Mining selective leakage protection method with arc detection function

Abstract

The invention discloses a mining selective electric leakage protection method with an electric arc detection function, which relates to the technical field of electric arc detection, is used for improving the problem of electric arc detection omission caused by timeliness of instantaneous electric arcs, and comprises the following steps: electrifying mining equipment to carry out electric leakage monitoring, and obtaining zero-sequence voltage and zero-sequence current under basic harmonic waves; constructing a ridge regression model according to the zero-sequence voltage and the zero-sequence current, calculating an electric power reference index of the mining equipment, setting an electric leakage threshold value, and calculating electric leakage coefficients at different time points during electric leakage monitoring according to the electric power reference index and the electric leakage threshold value; analyzing the electric leakage instantaneity of different time points by using fuzzy reasoning according to the electric leakage coefficient and the electric leakage time interval, selecting whether to collect zero-sequence current under high-frequency harmonic waves according to an analysis result, comparing the zero-sequence current of each marked time point with a preset alarm threshold value, and obtaining the grade of each marked time point; and judging whether electric leakage alarm is needed or not according to the grade of each marked time point.

Description

Technical Field

[0001] The present invention relates to the technical field of arc detection, and more specifically, to a selective leakage protection method for mines with arc detection function. Background Art

[0002] Arc detection technology is used to identify and monitor arc phenomena. In the power system, arc monitoring technology is used to reduce the risk of equipment damage. Applying arc detection technology to selective leakage in mines can reduce mine fires caused by arcs and other safety hazards.

[0003] The prior art has the following deficiencies:

[0004] Previous arc detection technologies used fundamental harmonics to detect zero-sequence current, determined whether the zero-sequence current was abnormal, and analyzed the leakage arc. However, the time sensitivity of the instantaneous arc was not taken into account. When the leakage arc was highly instantaneous, the zero-sequence current collected by the fundamental harmonics was not identified in time, which would damage the mining equipment and cause safety hazards. Summary of the invention

[0005] In order to overcome the above-mentioned defects of the prior art, an embodiment of the present invention provides a selective leakage protection method for mining with an arc detection function. After setting a threshold for judging leakage by analyzing the zero-sequence current and zero-sequence voltage under the basic harmonic, the leakage conditions of mining equipment at different time points are determined for screening and grading. According to the level at each time point, it is considered whether to select high-frequency harmonics to detect the zero-sequence current and issue an alarm to solve the problems raised in the above-mentioned background technology.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] A selective leakage protection method for mines with arc detection function comprises the following steps:

[0008] Step S1: Connect the mining equipment, power on the mining equipment for leakage monitoring and set the leakage monitoring time, and use the zero-sequence transformer to obtain the zero-sequence current and zero-sequence voltage of the mining equipment under the basic harmonic in real time;

[0009] Step S2: constructing a ridge regression model based on zero-sequence voltage and zero-sequence current to calculate the power benchmark index at different time points and set the leakage threshold of mining equipment, comparing the power benchmark index with the leakage threshold in real time, calculating the leakage arc quantization value at different time points and screening and marking each time point;

[0010] Step S3: obtaining the leakage time interval within the leakage monitoring time according to the marked time point, calculating the leakage coefficient of each marked time point by using the leakage arc quantization value, and analyzing the leakage instantaneity of each marked time point by using fuzzy reasoning based on the leakage coefficient and the leakage time interval;

[0011] Step S4: According to the instantaneous leakage analysis results of each marked time point, select whether to collect the zero-sequence current under the high-frequency harmonics. After obtaining the zero-sequence current under the high-frequency harmonics at each marked time point, compare it with the preset alarm threshold, classify each marked time point according to the comparison result, and determine whether a leakage alarm is needed according to the classification result.

[0012] In a preferred embodiment, when the zero-sequence voltage and zero-sequence current of the incoming mining equipment under the fundamental harmonic are acquired in real time, a period of time is selected as the analysis time, n time points are set within the analysis time, and the zero-sequence voltage and zero-sequence current of the mining equipment at each time point are acquired and merged into a voltage data set and a current data set respectively.

