosis, and in particular to a power grid diagnosis method and system based on discrete median filter and mathematical morphology. Background Art

[0002] With the rapid development of renewable energy, direct current distribution system (DCDS) has attracted increasing attention. DCDS has the advantages of high efficiency, low loss, and easy integration of renewable energy, and is considered to be an important development direction of future power systems. However, the fault diagnosis of DCDS faces some challenges. However, the existing fault diagnosis methods have certain limitations when conducting fault diagnosis for DC distribution systems. First, the interference of measurement noise will affect the accuracy of fault detection, resulting in false alarms or missed alarms; second, the rapid rise of DC fault current requires rapid detection and isolation, and the existing technology is difficult to meet the real-time requirements; third, the lack of zero crossing points in the DC system increases the difficulty of fault detection and makes it difficult to distinguish between short-circuit faults and load mutations; in addition, the access of distributed power sources will change the topology of the system and affect the accuracy of the fault diagnosis algorithm; finally, some fault diagnosis technologies require expensive equipment, which increases the cost of the system. Therefore, it is necessary to study a DC distribution system fault diagnosis method that can accurately identify various types of DC faults and detect and locate faults in a short time.

[0003] Existing fault diagnosis methods for DC distribution systems can be mainly divided into three categories: methods based on current and voltage measurements, methods based on traveling waves, and methods based on artificial intelligence.

[0004] Methods based on current and voltage measurements identify fault types and locations by analyzing the waveform characteristics of current and voltage. However, these methods are susceptible to noise interference and have difficulty distinguishing different types of faults.

[0005] The traveling wave-based method identifies the fault type and location by analyzing the traveling waves generated by the fault. However, the traveling wave method requires complex calculations and is difficult to apply to DC microgrids with a large number of distributed power sources.

[0006] Artificial intelligence-based methods use machine learning algorithms to analyze and identify fault data. Artificial intelligence methods require a large amount of training data, and the generalization ability of the model needs to be improved.

[0007] Among the existing patent technologies, some patents also propose fault diagnosis methods based on mathematical morphology or discrete median filters. Patent CN104033583A proposes a DC microgrid fault diagnosis method based on mathematical morphology, but this method is mainly aimed at single-phase DC microgrids and does not consider the influence of noise interference. Patent CN109912533A proposes a DC microgrid fault diagnosis method based on discrete median filters, but this method does not consider the advantages of mathematical morphology in feature extraction. Summary of the invention

[0008] The purpose of the present invention is to overcome the deficiencies of the above-mentioned prior art and provide a high-precision, noise-resistant, fast-response power grid diagnosis method and system, which can effectively solve the challenges of fault diagnosis of DC power distribution systems and has broad application prospects. The present invention is specifically implemented by the following technical solutions:

[0009] The power grid diagnosis method based on discrete median filter and mathematical morphology comprises the following steps:

[0010] Step 1) collecting current signals from a measuring unit of a grid protection relay and performing preprocessing;

[0011] Step 2) extracting fault features from the preprocessed signal through a mathematical morphology algorithm and calculating the DC residual;

[0012] Step 3) cross-matching the DC residual with a preset threshold to determine whether a fault exists, and if it exceeds the threshold, triggering the fault isolation logic;

[0013] Step 4) Calculate the energy value based on the mathematical morphology algorithm, determine the type of fault according to the polarity of the energy value, and locate the specific location where the fault occurs.

[0014] The further design of the power grid diagnosis method based on discrete median filter and mathematical morphology is that the preprocessing in step 1) includes:

[0015] The digital current signal is subjected to anti-aliasing filtering using a Bessel filter to remove signal components above the Nyquist frequency; the analog signal is converted into a digital signal through an analog-to-digital converter and then discretized;

[0016] A discrete median filter is used to estimate the true state of the current signal based on the values ​​of current and past samples; and the signal is smoothed by selecting the median value within each sliding window to remove spikes and pulse noise in the signal.

