Grounding grid non-structural weakness risk identification method

Abstract

The application discloses a grounding grid non-structure weak point risk identification method, comprising the following steps: collecting grounding grid basic data; establishing a unified three-dimensional coordinate system; dividing the grounding grid branch into multiple candidate units; building a defect-free reference working condition simulation model, calculating the structure background influence rate of each candidate unit; calculating the structure node distance, structure influence threshold and distance threshold, and screening out non-structure candidate units in combination with the structure background influence rate; constructing multiple working conditions, extracting electrical response parameters, and calculating difference characteristics; according to the difference characteristics, respectively calculating the risk probabilities of three types of risks, namely, point corrosion, breakpoint / high resistance and abnormal conductor disturbance, dividing the risk grades, and completing the risk identification. The method solves the problems that the prior art is difficult to identify non-structure position defects such as ordinary straight line segments of the grounding grid, is easily disturbed by external conductive bodies, and the model cannot adapt to different station working conditions.

Description

Technical Field

[0001] The present invention belongs to the technical field of state assessment of grounding grids in power systems, and relates to an unstructured weakness risk identification method for candidate units of a grounding grid. Background Art

[0002] A grounding grid is a core infrastructure for ensuring the safety of equipment and personnel in substations, power plants, converter stations and industrial stations. It is mainly composed of buried conductors, grounding down conductors, welding spots, connection points and boundary branches, and can provide a reliable current discharge channel under working conditions such as fault current, lightning current and equipment discharge.

[0003] There are obvious technical shortcomings in engineering applications: First, the qualified overall grounding resistance of a grounding grid cannot reflect local pitting, cross-section reduction, high-resistance defects or strand breakage of ordinary straight segments. Such local defects only change the current-carrying capacity of a single conductor and the surface potential distribution, which are difficult to detect through overall indicators, but can cause safety hazards such as excessive step voltage, excessive contact voltage and conductor thermal stability failure. Second, traditional detection methods focus on inspection for structural weaknesses such as welding spots, connection points, corners and grounding down conductors, and generally ignore ordinary straight segments far away from structural nodes; such areas are easily affected by soil corrosion, intersection of multiple current discharge sources and disturbance of external abnormal conductors, forming unstructured weaknesses. Third, a large number of conductive objects such as construction leftover metal sheets, metal pipelines, building steel bars and cable metal sheaths on site will change the local current discharge path of the grounding grid, causing abnormal deviation of current density, surface potential and magnetic field response. Such disturbances are easily determined as measurement noise, which further increases the difficulty of identifying unstructured weaknesses. Fourth, there are great differences in soil resistivity, moisture content, corrosion degree, grounding grid topology and buried conditions among different stations. Most of the existing technologies adopt fixed thresholds and general models, which are prone to produce model migration deviations; meanwhile, there is currently a lack of an integrated closed-loop identification system for fine division of candidate units, unstructured weakness screening, disturbance differential analysis, risk probability output and on-site feedback correction, resulting in insufficient identification accuracy and practicability. Summary of the Invention

[0004] The purpose of the present invention is to provide an unstructured weakness risk identification method for a grounding grid, which solves the problems that the prior art is difficult to identify defects at unstructured positions such as ordinary straight segments of the grounding grid, is easily interfered by external conductor disturbance, and the model cannot adapt to the working conditions of different stations.

[0005] The technical scheme adopted by the present invention is an unstructured weakness risk identification method for a grounding grid, which is specifically implemented according to the following steps: Step 1, collect basic data of the grounding grid; Step 2, establish a unified three-dimensional coordinate system; Step 3, divide the grounding grid branches into a plurality of candidate units; Step 4: Build a flawless baseline simulation model and calculate the structural background influence rate of each candidate unit; Step 5: Calculate the structural node distance, structural influence threshold, and distance threshold, and filter out non-structural candidate units based on the structural background influence rate; Step 6: Construct multiple operating conditions and extract electrical response parameters, and calculate differential characteristics; Step 7: Calculate the probability of three types of risks—pitting corrosion, breakpoint / high resistance, and abnormal conductor disturbance—based on the differential characteristics, classify the risk levels, and complete the risk identification.

[0006] The invention is further characterized by: Step 1 is as follows: Collect the three-dimensional coordinates of grounding grid nodes, branch start and end relationships, conductor material and cross-sectional dimensions, burial depth, structural node locations, soil resistivity of the site area, and data on known external conductors, and store the data in a standardized manner according to five types of tables: node table, branch table, structural node table, soil parameter table, and abnormal conductor table.

[0007] Step 2 is as follows: A global three-dimensional coordinate system is established with the geometric center of the grounding grid as the origin. The X-axis corresponds to the east-west direction of the site, the Y-axis corresponds to the north-south direction of the site, and the Z-axis is the vertical direction.