[0013] In a preferred embodiment, the voltage data set and the current data set are integrated to construct a ridge regression model to calculate the power benchmark index of mining equipment. The specific steps are as follows:

[0014] Standardization feature: Use the standardization formula to standardize the data in the voltage data set and the current data set: x nore =x-μ / σ, where x is the data in the voltage data set or the current data set, μ is the average value of the data in the voltage data set or the current data set, σ is the standard deviation of the data in the voltage data set or the current data set, and x nore The normalized value of the data in the voltage data set or the current data set;

[0015] Data matching: give the same mark to the zero-sequence voltage or zero-sequence current at the same time point in the voltage data set and the current data set;

[0016] Model construction: The ridge regression model is constructed by integrating the data with the same labels in the voltage dataset and the current dataset: L(β)=||y-Xβ|| 2 +λ||β|| 2 Among them, ||y-Xβ|| 2 The loss amount obtained by the least squares method, λ||β|| 2 is the L2 regularization term used to control the model complexity, X is the characteristic matrix composed of zero-sequence voltage and zero-sequence current with the same label, and y is the preset calibration parameter;

[0017] Select regularization parameters: Use K-fold cross validation to select regularization parameters, and import the obtained regularization parameters into the ridge regression model;

[0018] Calculate the power benchmark index: use the ridge regression model calculation result L(β) as the power benchmark index at each time point;

[0019] Setting the leakage threshold: taking the average value of the power reference index at each time point and marking it as d, setting the leakage threshold: L(d)=cd, where c is the mapping amount and L(d) is the leakage threshold.

[0020] In a preferred embodiment, the power reference index at different time points is compared with the leakage threshold in real time, the difference between the power reference index and the leakage threshold is calculated as the leakage arc quantization value, the different time points are assigned and sorted in chronological order from small to large, the first time point is assigned 1, and the subsequent time points are added by 1 one by one. When the leakage arc quantization value of the corresponding time point is a positive number, it is screened out and marked.

[0021] In a preferred embodiment, in the marked time points, the time points with a sequence number difference of 1 are merged to obtain a merged time point, the time interval between the two time points is obtained as the leakage time interval, and the two time points are merged to have the sequence number characteristics of the two time points, and the marked time points are continuously traversed to obtain the merged time points and the leakage time interval, until all the marked time points are traversed to obtain multiple leakage time intervals;

[0022] The ratio of the leakage arc quantization value calculated at each marked time point to the standard judgment arc is taken as the leakage coefficient of the corresponding marked time point.

[0023] In a preferred embodiment, the leakage coefficient and the leakage time interval are combined to use fuzzy reasoning to analyze the leakage instantaneity of mining equipment, and the specific steps are as follows:

[0024] Input preprocessing: Use the Max-Min normalization algorithm to process the leakage time interval of the merged time point to obtain the time coefficient corresponding to the merged time point;

[0025] Define input: Take the time coefficient of the merged time point as the time coefficient of all the marked time points contained in it, define the leakage coefficient and time coefficient of each marked time point as input variables, and divide them into different fuzzy sets;

[0026] Define output: Define the instantaneous nature of leakage at the marked time point as the output variable;

[0027] Formulate rules: formulate fuzzy rules to describe the impact of different input variables on output variables;

[0028] Perform fuzzy reasoning: Classify the instantaneous nature of leakage at each marked time point as high or low according to fuzzy rules.

[0029] In a preferred embodiment, after receiving the leakage instantaneousness of each marked time point, statistics are performed, and the ratio of the number of marked time points with high leakage instantaneousness to the number of marked time points with low leakage instantaneousness is used as the harmonic selection coefficient. When the harmonic selection coefficient exceeds 1, the zero-sequence current under the high-frequency harmonic is collected within the merged time point where the corresponding marked time point is located.

[0030] In a preferred embodiment, if the harmonic selection coefficient of the marked time point is lower than 1, the zero-sequence current under the fundamental harmonic at the corresponding marked time point is compared with the alarm threshold; if the harmonic selection coefficient of the marked time point is higher than 1, the zero-sequence current under the high-frequency harmonic at the corresponding marked time point is compared with the alarm threshold.