[0017] The further design of the power grid diagnosis method based on discrete median filter and mathematical morphology is that the step 2) specifically includes:

[0018] The pre-processed current signal is processed by using mathematical morphology opening operation and mathematical morphology closing operation in sequence;

[0019] Calculate the difference between the current signal after the opening operation and the closing operation and the original current signal to obtain the DC residual;

[0020] The mathematical morphology closing operation is analyzed to extract the fault features including the peak value, average value, variance and waveform characteristics of DCR.

[0021] The further design of the power grid diagnosis method based on discrete median filter and mathematical morphology is that the step 3) specifically includes:

[0022] Threshold setting: pre-set one or more thresholds based on system parameters and fault characteristics, combined with noise level, fault current size and fault type;

[0023] DC residual analysis: Compare the calculated DC residual with the preset threshold, analyze the change trend and amplitude of the DC residual, and determine whether there is a fault: if the DC residual exceeds the threshold, it is determined that there is a fault and proceed to the next step; if the DC residual is lower than the threshold, it is determined that there is no fault and the fault detection process ends;

[0024] Fault isolation logic triggering: If the DC residual exceeds the threshold, the fault isolation logic is triggered. The fault isolation logic includes: disconnecting the fault branch, tripping the protection relay, and isolating the faulty equipment. The selection of the fault isolation logic includes factors such as the fault type, fault location, and system topology.

[0025] The further design of the power grid diagnosis method based on discrete median filter and mathematical morphology is that the step 4) specifically includes:

[0026] Step 4-1) Processing the current signal estimated by the discrete median filter using a mathematical morphology algorithm to obtain an energy value;

[0027] Step 4-2) setting a polarity determination rule based on the fault type, and analyzing the polarity of the energy value according to the polarity determination rule;

[0028] Step 4-3) classifying the fault according to the polarity of the energy value, and determining the fault type by comparing with the preset fault classification rules;

[0029] Step 4-4) locating the fault according to the fault type and the polarity of the energy value;

[0030] Step 4-5) Output the information including the fault type and fault location to the human-machine interface or monitoring system.

[0031] A further design of the power grid diagnosis method based on discrete median filter and mathematical morphology is that the polarity determination rule is: an inter-pole short circuit fault will cause the energy value to be negative at both ends, a positive pole grounding fault will cause the energy value to be positive at both ends, and a negative pole grounding fault will cause the energy value to be negative at both ends.

[0032] A further design of the power grid diagnosis method based on discrete median filter and mathematical morphology is that the fault classification rule is: the energy value of the pole-to-pole short circuit fault is a positive value, the energy value of the positive pole-to-ground short circuit fault is a negative value, and the energy value of the negative pole-to-ground short circuit fault is a negative value.

[0033] The present invention also provides a power grid diagnosis system based on discrete median filter and mathematical morphology, comprising the following modules:

[0034] The signal acquisition and preprocessing module obtains the current signal from the measurement unit of the protection relay and performs A / D conversion; a discrete median filter is used to preprocess the signal for state estimation and noise reduction;

[0035] The fault feature extraction module performs morphological opening and closing operations on the preprocessed signal through a mathematical morphological algorithm to extract the fault features in the signal and calculate the DC residual.

[0036] The fault detection and classification module cross-matches the calculated DC residual with the preset threshold to determine whether there is a fault; and triggers the fault isolation logic instruction when the DC residual value exceeds the threshold; at the same time, the polarity of the energy value based on the mathematical morphology algorithm is used to classify and locate the fault and identify different types of faults.

[0037] Fault isolation module: After receiving the instruction from the fault isolation logic, it isolates the fault area to ensure the stable operation of the power grid.

[0038] The present invention also provides a computer-readable storage medium, wherein the storage medium stores a computer program, and the computer program is executed by a processor to implement the power grid diagnosis method based on discrete median filter and mathematical morphology.