[0008] Step 3 specifically involves: The grounding grid branches are divided into several candidate units according to a fixed length. Each candidate unit is assigned a unit number, start and end coordinates, center coordinates, unit length, conductor cross-section, whether it is a pure straight line segment, and corresponding ground surface measurement point number information to form a candidate unit information table.

[0009] Step 4 specifically involves: A flawless baseline simulation was built, and the current density J and surface potential gradient G parameters of each candidate element under this condition were extracted. The structural influence ratio and structural background influence rate A were then calculated. i ; Among them, the structural influence ratio includes the current density influence ratio R. J The influence ratio of surface potential gradient to R G The calculation formula is as follows:

[0010] In the formula, X i0 X represents the current density or surface potential gradient response of candidate cell i under a defect-free baseline condition; node.peak Indicates the peak response near the nearest structural node; X far ε represents the average background response of a normal straight line segment far from structural nodes; ε is a small, stable quantity to prevent the denominator from returning to zero. Structural background influence rate A i=max(R J ,R G ).

[0011] Step 5 specifically involves: Regarding the structural background influence rate A obtained in step 4 i For candidate cells with a value <0.30, calculate the distance d from the center point of each candidate cell to the solder joint, connection point, corner point, and grounding lead connection point. 焊点 d 连接点 d 边角点 d 引下线 The minimum value is taken as the distance d between structural nodes. i ,Right now:

[0012] Plot the structural background influence rate A of each candidate unit i With distance d i The decay curve is used to calculate the rate of change of influence of adjacent candidate units. When two consecutive adjacent intervals of ΔA i When all values ​​are ≤0.03 to 0.05, the influence decay is determined to have transitioned from a rapid and steep decline phase to a gradual low-influence range. The structural background influence rate corresponding to the starting position of this gradual low-influence range is extracted as the structural influence threshold A. th and A th The corresponding d i As the distance threshold D th ; If the structural node of a candidate unit is d i ≥D th And the structural background influence rate A i ≤A th If so, the candidate element is determined to be an unstructured candidate element; for distance d i ≥D th Impact rate A i Elements in the 0.15~0.30 transition range are marked as elements to be reviewed. By densifying the surface measurement points or conducting continuity retests, it is confirmed whether the element can be included in the unstructured candidate element set.

[0013] Step 6 specifically involves: For the unstructured candidate units selected in step 5, three simulation conditions were built: pitting corrosion, breakpoint / high resistance, and abnormal conductor disturbance. Electrical response data were collected, and relative differential features ΔJ, ΔG, current continuity anomaly C, and hot spot migration distance M were calculated with reference to the baseline condition. All relative differential features were normalized to the interval of 0 to 1 to obtain the normalized feature value N. J N G N C N M .

[0014] Step 7 specifically includes: N J N G N C and N M Substituting the values ​​into the weighting formula, the pitting risk P of each unstructured candidate unit is calculated sequentially. pit Breakpoint / High Resistance Risk P break Risk of abnormal conductor disturbance P dist The maximum value is taken to obtain the overall detection priority: Score = max(P) pit P break P dist ); Among them, P pit =a1N J +a2N G +a3(1-N C )+a4(1-N M ); In the formula, a1+a2+a3+a4=1; a1 takes values ​​of 0.30~0.50, a2 takes values ​​of 0.25~0.45, a3 takes values ​​of 0.05~0.20, and a4 takes values ​​of 0.05~0.15; P break =b1N C +b2N G +b3N J +b4N M ; In the formula, b1+b2+b3+b4=1; b1 takes values ​​of 0.35~0.60, b2 takes values ​​of 0.15~0.35, b3 takes values ​​of 0.10~0.30, and b4 takes values ​​of 0.05~0.20; P dist =c1N M +c2N G +c3N J +c4(1-N C ); In the formula, c1+c2+c3+c4=1; c1 takes values ​​of 0.35~0.60, c2 takes values ​​of 0.15~0.35, c3 takes values ​​of 0.10~0.30, and c4 takes values ​​of 0.05~0.20; After calculating the score, perform on-site operations according to the threshold levels: (1) Score ≥ 0.75: First priority, conduct continuity retesting first; (2) 0.55≤Score<0.75: Secondary priority, densify surface measuring points, and remeasure soil parameters and electrical response; (3) 0.35≤Score<0.55: Level 3 priority, included in the daily key inspection log, and tracked regularly; (4) Score < 0.35: No abnormal risks were found. Perform routine inspections as usual.

[0015] The method of the present invention also includes a field feedback closed-loop update mechanism: based on the results of field excavation verification, continuity testing, soil retesting, and inspection misjudgment and omission, the structural influence threshold, distance threshold, risk calculation weight, soil parameters, and abnormal conductor disturbance judgment rules are corrected to achieve localized iterative optimization of the model.