[0031] In a preferred embodiment, the difference between the zero-sequence current at each marked time point and the alarm threshold is used as the classification level. When the classification level of the marked time point exceeds the difference between the average value of the zero-sequence current at the marked time point under all fundamental harmonics and the alarm threshold, the marked time point is set as a high-level alarm point; otherwise, the marked time point is set as a low-level alarm point.

[0032] When the marked time point is a low-level alarm point, no leakage alarm is issued; when the marked time point is a high-level alarm point, a leakage alarm is issued.

[0033] The technical effects and advantages of the mine selective leakage protection method with arc detection function of the present invention are as follows:

[0034] The present invention performs leakage monitoring on power-on mining equipment, captures zero-sequence voltage and zero-sequence current of mining equipment under basic harmonics, calculates power reference index of mining equipment according to zero-sequence voltage and zero-sequence current and sets leakage threshold, compares power reference index with leakage threshold to calculate leakage arc, adds zero-sequence voltage factor to calculate leakage arc to improve data accuracy, calculates leakage coefficient at different time points according to leakage arc, analyzes leakage instantaneity at different time points based on leakage coefficient and leakage time interval, effectively locks each monitoring time point by analyzing leakage instantaneity, selects whether to collect zero-sequence current under high-frequency harmonics according to analysis result, compares zero-sequence current at each marked time point with preset alarm threshold and obtains level of each marked time point, determines whether leakage alarm is needed according to level of each marked time point, and improves safety guarantee of mining equipment in case of instantaneous leakage. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1 The present invention is a schematic diagram of a selective leakage protection method for mines with arc detection function. DETAILED DESCRIPTION

[0036] 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.

[0037] The present invention performs leakage monitoring on power-on mining equipment, captures zero-sequence voltage and zero-sequence current of mining equipment under basic harmonics, calculates power reference index of mining equipment according to zero-sequence voltage and zero-sequence current and sets leakage threshold, compares power reference index with leakage threshold to calculate leakage arc, calculates leakage coefficient at different time points according to leakage arc, analyzes leakage instantaneity at different time points by comprehensively considering leakage coefficient and leakage time interval, selects whether to collect zero-sequence current under high-frequency harmonics according to the analysis result, compares zero-sequence current at each marked time point with preset alarm threshold and obtains level of each marked time point, determines whether leakage alarm is needed according to level of each marked time point, and improves safety guarantee of mining equipment in case of instantaneous leakage.

[0038] Embodiment, a selective leakage protection method for mine with arc detection function, such as Figure 1 As shown, the following steps are included:

[0039] Step S1: Connect the mining equipment, power on the mining equipment for leakage monitoring and set the leakage monitoring time, and use the zero-sequence transformer to obtain the zero-sequence current and zero-sequence voltage of the mining equipment under the basic harmonic in real time;

[0040] Step S2: constructing a ridge regression model based on zero-sequence voltage and zero-sequence current to calculate the power benchmark index at different time points and set the leakage threshold of mining equipment, comparing the power benchmark index with the leakage threshold in real time, calculating the leakage arc quantization value at different time points and screening and marking each time point;

[0041] Step S3: obtaining the leakage time interval within the leakage monitoring time according to the marked time point, calculating the leakage coefficient of each marked time point by using the leakage arc quantization value, and analyzing the leakage instantaneity of each marked time point by using fuzzy reasoning based on the leakage coefficient and the leakage time interval;

[0042] Step S4: According to the instantaneous leakage analysis results of each marked time point, select whether to collect the zero-sequence current under the high-frequency harmonics. After obtaining the zero-sequence current under the high-frequency harmonics at each marked time point, compare it with the preset alarm threshold, classify each marked time point according to the comparison result, and determine whether a leakage alarm is needed according to the classification result.