[0039] The advantages of the present invention are as follows:

[0040] Compared with similar power grid fault diagnosis methods, the method of the present invention combines discrete median filters and mathematical morphology. Discrete median filters can effectively eliminate noise interference, and mathematical morphology algorithms can effectively extract fault features. The combination of the two can improve the accuracy and robustness of fault diagnosis. At the same time, the method of the present invention can identify various types of faults such as pole-to-pole faults, positive pole-to-ground faults, negative pole-to-ground faults, and can adapt to different fault impedances. In addition, the method of the present invention has ultra-high-speed fault diagnosis capabilities and can complete fault diagnosis within 2.5 milliseconds, meeting the requirements of DC distribution systems for rapid response. Therefore, the method of the present invention has the characteristics of high precision, noise resistance, and rapid response. It can be applied to various DC distribution systems such as DC microgrids, medium-voltage DC distribution systems, and high-voltage DC transmission systems. It has broad application prospects and can provide technical guarantees for the safe and stable operation of DC distribution systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] Figure 1 A schematic flow chart of a power grid diagnosis method based on discrete median filter and mathematical morphology provided in an embodiment of the present invention.

[0042] Figure 2 A schematic diagram of modules of a power grid diagnosis system based on discrete median filter and mathematical morphology provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0043] The technical solution of the present invention is described in detail below.

[0044] Example 1

[0045] The power grid diagnosis method based on discrete median filter and mathematical morphology of this embodiment includes the following steps:

[0046] Step 1) collect current signals from the measuring unit of the grid protection relay and perform preprocessing.

[0047] Step 2) The preprocessed signal is subjected to a mathematical morphology algorithm to extract fault features from the signal and calculate the DC residual.

[0048] Step 3) Cross-match the DC residual with a preset threshold to determine whether a fault exists. If the threshold is exceeded, the fault isolation logic is triggered.

[0049] Step 4) Calculate the energy value based on the mathematical morphology algorithm, determine the type of fault according to the polarity of the energy value, and locate the specific location where the fault occurs.

[0050] The pretreatment in step 1) specifically includes:

[0051] A Bessel filter is used to perform anti-aliasing filtering on the digital current signal to remove signal components higher than the Nyquist frequency;

[0052] After the analog signal is converted to a digital signal through an analog-to-digital converter, it is discretized; anti-aliasing filtering can also be completed using a low-pass filter. The sampling frequency of the analog-to-digital converter is set to 1600Hz to ensure that sufficient signal details can be captured.

[0053] The discrete median filter is used to estimate the true state of the current signal based on the values ​​of the current and past samples, thereby eliminating the influence of measurement errors and noise. At the same time, the discrete median filter is used to smooth the signal by selecting the median in each sliding window, removing the spikes and impulse noise in the signal, and improving the stability of the signal. Thus, the current signal preprocessed by the discrete median filter is obtained, which has high accuracy and low noise level.

[0054] Step 2) specifically includes:

[0055] The pre-processed current signal is processed by mathematical morphology opening operation and mathematical morphology closing operation. The opening operation can eliminate the spikes and burrs in the signal, smooth the signal, and retain the fault characteristics of the signal. The closing operation can fill the small holes in the signal, connect the disconnected contours, and retain the fault characteristics of the signal.

[0056] The difference between the current signal after opening and closing operation and the original current signal is calculated to obtain the DC residual; this signal reflects the part of the signal that is significantly different from the normal state, can highlight the fault characteristics in the signal, and is used for fault detection and classification.

[0057] The mathematical morphology closing operation is analyzed to extract the fault features including the peak value, average value, variance and waveform characteristics of DCR.

[0058] Step 3) specifically includes:

[0059] Threshold setting: one or more thresholds are pre-set based on system parameters and fault characteristics, combined with noise level, fault current size and fault type; in this embodiment, the threshold can be a fixed value or dynamically adjusted to adapt to different operating conditions.

[0060] DC residual analysis: Compare the calculated DC residual with the preset threshold, analyze the change trend and amplitude of the DC residual, and determine whether there is a fault: if the DC residual exceeds the threshold, it is determined that there is a fault and proceed to the next step; if the DC residual is lower than the threshold, it is determined that there is no fault and the fault detection process ends;

[0061] Fault isolation logic triggering: If the DC residual exceeds the threshold, the fault isolation logic is triggered. The fault isolation logic includes: disconnecting the fault branch, tripping the protection relay, and isolating the faulty equipment. The selection of the fault isolation logic needs to be determined based on factors such as the fault type, fault location, and system topology.