[0016] The beneficial effects of this invention are: (1) The method of the present invention performs a fine division of candidate units for grounding grid branches and completes the screening by combining the structural background influence rate and adaptive distance threshold. It can accurately identify non-structural weaknesses of ordinary straight segments far away from weld points, connection points, corners and grounding leads. It solves the problem that traditional technology relies on overall grounding resistance and overall corrosion evaluation, which cannot locate local pitting corrosion, high resistance and open circuit break points, and easily misses defects in non-structural areas. (2) Constructing multi-condition differential features to quantitatively describe the changes in leakage paths caused by abnormal conductors and multiple leakage sources, making up for the lack of candidate unit-level feature representation in the existing technology, effectively distinguishing three types of anomalies: pitting corrosion, conductor breakage and external conductor disturbance, and improving the interpretability of risk identification results. (3) The risk probabilities of three categories are classified and output: pitting corrosion, breakpoint / high resistance, and abnormal conductor disturbance. The single and general corrosion evaluation method is abandoned, and the identification results are more comprehensive. At the same time, the detection priority and on-site handling rules are provided to directly guide on-site operations such as retesting, excavation, and inspection, which is highly practical for engineering. (4) Add a closed-loop update mechanism for on-site feedback. Based on on-site measurements, excavation verification and other data, continuously correct the model parameters and iterate the general model into a localized model that adapts to different soil environments, topological structures and external conductor distributions of different sites. This effectively reduces model migration deviation and reduces misjudgment and omissions in the identification process. Attached Figure Description

[0017] Figure 1 This is a schematic diagram of the grounding grid candidate unit division in Embodiment 1 of the present invention; Figure 2 This is a current density distribution diagram of each candidate unit in Embodiment 1 of the present invention under a reference operating condition; Figure 3 This is a map showing the surface potential gradient distribution of each candidate unit under the baseline working condition in Embodiment 1 of the present invention; Figure 4 This is a graph showing the change in current continuity of candidate cell S03 under high resistance / breakpoint conditions in Embodiment 1 of the present invention. Figure 5 This is a differential feature diagram of the pitting, high resistance, and abnormal conductor disturbance conditions of candidate unit S03 in Embodiment 1 of the present invention relative to the reference condition. Figure 6 This is a schematic diagram of the differential hotspot migration caused by the proximity of an abnormal conductor to the candidate unit S03 in Embodiment 1 of the present invention. Detailed Implementation

[0018] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments.

[0019] The method for identifying non-structural weaknesses in grounding grids according to the present invention is implemented according to the following steps: Step 1: Collect and organize basic grounding grid data For the target substation grounding grid, collect the three-dimensional coordinates of each node, branch start and end relationships, conductor material, cross-sectional dimensions, and burial depth; identify the locations of structural nodes such as weld points, connection points, corner points, and grounding down conductors. Simultaneously collect soil resistivity data from different areas of the site; investigate and register external conductors such as metal pipelines, building steel bars, and cable metal sheaths in the site.

[0020] As shown in Table 1, data is stored in a standardized manner according to five types of tables: node table, branch table, structural node table, soil parameter table, and abnormal conductor table. For abnormal conductors whose specific locations cannot be determined, no preliminary data entry is performed; instead, they are identified by reverse engineering based on subsequent electrical response and hotspot migration characteristics.

[0021] Table 1 Basic Data Table

[0022] Step 2: Establish a unified three-dimensional coordinate system A global three-dimensional coordinate system is established with the geometric center of the grounding grid as the origin: the X-axis corresponds to the east-west direction of the site (horizontal direction on the drawing), the Y-axis corresponds to the north-south direction of the site (vertical direction on the drawing), and the Z-axis is the vertical direction; the ground surface plane coordinates are defined as Z=0, and the burial depth of the grounding grid is uniformly set to Z=-0.8.

[0023] Step 3, Divide the candidate units for the grounding grid. Each branch of the grounding grid is divided into several candidate units, which can be divided according to fixed lengths of 0.5m, 1m, and 2m or the spacing between field measurement points. The candidate unit is used as the smallest analysis unit. The unit number, start and end coordinates, center coordinates, unit length, conductor cross-section, whether it is a pure straight line segment, and corresponding ground measurement point number of each candidate unit are recorded one by one to form a candidate unit information table as shown in Table 2, so as to ensure that all electrical measurement data can be accurately assigned to a single unit.

[0024] Table 2 Candidate Unit Information Table

[0025] Step 4: Establish a defect-free baseline working condition and calculate the structural background influence rate. A defect-free baseline simulation model was built using COMSOL simulation software. The model included the grounding grid conductor, layered soil, and grounding down conductor, without adding any defects or external abnormal conductors. A standard test current was applied, and the simulation extracted two core parameters for each candidate unit: the current density J and the surface potential gradient G.

