Mine time shift resistivity monitoring poor grounding emission electrode discrimination method

Abstract

The invention relates to a mine time-shifting resistivity monitoring poor grounding emission electrode discrimination method, which can significantly improve the precision and automation level of electrode grounding state discrimination from data collection, model construction to real-time monitoring full-process automatic processing, realizes scientific, accurate and real-time identification of the electrode grounding state, and improves the reliability of the electrode grounding state discrimination. And the efficiency of the mine safety monitoring system is obviously improved.

Description

Technical Field

[0001] The invention belongs to the technical field of geophysical exploration, and in particular relates to a method for distinguishing a poorly grounded transmitting electrode in mine time-lapse resistivity monitoring. Background Art

[0002] Coal has long been the main energy source in my country, and the conditions of my country's coal resources are complex. Time-lapse resistivity monitoring of mines is an important means of predicting and preventing geological disasters in mines. Its principle is to detect the stratum structure, groundwater conditions, ore body distribution and potential signs of geological disasters (such as goaf collapse, karst collapse, fault activity, etc.) by transmitting a current of a specific frequency underground and measuring the resistivity distribution that changes with time. The monitoring system usually consists of ground transmitting electrodes, receiving electrodes, and signal acquisition and processing equipment. Good grounding of the transmitting electrode is one of the key factors to ensure the accuracy of monitoring data, which directly affects the effective injection of current and the accurate measurement of resistivity. Poor grounding of the transmitting electrode mainly includes increased grounding resistance, current leakage or poor impedance matching caused by factors such as grounding electrode corrosion, loose contact, and changes in geological structure. These problems will cause the following adverse effects: 1 Data deviation: Poor grounding causes the current to be unable to be effectively injected into the ground, causing the measured resistivity value to deviate from the true value, affecting the accuracy of geological structure analysis. 2 Monitoring blind area: Current leakage may cause the resistivity signal in a local area to weaken or disappear, forming a monitoring blind area, and unable to effectively detect potential geological disasters. 3 System stability: Poor grounding may lead to unstable current loop, causing monitoring signal fluctuations, increased noise, and reducing the overall stability of the system.

[0003] The traditional method of judging whether the transmitting electrode is poorly grounded mainly relies on manual inspection, regular grounding resistance test and experience judgment, which has the following limitations: 1. Lag: The manual inspection cycle is long, and it is difficult to detect sudden poor grounding problems in time. 2. Subjectivity: Experience judgment is easily affected by factors such as personal skill level and on-site conditions, and the accuracy is not high. 3. Inefficiency: Conventional grounding resistance testing requires interrupting normal monitoring work, and can only reflect the grounding status at a certain moment, and cannot achieve continuous monitoring. Summary of the invention

[0004] In view of the above problems, the purpose of the present invention is to provide a method for identifying poorly grounded transmitting electrodes in mine time-lapse resistivity monitoring, so as to solve the problems of lag, subjectivity, inefficiency, etc. in the existing methods for identifying poorly grounded transmitting electrodes, and fail to meet the needs of real-time and accurate monitoring.

[0005] To achieve the above object, the technical solution adopted by the present invention includes:

[0006] A method for identifying a poorly grounded transmitting electrode in mine time-lapse resistivity monitoring comprises the following steps:

[0007] Step 1, collecting N groups of receiving electrodes corresponding to transmitting current sets, including M groups of transmitting electrodes in good ground coupling state and NM groups of transmitting electrodes in poor ground coupling state, selecting the corresponding receiving electrode transmitting current set corresponding to the M0 group of transmitting electrodes in good ground coupling state and the corresponding receiving electrode transmitting current set corresponding to the M1 group of transmitting electrodes in poor ground coupling state as training sets, and selecting the corresponding receiving electrode transmitting current set corresponding to the N0 group of transmitting electrodes in good ground coupling state and the corresponding receiving electrode transmitting current set corresponding to the N1 group of transmitting electrodes in poor ground coupling state as test sets;

[0008] Wherein, M0+N0=M, M1+N1=NM; any set of receiving electrodes corresponding to the transmitting current set includes the transmitting currents corresponding to n pairs of receiving electrodes;