[0062] The further design of the power grid diagnosis method based on discrete median filter and mathematical morphology is that step 4) specifically includes:

[0063] Step 4-1) The current signal estimated by the discrete median filter is processed by a mathematical morphology algorithm to obtain an energy value; the energy value is the sum of the signal energy after filtering by the mathematical morphology algorithm, reflecting the shape and structural characteristics of the signal.

[0064] Step 4-2) setting a polarity determination rule based on the fault type, and analyzing the polarity of the energy value according to the polarity determination rule;

[0065] Step 4-3) classifying the fault according to the polarity of the energy value, and determining the fault type by comparing with the preset fault classification rules;

[0066] Step 4-4) The fault is located according to the fault type and the polarity of the energy value; for example, for an inter-pole short circuit fault, the location of the fault can be determined according to the negative value range of the energy value.

[0067] Step 4-5) Output the information including the fault type and fault location to the human-machine interface or monitoring system. The fault information output may include text description, graphic display, sound alarm, etc.

[0068] The polarity determination rule in this embodiment is: an inter-pole short circuit fault will cause the energy values ​​at both ends to be negative, a positive pole grounding fault will cause the energy values ​​at both ends to be positive, and a negative pole grounding fault will cause the energy values ​​at both ends to be negative.

[0069] The fault classification rule in this embodiment is: the energy value of a pole-to-pole short circuit fault is a positive value, the energy value of a positive pole-to-ground short circuit fault is a negative value, and the energy value of a negative pole-to-ground short circuit fault is a negative value.

[0070] Example 2

[0071] like Figure 2 ,This embodiment provides a power network diagnosis system based on discrete median filter and ,mathematical morphology, which is mainly composed of a signal acquisition and ,preprocessing module, a fault feature extraction module, a fault detection and ,classification module, and a fault isolation module.

[0072] The signal acquisition and preprocessing module obtains the current signal from the measuring unit of the protection relay and performs A / D conversion; a discrete median filter is used to preprocess the signal for state estimation and noise reduction.

[0073] The fault feature extraction module performs morphological opening and closing operations on the preprocessed signal through a mathematical morphological algorithm, extracts the fault features in the signal, and calculates the DC residual.

[0074] The fault detection and classification module cross-matches the calculated DC residual with the preset threshold to determine whether there is a fault; and triggers the fault isolation logic instruction when the DC residual value exceeds the threshold; at the same time, the polarity of the energy value based on the mathematical morphology algorithm is used to classify and locate the fault and identify different types of faults.

[0075] The fault isolation module, after receiving the instruction from the fault isolation logic, isolates the fault area to ensure the stable operation of the power grid.

[0076] Example 3

[0077] This embodiment provides a computer-readable storage medium, which stores a computer program. The computer program is executed by a processor to implement a power grid diagnosis method based on a discrete median filter and mathematical morphology.

[0078] The above is only a preferred specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by a person skilled in the art within the technical scope disclosed by the present invention should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims

1. A power grid diagnosis method based on discrete median filter and mathematical morphology, characterized in that: The steps include: Step 1) collecting current signals from a measuring unit of a grid protection relay and performing preprocessing; Step 2) extracting fault features from the preprocessed signal through a mathematical morphology algorithm and calculating the DC residual; Step 3) cross-matching the DC residual with a preset threshold to determine whether a fault exists, and if it exceeds the threshold, triggering the fault isolation logic; Step 4) Calculate the energy value based on the mathematical morphology algorithm, determine the type of fault according to the polarity of the energy value, and locate the specific location where the fault occurs.