[0026] Calculate the current density influence ratio R according to the structural influence ratio formula. J The influence ratio of surface potential gradient to R G Subsequently, to avoid a situation where a certain type of response is still affected by structural nodes but weakened by the average value, the method of this invention uses the conservative maximum value method to determine the comprehensive structural background influence rate A. i A i =max(R J ,R G When R J Or R G If any one of the indicators is high, the candidate element is considered to still be affected by the background field of the structural nodes; only when R is high... J and R G Only when both are low is the candidate unit considered to be closer to the background response of a normal straight line segment. The formula for calculating the structural influence ratio is:

[0027] In the formula, X i0 X represents the response of candidate unit i under a defect-free baseline condition; node.peak Indicates the peak response near the nearest structural node; X far ε represents the average background response of a normal straight line segment far from the structural nodes; ε is a small, stable quantity to prevent the denominator from returning to zero.

[0028] Based on Table 3, complete the preliminary classification and labeling of unstructured candidate units.

[0029] Table 3 Preliminary classification criteria for unstructured candidate units

[0030] Step 5: Select unstructured candidate units Regarding step 4, A i For candidate elements with a distance <0.30, calculate the distance from the center point of each candidate element to the weld point, connection point, corner point, and grounding lead connection point, and take the minimum value as the structural node distance d. i ,Right now:

[0031] Next, plot the structural background influence rate A. i With distance d i The decay curve is used to calculate the rate of change of influence ΔA between adjacent candidate units. i :

[0032] Identify the location where the structural influence transitions from a significant attenuation to a region of gradual, low influence, when ΔA i When two consecutive adjacent intervals are no greater than 0.03 to 0.05 (any value within this range can be fixed as the critical limit as needed), the change of Ai within this interval can be considered to be relatively flat. The interval where Ai has dropped to a low level and subsequent changes meet the flattening criterion is defined as the low-influence zone, and the upper limit value A of this low-influence zone is set as follows. i As the structural influence threshold A th Determine A th After that, A will be satisfied for the first time. i ≤A th Furthermore, the distance d corresponding to the candidate elements that continuously satisfy the conditions of low impact and gradual change is... i , as the distance threshold D th。

[0033] Subsequently, according to d i A i A th and D th Unstructured candidate elements that are ordinary straight line segments are selected based on the following selection rule: the distance d from the candidate element to the nearest structural node. i ≥D th And the structural background influence rate A i ≤A th Those that meet this condition are considered unstructured candidate units.

[0034] In addition, distance d i Meets the standard but has an impact rate of A i Elements falling within the 0.15–0.30 transition range are not directly classified as unstructured candidate elements, but are instead marked as elements to be reviewed. The influence of structural nodes on these elements may have diminished, but has not yet stabilized at the background level of ordinary straight lines. In practical engineering applications, on-site verification methods such as denser surface measurement point re-measurement and continuity re-measurement should be used to confirm whether the element can be included in the unstructured candidate element set.

[0035] Step 6: Multi-condition response acquisition and differential feature construction For the selected unstructured candidate units, three simulation conditions were established: pitting corrosion, breakpoint / high resistance, and anomalous conductor disturbance. The average current density J, surface potential gradient G, branch current continuity, and hotspot coordinates of each unit under each condition were obtained by combining field measurements. In implementations with magnetic field simulation or on-site magnetic response measurement capabilities, magnetic response data B can also be obtained.

[0036] Based on the defect-free baseline working condition, using the formula Calculate the relative difference characteristics of each variable, including ΔJ, ΔG, and ΔB, and calculate the current continuity anomaly C and hotspot migration distance M; linearly normalize all difference indices to the 0~1 interval to obtain the normalized characteristic value N of each variable. J N G N C N M and N B Among them, N J N G N C and N M As an essential feature for risk level calculation, the magnetic response difference ΔB and its normalized value N are... B As an optional enhancement feature, it can be used to assist in identifying flow around discontinuities and abnormal conductor disturbances when employing magnetic field simulation or in-situ magnetic response measurement. Table 4 shows the applications of each differential feature.

[0037] Table 4. Applications of Differential Features

[0038] Step 7: Risk Calculation and Detection Priority Determination N J N G N C and N M Substituting the values ​​into the weighting formula, the pitting risk P of each unstructured candidate unit is calculated sequentially. pit Breakpoint / High Resistance Risk P break Risk of abnormal conductor disturbance P dist The maximum value is taken to obtain the overall detection priority: Score = max(P) pit P break P dist ).

[0039] P pit =a1N J +a2N G +a3(1-N C )+a4(1-N M ); Where a1+a2+a3+a4=1; a1 takes values ​​of 0.30~0.50, a2 takes values ​​of 0.25~0.45, a3 takes values ​​of 0.05~0.20, and a4 takes values ​​of 0.05~0.15.