[0043] It should be noted that the zero-sequence current includes the leakage arc of mining equipment. The leakage arc is estimated by comparing the zero-sequence current with the leakage threshold. The zero-sequence voltage is added to calculate the power reference index to improve the accuracy of the leakage arc estimation.

[0044] In step S1, when the mining equipment is powered on for a working test, the voltage applied to the mining equipment is first converted by using a PT voltage transformer, and then the zero-sequence voltage and zero-sequence current passing through the mining equipment are collected by using a zero-sequence transformer.

[0045] It should be noted that the PT voltage transformer is a conversion device in the power system, which is used to convert high voltage into low voltage for measurement and monitoring. In this example, the PT voltage transformer is used to convert the voltage passed into the mining equipment; the zero-sequence transformer is a device that detects and measures zero-sequence current and zero-sequence voltage in the power system, and is mainly used for monitoring unbalanced current and voltage.

[0046] In step S2, the zero-sequence voltage and zero-sequence current of the mining equipment under the fundamental harmonic are received in real time, a period of time is selected as the analysis time, n time points are set within the analysis time, and the zero-sequence voltage and zero-sequence current of the mining equipment at each time point are obtained and merged into a voltage data set and a current data set respectively;

[0047] The voltage data set and the current data set are combined to construct a ridge regression model to calculate the power benchmark index of mining equipment. The specific steps are as follows:

[0048] Standardization feature: Use the standardization formula to standardize the data in the voltage data set and the current data set: x nore =x-μ / σ, where x is the data in the voltage data set or the current data set, μ is the average value of the data in the voltage data set or the current data set, σ is the standard deviation of the data in the voltage data set or the current data set, and x nore The normalized value of the data in the voltage data set or the current data set;

[0049] Data matching: give the same mark to the zero-sequence voltage or zero-sequence current at the same time point in the voltage data set and the current data set;

[0050] Model construction: The ridge regression model is constructed by integrating the data with the same labels in the voltage dataset and the current dataset: L(β)=||y-Xβ||| 2 +λ||β|| 2 Among them, ||y-Xβ|| 2 The loss amount obtained by the least squares method, λ||β|| 2 is the L2 regularization term used to control the model complexity, X is the characteristic matrix composed of zero-sequence voltage and zero-sequence current with the same label, and y is the preset calibration parameter;

[0051] Select regularization parameters: Use K-fold cross validation to select regularization parameters, and import the obtained regularization parameters into the ridge regression model;

[0052] Calculate the power benchmark index: use the ridge regression model calculation result L(β) as the power benchmark index at each time point;

[0053] Setting the leakage threshold: taking the average value of the power reference index at each time point and marking it as d, setting the leakage threshold: L(d)=cd, where c is the mapping amount and L(d) is the leakage threshold.

[0054] It should be explained that the above-mentioned standardized features process the data in the voltage data set and the current data set to make their mean 0 and variance 1, thereby improving the stability of the model and setting the leakage threshold using the power benchmark index at different time points, which deeply improves the accuracy of subsequent leakage judgments. K-fold cross-validation is a model evaluation technology that reduces the randomness and bias in model evaluation by dividing the data set into K subsets.

[0055] The power benchmark index at different time points is compared with the leakage threshold in real time, and the difference between the power benchmark index and the leakage threshold is calculated as the leakage arc quantization value. The different time points are assigned values ​​and sorted in chronological order from small to large. The first time point is assigned a value of 1, and the subsequent time points are added by 1 one by one. When the leakage arc quantization value of the corresponding time point is a positive number, it is screened out and marked.

[0056] In step S3, for the marked time points, the time points with a sequence number difference of 1 are merged to obtain a merged time point, and the time interval between the two time points is obtained as the leakage time interval. In addition, after the two time points are merged, they have the sequence number characteristics of the two time points, and the marked time points are continuously traversed to obtain the merged time points and the leakage time interval, until all the marked time points are traversed to obtain multiple leakage time intervals;

[0057] The ratio of the leakage arc quantization value calculated at each marked time point to the standard judgment arc is taken as the leakage coefficient of the corresponding marked time point.