[0009] Step 2: Calculate the good contact mean vector μ0 and the bad contact mean vector μ1 for the corresponding transmitting current set of the receiving electrodes in the M0 group of transmitting electrodes with good ground coupling and the corresponding transmitting current set of the receiving electrodes in the M1 group of transmitting electrodes with poor ground coupling in the training set, and calculate the intra-class divergence matrix S w ; Calculate the mean vector μ for all the receiving electrodes corresponding to the transmitting current set in the training set, and calculate the inter-class divergence matrix S b ;

[0010] Step 3: Set the criterion function as shown in formula (3) and calculate when J weighted The projection vector corresponding to the maximum value of (w) is the optimal projection vector w;

[0011] J weighted(w) =λ(w T S b w)-(1-λ)(w T S w w) (3)

[0012] Among them, λ represents the weight factor, λ∈[0,1];

[0013] Step 4: project the transmitting current x corresponding to the receiving electrode whose coupling state is to be determined onto the optimal projection vector w through equation (4) to obtain a one-dimensional projection value z;

[0014] z=w T ×x (4)

[0015] If z is greater than the discrimination threshold of 0.5, it is predicted that the transmitting electrode is poorly coupled to the ground; if z≤0.5, it is predicted that the transmitting electrode is well coupled to the ground.

[0016] Preferably, λ in step 3=0.6.

[0017] Preferably, in step 2, the intra-class scatter matrix S is calculated by formula (1): w ;

[0018]

[0019] Among them, x i represents the characteristic vector of the transmitting current set corresponding to the i-th group of receiving electrodes; y i Represents the category label of the i-th group of data. When y i = 0 means that the transmitting electrode ground coupling is good. i =1 indicates that the transmitting electrode ground coupling is poor.

[0020] Preferably, in step 2, the inter-class scatter matrix S is calculated by formula (2): b ;

[0021] S b =M0(μ0-μ)(μ0-μ) T +M1(μ1-μ)(μ1-μ) T (2).

[0022] Compared with the prior art, the advantages of the present invention are:

[0023] The present application discloses a method for identifying transmitting electrodes with poor grounding in mine time-lapse resistivity monitoring, which automates the entire process from data collection, model construction to real-time monitoring, significantly improves the accuracy and automation level of electrode grounding status identification, realizes scientific, accurate and real-time identification of electrode grounding status, and significantly improves the efficiency of mine safety monitoring systems. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the following specific embodiments, they are used to explain the present invention but do not constitute a limitation of the present invention. In the accompanying drawings:

[0025] Figure 1 Verification accuracy for different λ values. DETAILED DESCRIPTION

[0026] The invention is not limited to the following specific embodiments, and all equivalent changes made on the basis of the technical solution of this application fall within the protection scope of the invention.

[0027] All components and devices in the present invention, unless otherwise specified, are components and devices known in the prior art.

[0028] Example

[0029] A method for identifying a poorly grounded transmitting electrode in mine time-lapse resistivity monitoring comprises the following steps:

[0030] Step 1, collecting N groups of receiving electrodes corresponding to transmitting current sets, including M groups of transmitting electrodes in good ground coupling state and NM groups of transmitting electrodes in poor ground coupling state, selecting the corresponding receiving electrode transmitting current set corresponding to the M0 group of transmitting electrodes in good ground coupling state and the corresponding receiving electrode transmitting current set corresponding to the M1 group of transmitting electrodes in poor ground coupling state as training sets, and selecting the corresponding receiving electrode transmitting current set corresponding to the N0 group of transmitting electrodes in good ground coupling state and the corresponding receiving electrode transmitting current set corresponding to the N1 group of transmitting electrodes in poor ground coupling state as test sets;

[0031] Among them, M0+N0=M, M1+N1=NM; the transmitting current set corresponding to any group of receiving electrodes includes the transmitting currents corresponding to n pairs of receiving electrodes.

[0032] This embodiment uses a monopole-dipole device to monitor the resistivity of a single tunnel mine. Before carrying out the time-shift resistivity monitoring project, data supplementation is purposefully collected. A reasonable resistivity monitoring experiment is designed to simulate different grounding coupling states and record the corresponding transmission current change data of the receiving electrode.