[0040] P break =b1N C +b2N G +b3N J +b4N M ; Where b1+b2+b3+b4=1; b1 takes values ​​of 0.35 to 0.60, b2 takes values ​​of 0.15 to 0.35, b3 takes values ​​of 0.10 to 0.30, and b4 takes values ​​of 0.05 to 0.20.

[0041] P dist =c1N M +c2N G +c3N J +c4(1-N C ).

[0042] Where c1+c2+c3+c4=1; c1 takes values ​​of 0.35~0.60, c2 takes values ​​of 0.15~0.35, c3 takes values ​​of 0.10~0.30, and c4 takes values ​​of 0.05~0.20.

[0043] In actual engineering projects, adjustments can be made within the above range based on simulation samples, continuity retesting, excavation verification, and abnormal conductor confirmation results.

[0044] Pitting corrosion risk is mainly manifested as an increase in local current density and surface potential gradient, but the current path is not significantly interrupted. Therefore, N is increased in the pitting corrosion risk calculation. J and N G The weights are reduced, while N is decreased. C and N M The impact of discontinuity / high resistance risk is mainly manifested in the disruption of current continuity, sudden drop in branch current, and potential abrupt change. Therefore, increasing N in the calculation of discontinuity / high resistance risk is crucial. C The weight of N is determined by the risk of anomalous conductor disturbance. The main manifestations of anomalous conductor disturbance risk are hotspot location migration and response distribution shift towards external conductors. The continuity of the candidate cell's current may not be significantly disrupted. Therefore, in the calculation of anomalous conductor disturbance risk, the weight of N is increased. M The weights, combined with N J and N G Make a judgment.

[0045] After calculating the score, perform on-site operations according to the threshold levels: (1) Score≥0.75: First priority, conduct continuity retesting first, and excavate to check conductor condition if necessary; (2) 0.55≤Score<0.75: Secondary priority, densify surface measuring points, and remeasure soil parameters and electrical response; (3) 0.35≤Score<0.55: Level 3 priority, included in the daily key inspection log, and tracked regularly; (4) Score < 0.35: No abnormal risks were found. Perform routine inspections as usual.

[0046] Step 8: On-site feedback and closed-loop update Based on the results of on-site excavation verification, continuity testing, corrosion detection, soil retesting, and cases of misjudgment and omission discovered during inspections, the model parameters were corrected in reverse: the structural influence threshold A was adjusted. th Distance threshold D th The calculation weights for the three types of risks were optimized, and the soil zoning parameters and abnormal conductor disturbance judgment rules were updated. The updated rules are directly applied to the next round of grounding grid risk identification, achieving localized adaptive optimization of the model. Specific update actions can be found in Table 5.

[0047] Table 5. Comparison of Feedback Status and Update Actions

[0048] This invention achieves accurate identification of non-structural weaknesses such as ordinary straight sections of the grounding grid through refined unit division, structural influence elimination, multi-condition differential analysis, risk output by type and closed-loop update. It is adaptable to the complex field environment of different sites and effectively makes up for the shortcomings of traditional detection technology.

[0049] Example 1: This embodiment presents a three-dimensional Electric Currents model that can be reproduced in COMSOL 6.3; it fixes candidate elements S01-S06 and verifies that S03 and S04 meet the screening conditions for unstructured candidate elements, and selects S03 as the key defect implantation element; it constructs three types of anomalies: pitting, high resistance / breakpoint, and abnormal conductor disturbance; and outputs the corresponding CSV table and corresponding image.

[0050] (1) Overall COMSOL modeling process Four independent models were created using COMSOL, corresponding to four operating conditions. Each operating condition was studied using ElectricCurrents physics and Stationary steady-state methods. The four models were saved as baseline.mph, pitting.mph, breakpoint / high resistance.mph, and anomalous disturbance.mph, respectively. Post-processing was used to read the potential V, current density J, and electric field strength E and generate tables and images.

[0051] (2) Geometric model, material parameters and candidate element division like Figure 1 As shown, the soil computational domain is set as a cuboid with dimensions of 18m × 18m × 5m, and the coordinate range is X = -9~9m, Y = -9~9m, Z = -5~0m. The grounding grid is buried at a depth of Z = -0.8m and is set as a 3×3 grid: the horizontal conductors are located at Y = -6, 0, 6m, and the vertical conductors are located at X = -6, 0, 6m, with an equivalent conductor radius of 0.025m. The center down conductor is located at X = 0, Y = 0 and is vertically arranged along Z = -0.8~0m.

[0052] The intermediate horizontal branch with Y=0, Z=-0.8, and X=0~6m is divided into six 1m candidate units, S01-S06. S01 and S06 are close to structural nodes or downleaders and are easily affected by the structural background; S03 and S04 are located on ordinary straight lines and are suitable for non-structural weakness identification.

[0053] The specific parameters are shown in Table 6.