[0058] It should be noted that the larger the leakage coefficient of the marked time electricity, the larger the leakage arc quantization value, the more serious the leakage situation, the shorter the leakage time interval, and the higher the instantaneous leakage at the corresponding merged time point. The standard judgment arc is set by professionals in this field based on the acceptable leakage amount, and will not be elaborated here.

[0059] The leakage coefficient and leakage time interval are combined to use fuzzy reasoning to analyze the leakage instantaneity of mining equipment. The specific steps are as follows:

[0060] Input preprocessing: Use the Max-Min normalization algorithm to process the leakage time interval of the merged time point: x nore =xx min / x max -x min , where x is the leakage time interval at the combined time point, x min is the minimum value of the leakage time interval of all combined time points, x max is the maximum value of the leakage time interval of all combined time points, x nore The normalized result of the leakage time corresponding to the combined time point is used as the time coefficient corresponding to the combined time point;

[0061] Define input: The time coefficient of the merged time point is used as the time coefficient of all the marked time points contained in it, and the leakage coefficient and time coefficient of each marked time point are defined as input variables, and divided into different fuzzy sets. For example, "High" and "Low" correspond to the leakage coefficient of the marked time point, and "Long" and "Short" correspond to the time coefficient of the marked time point.

[0062] Define output: Define the instantaneous leakage at the marked time point as the output variable. For example, define "Strength" and "Weakness" as the instantaneous leakage at the marked time point.

[0063] Formulate rules: Formulate fuzzy rules to describe the impact of different input variables on output variables. For example, mark the leakage coefficient at the marked time point as L, mark the time coefficient at the marked time point as S, and mark the instantaneous leakage at the corresponding marked time point as E, then you can define

[0064] Rule1: IF(L is High)AND(S is Short)THEN(E is Strength)

[0065] Rule2: IF(L is Low)AND(S is Long)THEN(E is Weakness) ...

[0067] Perform fuzzy reasoning: when the output result is "Strength", the instantaneous leakage of the corresponding marked time point is judged to be high; when the output result is "Weakness", the instantaneous leakage of the corresponding marked time point is judged to be low.

[0068] It should be noted that the classification of fuzzy sets can be adjusted according to actual conditions in order to improve the accuracy of harmonic selection. This example takes two fuzzy sets as an example, and three or more fuzzy sets can also be used for classification. In addition, the high and low input variables in this example can be judged by setting the accuracy threshold. For example, when the leakage coefficient of the marked time point exceeds 1, the leakage coefficient of the marked time point is calibrated as "High"; when the time coefficient of the marked time point is lower than the average time coefficient of all marked time points, it is calibrated as "Short", etc., and no further analysis will be made here.

[0069] In step S4, the leakage instantaneousness of each marked time point is received and then counted, and the ratio of the number of marked time points with high leakage instantaneousness to the number of marked time points with low leakage instantaneousness is used as the harmonic selection coefficient. When the harmonic selection coefficient exceeds 1, the zero-sequence current under the high-frequency harmonic is collected within the merged time point where the corresponding marked time point is located.

[0070] When collecting zero-sequence current under high-frequency harmonics, the transient fortieth harmonic difference current principle is used for detection. The specific steps are as follows:

[0071] Step 1: Use a high sampling rate analog-to-digital converter to collect data at a frequency exceeding 51.2 kHz;

[0072] Step 2: Connect the zero-sequence current sensor and install it to the neutral point or grounding wire of the transformer to monitor the zero-sequence current in real time;

[0073] Step 3: Use ADC to collect zero-sequence current signals in real time, measure the transformer's incoming and outgoing currents respectively, and calculate the zero-sequence current using the following formula: I0 = I in +I out +I neutral , where I0 is the zero sequence current, I in is the incoming line current, I out is the outgoing current, I neutral is the neutral current;

[0074] Step 4: Collect the zero-sequence current signal at each marked time point, as well as the amplitude and phase of the fortieth harmonic, and use fast Fourier transform to match and analyze the collected zero-sequence current information to obtain the zero-sequence current at each marked time point.