[0033] Before the start of the mine resistivity monitoring project, 20 groups of receiving electrodes corresponding to the transmitting current set were measured, of which 10 groups of transmitting electrodes had good ground coupling and 10 groups of transmitting electrodes had poor ground coupling. The collected data are shown in Table 1. In Table 1, the transmitting electrode ground coupling state of 0 indicates good ground coupling (marked as category 0), and the ground coupling state of 1 indicates poor ground coupling (marked as category 1); 8 groups of transmitting electrode ground coupling good data and 8 groups of transmitting electrode ground coupling poor data were selected as training sets, and the remaining 2 groups of transmitting electrode ground coupling good data and 2 groups of transmitting electrode ground coupling poor data were used as verification data sets.

[0034] Table 1 Collected emission electrode current data and its coupling state

[0035]

[0036]

[0037] Step 2: Calculate the good contact mean vector μ0 and the bad contact mean vector μ1 for the 8 sets of receiving electrodes corresponding to the good ground coupling state and the 8 sets of receiving electrodes corresponding to the bad ground coupling state in the training set, and calculate the intra-class divergence matrix S by formula (1): w ;

[0038]

[0039] Among them, x iThe characteristic vector representing the transmitting current set corresponding to the i-th group of receiving electrodes;

[0040] y i Represents the category label of the i-th group of data. When y i = 0 means that the transmitting electrode ground coupling is good. i =1 indicates that the transmitting electrode ground coupling is poor.

[0041] The mean vector μ is calculated for all the transmitting current sets corresponding to the receiving electrodes in the training set, and the inter-class divergence matrix Sb is calculated by formula (2);

[0042] S b =M0(μ0-μ)(μ0-μ) T +M1(μ1-μ)(μ1-μ) T (2)

[0043] Step 3: Set the criterion function as shown in formula (3) and calculate when J weighted The projection vector corresponding to the maximum value of (w) is the optimal projection vector w;

[0044] J weighted(w) =λ(w T S b w)-(1-λ)(w T S w w) (3)

[0045] Among them, λ represents the weight factor, λ∈[0,1]. When λ tends to 1, it emphasizes the separation between classes; when λ tends to 0, it emphasizes the compactness within the class. The λ value range is [0.1, 0.2, ..., 0.9]. Combined with the validation data set, the optimal λ value is determined. For each λ value, the weighted discriminant analysis function is performed on the training set to calculate J weighted (w) and obtain the corresponding projection vector w. Using the obtained w, the verification data is projected into the new feature space and the classification accuracy is calculated. The accuracy of verification for different λ values ​​is as follows Figure 1 shown. Figure 1 It can be seen that when λ=0.6, all four sets of verification data are correct. Therefore, in the subsequent test data of this embodiment, λ=0.6 is used for discriminant analysis.

[0046] Step 4: project the transmitting current x corresponding to the receiving electrode of the coupling state to be determined onto the optimal projection vector w through equation (4) to obtain a one-dimensional projection value z;

[0047] z=w T ×x (4)

[0048] If z is greater than the discrimination threshold of 0.5, it is predicted that the transmitting electrode ground coupling is poor;

[0049] If z≤0.5, it is predicted that the transmitting electrode is well coupled to the ground.

[0050] In the time-lapse resistivity monitoring project of this embodiment, the current data of the two transmitting electrodes collected in real time are shown in Table 2.

[0051] Table 2 Transmitting electrode current data to be determined for the transmitting electrode state

[0052]

[0053] S3.2: Contact coupling state determination: Substitute the real-time current data in Table 2 into formula (4), and the projection value of sample number 1 to be determined is 0.375, which is predicted to be good contact. The projection value of sample number 2 to be determined is 0.64, which is predicted to be poor contact.

[0054] As the monitoring project progresses and new data accumulates, the discriminant function is retrained or updated regularly every month to adapt to changes in geological, environmental and other factors. The performance of the discriminant model on new data is continuously monitored. If performance degradation is found, the data set should be updated or the discriminant method should be changed in a timely manner. When the discrimination result shows that there is a poorly grounded electrode, the system immediately generates an early warning message and notifies relevant personnel through SMS, email, APP push, etc.