[0054] Table 6 Model Parameters

[0055] (3) Physics, boundary conditions and solution settings This embodiment uses the Electric Currents interface to describe the electrical conductivity process between the soil and the metal conductor. The top end face of the center lead is set to Terminal, the terminal type is Current, and the injected current is 100A; the bottom surface of the soil at Z=-5m is set to Ground, and the potential is 0V; the remaining outer boundaries maintain the default electrical insulation conditions of Electric Currents.

[0056] Each operating condition was studied using a Stationary steady-state approach. The mesh was set to COMSOL automatic meshing level 7 to ensure stable solutions for all four operating conditions within limited memory. The settings for the four simulation operating conditions are shown in Table 7.

[0057] Table 7 Specific settings for the four simulation conditions

[0058] (4) Structural background exclusion Structural background exclusion uses distance index d i Compared with background intensity index A i Joint judgment. In this embodiment, the threshold is set to D. th =2.0m, A th =0.15; Take A i =max(RJ,RG). Table 8 shows the simulation results. It can be seen that the d of S03 i =2.5m, A i =0.001, d of S04 i =2.5m, A i =0.141, all satisfying d i ≥D th And A i ≤A th Therefore, S03 and S04 are both unstructured candidate units.

[0059] Table 8 Simulation Results

[0060] (5) Simulation field quantity extraction and difference index This step uses candidate element S03 as an example, reading the potential V, current density J, and electric field intensity E from each model, where E is used as an approximation of the surface potential gradient G. For element S03, the cross-condition differential characteristics ΔJ, ΔG, continuity characteristic C, and hotspot migration characteristic M are calculated. For example... Figure 2 and Figure 3 The figures show the current density distribution and surface potential gradient distribution of each candidate unit under the baseline operating conditions.

[0061] C is calculated using the local current density surrogate from the left, middle, and right sides of S03, reflecting the left-right imbalance caused by breakpoints or high resistance, as well as the sudden drop in the middle. M represents the normalized migration distance of the cross-unit differential hotspot relative to S03, used to identify abnormal conductor disturbances. Specific calculation results are shown in Table 9. Figure 4 , Figure 5 , Figure 6 As shown, under the pitting corrosion condition, the current density difference ΔJ in S03 increases significantly, but the continuity anomaly C is small, and the hot spot is still located in S03. This indicates that the reduction in local cross-section mainly causes changes in local current carrying capacity and has not yet caused obvious discontinuity damage. Under the high resistance / discontinuity condition, the continuity anomaly C increases significantly, indicating that there is a significant imbalance in the current carrying state on the left, right and middle parts of S03, which corresponds to the high resistance or discontinuity setting. Under the abnormal conductor disturbance condition, the increases in ΔJ and ΔG are small, but the differential hot spot migrates from S03 to S05, and the migration distance M of the hot spot increases significantly. This indicates that the anomaly is mainly manifested as the displacement of the discharge path or field distribution, rather than pitting corrosion or breakage of the grounding grid conductor itself.

[0062] Table 9 Calculation Results

[0063] (6) Risk scoring methods and results Normalize ΔJ, ΔG, C, and M for abnormal operating conditions to obtain N. J N G N C and N M The three risk scores are calculated as follows: P pit =0.40N J +0.35N G +0.15(1-N C )+0.10(1-N M ); P break =0.45N C +0.25N G +0.20N J +0.10N M ; Pdist =0.45N M +0.25N G +0.20N J +0.10(1-N C ).

[0064] The final score is the maximum score across the three categories.

[0065] The risk level thresholds are as follows: Score ≥ 0.75 is considered high risk; 0.55 ≤ Score < 0.75 is considered medium-high risk; 0.35 ≤ Score < 0.55 is considered medium risk; and Score < 0.35 is considered low risk.

[0066] The calculation results and risk level assessments are shown in Table 10.

[0067] Table 10 Calculation Results and Risk Levels

[0068] In this embodiment, S03 pitting corrosion, S03 high resistance / breakpoint, and adjacent anomalous conductor disturbance conditions were constructed in COMSOL. The extracted current density difference, surface potential gradient difference, continuity anomaly, and hotspot migration distance under each condition were input into the risk scoring method of this invention for calculation. The calculation results show that under the pitting corrosion condition, P... pit At its maximum, the method of this invention determines the risk type as pitting corrosion; under high resistance / breakpoint conditions, P break Maximum, the determination type is high resistance / breakpoint; P under abnormal conductor disturbance conditions dist The maximum value was determined to be an abnormal conductor disturbance. The results of the method in this invention for all three operating conditions were consistent with the pre-set anomaly types in COMSOL, indicating that the method of this invention can distinguish and identify pitting corrosion on ordinary straight segments, high resistance / breakpoints, and abnormal conductor disturbance risks based on the differential characteristics of candidate units.