[0075] If the harmonic selection coefficient of the marked time point is lower than 1, the zero-sequence current under the fundamental harmonic at the corresponding marked time point is compared with the alarm threshold. If the harmonic selection coefficient of the marked time point is higher than 1, the zero-sequence current under the high-frequency harmonic at the corresponding marked time point is compared with the alarm threshold.

[0076] It should be noted that, in this example, the fundamental harmonic is 50 Hz and the high-frequency harmonic is 2000 Hz. The acquisition efficiency of the high-frequency harmonic is much greater than that of the fundamental harmonic, which improves the accuracy and anti-interference ability of the data acquisition and detection process.

[0077] The difference between the zero-sequence current and the alarm threshold at each marked time point is taken as the classification level. When the classification level of the marked time point exceeds the difference between the average value of the zero-sequence current at the marked time point under all fundamental harmonics and the alarm threshold, the marked time point is set as a high-level alarm point; otherwise, the marked time point is set as a low-level alarm point.

[0078] When the marked time point is a low-level alarm point, a leakage alarm is issued; when the marked time point is a high-level alarm point, a leakage alarm is issued.

[0079] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination. When implemented by software, the above embodiments may be implemented in whole or in part in the form of a computer program product.

[0080] Those of ordinary skill in the art will appreciate that the modules and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application of the technical solution and the invention constraints. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0081] In addition, each functional module in each embodiment of the present application may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module.

[0082] The above is only a specific implementation of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art who is familiar with the present technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application. Therefore, the protection scope of the present application should be based on the protection scope of the claims.

[0083] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A selective leakage protection method for mines with arc detection function, characterized in that: The following steps are involved: Step S1: Connect the mining equipment, power on the mining equipment for leakage monitoring and set the leakage monitoring time, and use the zero-sequence transformer to obtain the zero-sequence current and zero-sequence voltage of the mining equipment under the basic harmonic in real time; Step S2: constructing a ridge regression model based on zero-sequence voltage and zero-sequence current to calculate the power benchmark index at different time points and set the leakage threshold of mining equipment, comparing the power benchmark index with the leakage threshold in real time, calculating the leakage arc quantization value at different time points and screening and marking each time point; Step S3: obtaining the leakage time interval within the leakage monitoring time according to the marked time point, calculating the leakage coefficient of each marked time point by using the leakage arc quantization value, and analyzing the leakage instantaneity of each marked time point by using fuzzy reasoning based on the leakage coefficient and the leakage time interval; Step S4: According to the instantaneous leakage analysis results of each marked time point, select whether to collect the zero-sequence current under the high-frequency harmonics. After obtaining the zero-sequence current under the high-frequency harmonics at each marked time point, compare it with the preset alarm threshold, classify each marked time point according to the comparison result, and determine whether a leakage alarm is needed according to the classification result.

2. A selective leakage protection method for mines with arc detection function according to claim 1, characterized in that: When the zero-sequence voltage and zero-sequence current of the incoming mining equipment under the fundamental harmonic are obtained in real time, a period of time is selected as the analysis time, n time points are set within the analysis time, and the zero-sequence voltage and zero-sequence current of the mining equipment at each time point are obtained and merged into a voltage data set and a current data set respectively.

3. The selective leakage protection method for mines with arc detection function according to claim 1, characterized in that: The voltage data set and the current data set are combined to construct a ridge regression model to calculate the power benchmark index of mining equipment. The specific steps are as follows: Standardization feature: Use the standardization formula to standardize the data in the voltage data set and the current data set: x nore =x-μ / σ, where x is the data in the voltage data set or the current data set, μ is the average value of the data in the voltage data set or the current data set, σ is the standard deviation of the data in the voltage data set or the current data set, and x nore The normalized value of the data in the voltage data set or the current data set; Data matching: give the same mark to the zero-sequence voltage or zero-sequence current at the same time point in the voltage data set and the current data set; Model construction: The ridge regression model is constructed by integrating the data with the same labels in the voltage dataset and the current dataset: L(β)=||y-Xβ|| 2 +λ||β|| 2 , where ||y-Xβ|| 2 The loss amount obtained by the least squares method, λ||β|| 2 is the L2 regularization term used to control the model complexity, X is the characteristic matrix composed of zero-sequence voltage and zero-sequence current with the same label, and y is the preset calibration parameter; Select regularization parameters: Use K-fold cross validation to select regularization parameters, and import the obtained regularization parameters into the ridge regression model; Calculate the power benchmark index: use the ridge regression model calculation result L(β) as the power benchmark index at each time point; Setting the leakage threshold: taking the average value of the power reference index at each time point and marking it as d, setting the leakage threshold: L(d)=cd, where c is the mapping amount and L(d) is the leakage threshold.