2. The power grid diagnosis method based on discrete median filter and mathematical morphology according to claim 1, characterized in that: The pretreatment in step 1) comprises: A Bessel filter is used to perform anti-aliasing filtering on the digital current signal to remove signal components higher than the Nyquist frequency; After the analog signal is converted into a digital signal through an analog-to-digital converter, it is discretized; A discrete median filter is used to estimate the true state of the current signal based on the values ​​of current and past samples; and the signal is smoothed by selecting the median value within each sliding window to remove spikes and pulse noise in the signal.

3. The power grid diagnosis method based on discrete median filter and mathematical morphology according to claim 1, characterized in that: The step 2) specifically includes: The pre-processed current signal is processed by using mathematical morphology opening operation and mathematical morphology closing operation in sequence; Calculate the difference between the current signal after the opening operation and the closing operation and the original current signal to obtain the DC residual; The mathematical morphology closing operation is analyzed to extract the fault features including the peak value, average value, variance and waveform characteristics of DCR.

4. The power grid diagnosis method based on discrete median filter and mathematical morphology according to claim 1, characterized in that: The step 3) specifically includes: Threshold setting: pre-set one or more thresholds based on system parameters and fault characteristics, combined with noise level, fault current size and fault type; DC residual analysis: Compare the calculated DC residual with the preset threshold, analyze the change trend and amplitude of the DC residual, and determine whether there is a fault: if the DC residual exceeds the threshold, it is determined that there is a fault and proceed to the next step; if the DC residual is lower than the threshold, it is determined that there is no fault and the fault detection process ends; Fault isolation logic triggering: If the DC residual exceeds the threshold, the fault isolation logic is triggered. The fault isolation logic includes: disconnecting the fault branch, tripping the protection relay, and isolating the faulty equipment. The selection of the fault isolation logic includes factors such as the fault type, fault location, and system topology.

5. The power grid diagnosis method based on discrete median filter and mathematical morphology according to claim 1, characterized in that: The step 4) specifically includes: Step 4-1) Processing the current signal estimated by the discrete median filter using a mathematical morphology algorithm to obtain an energy value; Step 4-2) setting a polarity determination rule based on the fault type, and analyzing the polarity of the energy value according to the polarity determination rule; Step 4-3) classifying the fault according to the polarity of the energy value, and determining the fault type by comparing with the preset fault classification rules; Step 4-4) locating the fault according to the fault type and the polarity of the energy value; Step 4-5) Output the information including the fault type and fault location to the human-machine interface or monitoring system.

6. The power grid diagnosis method based on discrete median filter and mathematical morphology according to claim 5, characterized in that: The polarity determination rule is: an inter-pole short circuit fault will cause the energy values ​​at both ends to be negative, a positive pole grounding fault will cause the energy values ​​at both ends to be positive, and a negative pole grounding fault will cause the energy values ​​at both ends to be negative.

7. The power grid diagnosis method based on discrete median filter and mathematical morphology according to claim 5, characterized in that: The fault classification rule is: the energy value of a pole-to-pole short-circuit fault is positive, the energy value of a positive pole-to-ground short-circuit fault is negative, and the energy value of a negative pole-to-ground short-circuit fault is negative.

8. A power grid diagnosis system based on discrete median filter and mathematical morphology, characterized in that: Includes the following modules: The signal acquisition and preprocessing module obtains the current signal from the measurement unit of the protection relay and performs A / D conversion; a discrete median filter is used to preprocess the signal for state estimation and noise reduction; The fault feature extraction module performs morphological opening and closing operations on the preprocessed signal through a mathematical morphological algorithm to extract the fault features in the signal and calculate the DC residual. The fault detection and classification module cross-matches the calculated DC residual with the preset threshold to determine whether there is a fault; and triggers the fault isolation logic instruction when the DC residual value exceeds the threshold; at the same time, the polarity of the energy value based on the mathematical morphology algorithm is used to classify and locate the fault and identify different types of faults. Fault isolation module: After receiving the instruction from the fault isolation logic, it isolates the fault area to ensure the stable operation of the power grid.

9. A computer-readable storage medium, characterized in that The storage medium stores a computer program, and the computer program is executed by a processor to implement the power grid diagnosis method based on discrete median filter and mathematical morphology as described in any one of claims 1 to 7.

Images