[0069] In summary, under the same grounding grid geometry and the same injection boundary conditions, this invention, through candidate unit division, structural background exclusion, multi-condition differential features, and three types of risk scoring, can obtain risk assessment results consistent with the simulation preset conditions, thus verifying the effectiveness and feasibility of the method of this invention.

[0070] Example 2: This embodiment of the method for identifying non-structural weaknesses in the grounding grid is implemented according to the following steps: Step 1: Collect basic data of the grounding grid; Step 2: Establish a unified three-dimensional coordinate system; Step 3: Divide the grounding grid branches into multiple candidate units; Step 4: Build a flawless baseline simulation model and calculate the structural background influence rate of each candidate unit; Step 5: Calculate the structural node distance, structural influence threshold, and distance threshold, and filter out non-structural candidate units based on the structural background influence rate; Step 6: Construct multiple operating conditions and extract electrical response parameters, and calculate differential characteristics; Step 7: Calculate the probability of three types of risks—pitting corrosion, breakpoint / high resistance, and abnormal conductor disturbance—based on the differential characteristics, classify the risk levels, and complete the risk identification.

[0071] Example 3: Based on Example 2, step 1 specifically includes: Collect the three-dimensional coordinates of grounding grid nodes, branch start and end relationships, conductor material and cross-sectional dimensions, burial depth, structural node locations, soil resistivity of the site area, and data on known external conductors, and store the data in a standardized manner according to five types of tables: node table, branch table, structural node table, soil parameter table, and abnormal conductor table.

[0072] Example 4: Based on Example 3, step 2 specifically includes: A global three-dimensional coordinate system is established with the geometric center of the grounding grid as the origin. The X-axis corresponds to the east-west direction of the site, the Y-axis corresponds to the north-south direction of the site, and the Z-axis is the vertical direction.

[0073] Example 5: Based on Example 4, step 3 specifically includes: The grounding grid branches are divided into several candidate units according to a fixed length. Each candidate unit is assigned a unit number, start and end coordinates, center coordinates, unit length, conductor cross-section, whether it is a pure straight line segment, and corresponding ground surface measurement point number information to form a candidate unit information table.

[0074] Example 6: Based on Example 5, step 4 specifically includes: A flawless baseline simulation was built, and the current density J and surface potential gradient G parameters of each candidate element under this condition were extracted. The structural influence ratio and structural background influence rate A were then calculated. i ; Among them, the structural influence ratio includes the current density influence ratio R. J The influence ratio of surface potential gradient to R G The calculation formula is as follows:

[0075] In the formula, X i0 X represents the current density or surface potential gradient response of candidate cell i under a defect-free baseline condition; node.peak Indicates the peak response near the nearest structural node; Xfar ε represents the average background response of a normal straight line segment far from structural nodes; ε is a small, stable quantity to prevent the denominator from returning to zero. Structural background influence rate A i =max(R J ,R G ).

Claims

1. A method for identifying risks of non-structural weaknesses in grounding grids, characterized in that, The specific steps are as follows: Step 1: Collect basic data of the grounding grid; Step 2: Establish a unified three-dimensional coordinate system; Step 3: Divide the grounding grid branches into multiple candidate units; Step 4: Build a flawless baseline simulation model and calculate the structural background influence rate of each candidate unit; Step 5: Calculate the structural node distance, structural influence threshold, and distance threshold, and filter out non-structural candidate units based on the structural background influence rate; Step 6: Construct multiple operating conditions and extract electrical response parameters, and calculate differential characteristics; Step 7: Calculate the probability of three types of risks—pitting corrosion, breakpoint / high resistance, and abnormal conductor disturbance—based on the differential characteristics, classify the risk levels, and complete the risk identification.

2. The method for identifying non-structural weaknesses in grounding grids according to claim 1, characterized in that, Step 1 is as follows: Collect the three-dimensional coordinates of grounding grid nodes, branch start and end relationships, conductor material and cross-sectional dimensions, burial depth, structural node locations, soil resistivity of the site area, and data on known external conductors, and store the data in a standardized manner according to five types of tables: node table, branch table, structural node table, soil parameter table, and abnormal conductor table.

3. The method for identifying non-structural weaknesses in grounding grids according to claim 1, characterized in that, Step 2 is as follows: A global three-dimensional coordinate system is established with the geometric center of the grounding grid as the origin. The X-axis corresponds to the east-west direction of the site, the Y-axis corresponds to the north-south direction of the site, and the Z-axis is the vertical direction.

4. The method for identifying non-structural weaknesses in grounding grids according to claim 1, characterized in that, Step 3 specifically involves: The grounding grid branches are divided into several candidate units according to a fixed length. Each candidate unit is assigned a unit number, start and end coordinates, center coordinates, unit length, conductor cross-section, whether it is a pure straight line segment, and corresponding ground surface measurement point number information to form a candidate unit information table.