4. A selective leakage protection method for mines with arc detection function according to claim 3, characterized in that: The power benchmark index at different time points is compared with the leakage threshold in real time, and the difference between the power benchmark index and the leakage threshold is calculated as the leakage arc quantization value. The different time points are assigned values ​​and sorted in chronological order from small to large. The first time point is assigned a value of 1, and the subsequent time points are added by 1 one by one. When the leakage arc quantization value of the corresponding time point is a positive number, it is screened out and marked.

5. A selective leakage protection method for mines with arc detection function according to claim 4, characterized in that: In the marked time points, the time points with a sequence number difference of 1 are merged to obtain a merged time point, and the time interval between the two time points is obtained as the leakage time interval. After the two time points are merged, the sequence number characteristics of the two time points are obtained, and the marked time points are continued to be traversed to obtain the merged time points and the leakage time interval, until all the marked time points are traversed to obtain multiple leakage time intervals; The ratio of the leakage arc quantization value calculated at each marked time point to the standard judgment arc is taken as the leakage coefficient of the corresponding marked time point.

6. A selective leakage protection method for mines with arc detection function according to claim 5, characterized in that: The leakage coefficient and leakage time interval are combined to use fuzzy reasoning to analyze the instantaneous leakage of mining equipment. The specific steps are as follows: Input preprocessing: Use the Max-Min normalization algorithm to process the leakage time interval of the merged time point to obtain the time coefficient corresponding to the merged time point; Define input: Take the time coefficient of the merged time point as the time coefficient of all the marked time points contained in it, define the leakage coefficient and time coefficient of each marked time point as input variables, and divide them into different fuzzy sets; Define output: Define the instantaneous nature of leakage at the marked time point as the output variable; Formulate rules: formulate fuzzy rules to describe the impact of different input variables on output variables; Perform fuzzy reasoning: Classify the instantaneous nature of leakage at each marked time point as high or low according to fuzzy rules.

7. A selective leakage protection method for mines with arc detection function according to claim 6, characterized in that: After receiving the leakage instantaneousness of each marked time point, statistics are performed, and the ratio of the number of marked time points with high leakage instantaneousness to the number of marked time points with low leakage instantaneousness is used as the harmonic selection coefficient. When the harmonic selection coefficient exceeds 1, the zero-sequence current under high-frequency harmonics is collected within the merged time point where the corresponding marked time point is located.

8. A selective leakage protection method for mines with arc detection function according to claim 7, characterized in that: If the harmonic selection coefficient of the marked time point is lower than 1, the zero-sequence current under the fundamental harmonic at the corresponding marked time point is compared with the alarm threshold. If the harmonic selection coefficient of the marked time point is higher than 1, the zero-sequence current under the high-frequency harmonic at the corresponding marked time point is compared with the alarm threshold.

9. A selective leakage protection method for mines with arc detection function according to claim 8, characterized in that: The difference between the zero-sequence current at each marked time point and the alarm threshold is used as the classification level. When the classification level of the marked time point exceeds the difference between the average value of the zero-sequence current at the marked time point under all fundamental harmonics and the alarm threshold, the marked time point is set as a high-level alarm point. Otherwise, the marked time point is set as a low-level alarm point; When the marked time point is a low-level alarm point, no leakage alarm is issued; When the marked time point is a high-level alarm point, a leakage alarm is issued.

Images