[0055] The method for identifying poorly grounded transmitting electrodes in mine time-lapse resistivity monitoring disclosed in this embodiment realizes the automation of the entire process from data collection, model building to real-time monitoring, and significantly improves the accuracy and automation level of electrode grounding status identification.

[0056] The preferred embodiments of the present disclosure are described in detail above in conjunction with the accompanying drawings; however, the present disclosure is not limited to the specific details in the above embodiments. Within the technical concept of the present disclosure, a variety of simple modifications can be made to the technical solution of the present disclosure, and these simple modifications all fall within the protection scope of the present disclosure.

[0057] It should also be noted that the various specific technical features described in the above specific embodiments can be combined in any suitable manner without contradiction. In order to avoid unnecessary repetition, the present disclosure will not further describe various possible combinations.

[0058] In addition, the various different implementation modes disclosed in this solution may be arbitrarily combined, and as long as they do not violate the concept of the present disclosure, they should also be regarded as the content invented by the present disclosure.

Claims

1. A method for identifying poorly grounded transmitting electrodes in mine time-lapse resistivity monitoring, characterized in that: The steps include: Step 1, collecting N groups of receiving electrodes corresponding to transmitting current sets, including M groups of transmitting electrodes in good ground coupling state and NM groups of transmitting electrodes in poor ground coupling state, selecting the corresponding receiving electrode transmitting current set corresponding to the M0 group of transmitting electrodes in good ground coupling state and the corresponding receiving electrode transmitting current set corresponding to the M1 group of transmitting electrodes in poor ground coupling state as training sets, and selecting the corresponding receiving electrode transmitting current set corresponding to the N0 group of transmitting electrodes in good ground coupling state and the corresponding receiving electrode transmitting current set corresponding to the N1 group of transmitting electrodes in poor ground coupling state as test sets; Wherein, M0+N0=M, M1+N1=NM; any set of receiving electrodes corresponding to the transmitting current set includes the transmitting currents corresponding to n pairs of receiving electrodes; Step 2: Calculate the good contact mean vector μ0 and the bad contact mean vector μ1 for the corresponding transmitting current set of the receiving electrodes in the M0 group of transmitting electrodes with good ground coupling and the corresponding transmitting current set of the receiving electrodes in the M1 group of transmitting electrodes with poor ground coupling in the training set, and calculate the intra-class divergence matrix S w ; The mean vector μ is calculated for all the receiving electrodes corresponding to the transmitting current set in the training set, and the inter-class divergence matrix S is calculated. b ; Step 3: Set the criterion function as shown in formula (3) and calculate when J weighted The projection vector corresponding to the maximum value of (w) is the optimal projection vector w; J weighted(w) =λ(w T S b w)-(1-λ)(w T S w w) (3) Among them, λ represents the weight factor, λ∈[0,1]; Step 4: project the transmitting current x corresponding to the receiving electrode whose coupling state is to be determined onto the optimal projection vector w through equation (4) to obtain a one-dimensional projection value z; from=in T ×x (4) If z is greater than the discrimination threshold of 0.5, it is predicted that the transmitting electrode ground coupling is poor; If z≤0.5, it is predicted that the transmitting electrode is well coupled to the ground.

2. The method for identifying poorly grounded transmitting electrodes in mine time-lapse resistivity monitoring according to claim 1, characterized in that: In step 3, λ=0.

6.

3. The method for identifying poorly grounded transmitting electrodes in mine time-lapse resistivity monitoring according to claim 1, characterized in that: In step 2, the intra-class scatter matrix S is calculated by formula (1): w ; Among them, x i The characteristic vector representing the transmitting current set corresponding to the i-th group of receiving electrodes; y i Represents the category label of the i-th group of data. When y i = 0 means that the transmitting electrode ground coupling is good. i =1 indicates that the transmitting electrode ground coupling is poor.

4. The method for identifying poorly grounded transmitting electrodes in mine time-lapse resistivity monitoring according to claim 1, characterized in that: In step 2, the inter-class scatter matrix S is calculated by formula (2): b ; S b =M0(μ0-μ)(μ0-μ) T +M1(μ1-μ)(μ1-μ) T (2)。

Images