5. The method for identifying non-structural weaknesses in grounding grids according to claim 1, characterized in that, Step 4 specifically involves: A flawless baseline simulation was built, and the current density J and surface potential gradient G parameters of each candidate element under this condition were extracted. The structural influence ratio and structural background influence rate A were then calculated. i ; The structural influence ratio includes the current density influence ratio R. J The influence ratio of surface potential gradient to R G The calculation formula is as follows: In the formula, X i0 X represents the current density or surface potential gradient response of candidate cell i under a defect-free baseline condition; node.peak Indicates the peak response near the nearest structural node; X far ε represents the average background response of a normal straight line segment far from structural nodes; ε is a small, stable quantity to prevent the denominator from returning to zero. Structural background influence rate A i =max(R J ,R G ).

6. The method for identifying non-structural weaknesses in grounding grids according to claim 1, characterized in that, Step 5 specifically involves: Regarding the structural background influence rate A obtained in step 4 i For candidate cells with a value <0.30, calculate the distance d from the center point of each candidate cell to the solder joint, connection point, corner point, and grounding lead connection point. 焊点 d 连接点 d 边角点 d 引下线 The minimum value is taken as the distance d between structural nodes. i ,Right now: Plot the structural background influence rate A of each candidate unit i With distance d i The decay curve is used to calculate the rate of change of influence of adjacent candidate units. When two consecutive adjacent intervals of ΔA i When all values ​​are ≤0.03 to 0.05, the influence decay is determined to have transitioned from a rapid and steep decline phase to a gradual low-influence range. The structural background influence rate corresponding to the starting position of this gradual low-influence range is extracted as the structural influence threshold A. th and A th The corresponding d i As the distance threshold D th ; If the structural node of a candidate unit is d i ≥D th And the structural background influence rate A i ≤A th If so, the candidate element is determined to be an unstructured candidate element; for distance d i ≥D th Impact rate A i Elements in the 0.15~0.30 transition range are marked as elements to be reviewed. By densifying the surface measurement points or conducting continuity retests, it is confirmed whether the element can be included in the unstructured candidate element set.

7. The method for identifying non-structural weaknesses in grounding grids according to claim 1, characterized in that, Step 6 specifically involves: For the unstructured candidate units selected in step 5, three simulation conditions were built: pitting corrosion, breakpoint / high resistance, and abnormal conductor disturbance. Electrical response data were collected, and relative differential features ΔJ, ΔG, current continuity anomaly C, and hot spot migration distance M were calculated with reference to the baseline condition. All relative differential features were normalized to the interval of 0 to 1 to obtain the normalized feature value N. J N G N C N M .

8. The method for identifying non-structural weaknesses in grounding grids according to claim 1, characterized in that, Step 7 specifically includes: N J N G N C and N M Substituting the values ​​into the weighting formula, the pitting risk P of each unstructured candidate unit is calculated sequentially. pit Breakpoint / High Resistance Risk P break Risk of abnormal conductor disturbance P dist The maximum value is taken to obtain the overall detection priority: Score = max(P) pit P break P dist ); Among them, P pit =a1N J +a2N G +a3(1-N C )+a4(1-N M ); In the formula, a1+a2+a3+a4=1; a1 takes values ​​of 0.30~0.50, a2 takes values ​​of 0.25~0.45, a3 takes values ​​of 0.05~0.20, and a4 takes values ​​of 0.05~0.15; P break =b1N C +b2N G +b3N J +b4N M ; In the formula, b1+b2+b3+b4=1; b1 takes values ​​of 0.35~0.60, b2 takes values ​​of 0.15~0.35, b3 takes values ​​of 0.10~0.30, and b4 takes values ​​of 0.05~0.20; P dist =c1N M +c2N G +c3N J +c4(1-N C ); In the formula, c1+c2+c3+c4=1; c1 takes values ​​of 0.35~0.60, c2 takes values ​​of 0.15~0.35, c3 takes values ​​of 0.10~0.30, and c4 takes values ​​of 0.05~0.20; After calculating the score, perform on-site operations according to the threshold levels: (1) Score ≥ 0.75: First priority, conduct continuity retesting first; (2) 0.55≤Score<0.75: Secondary priority, densify surface measuring points, and remeasure soil parameters and electrical response; (3) 0.35≤Score<0.55: Level 3 priority, included in the daily key inspection log, and tracked regularly; (4) Score < 0.35: No abnormal risks were found. Perform routine inspections as usual.

9. The method for identifying non-structural weaknesses in grounding grids according to claim 1, characterized in that, The method also includes a field feedback closed-loop update mechanism: based on the results of field excavation verification, continuity testing, soil retesting, and inspection misjudgment and omission, the structural influence threshold, distance threshold, risk calculation weight, soil parameters, and abnormal conductor disturbance judgment rules are corrected to achieve localized iterative optimization of the model.

Images