CN-122017360-A - Substation grounding grid health state assessment and early warning method based on multi-source data fusion

CN122017360ACN 122017360 ACN122017360 ACN 122017360ACN-122017360-A

Abstract

The invention relates to the technical field of power equipment detection, and particularly discloses a substation grounding grid health state assessment and early warning method based on multi-source data fusion, which comprises the following steps of gridding a grounding grid area, measuring grounding resistance and environmental parameters of each area, and calculating corrected resistance growth rate time sequence; detecting the resistivity, PH value and chloride ion content of soil, evaluating the corrosion grade of the soil, identifying a resistance sudden increase and strong corrosion area according to the time sequence of the resistance growth rate and the corrosion grade of the soil, marking the corrosion suspicious area, injecting test current into a node below the suspicious area, measuring voltage response, calculating the resistance parameter of a branch by inversion of an optimization algorithm, calculating the corrosion anomaly degree and generating a branch investigation priority sequence. According to the invention, the dynamic change of the ground resistance and the static evaluation of soil corrosion are fused, the preliminary identification and the dynamic early warning of the corrosion risk area are realized, the investigation range is reduced, and the inversion precision and the diagnosis efficiency are improved.

Inventors

DING YUEMING
LI WENJUN
CUI RONGXI
MA DECAO
JIA TINGBO
WANG WEI
SHEN CHEN
WANG FENG
LIU TIANCHENG
SHI XIAOXIA

Assignees

国网山东省电力公司日照供电公司

Dates

Publication Date: 20260512
Application Date: 20260130

Claims (10)

1. The substation grounding grid health state assessment and early warning method based on multi-source data fusion is characterized by comprising the following steps of: Dividing the ground network area into a plurality of areas, measuring the grounding resistance and the environmental parameters of each area in the current period, calling historical period measurement data, calculating the resistance increase rate based on the difference value of the grounding resistances in adjacent periods, and correcting by combining the environmental parameters to obtain the resistance increase rate time sequence of each area; soil sampling points are arranged in the areas, the soil resistivity, the PH value and the chloride ion content of the soil sample at the sampling points are detected, and the soil corrosiveness score of each area is evaluated; Identifying whether a resistance surge and strong corrosion area exists according to the resistance growth rate time sequence and the soil corrosion score of each area, if so, judging that the ground network is suspected to be corroded, marking the corrosion suspicious area, and if not, continuing to perform health monitoring of the next period; Injecting test current into a plurality of contactable nodes of the ground network below the corrosion suspicious region, measuring node voltage response, calculating resistance parameters of branches of the ground network by inversion of an optimization algorithm, calculating corrosion anomaly of each branch and generating a branch investigation priority sequence.
2. The method for evaluating and early warning the health state of the transformer substation grounding grid based on multi-source data fusion according to claim 1, wherein the environmental parameters comprise soil surface temperature and soil humidity.
3. The method for evaluating and early warning the health state of the transformer substation grounding network based on multi-source data fusion according to claim 2, wherein the method for acquiring the resistivity growth rate time sequence of the measuring area comprises the following steps: Calculating the grounding resistance increment of the measuring area in the current period, wherein the grounding resistance increment is a difference value obtained by subtracting the grounding resistance value of the previous period from the grounding resistance value of the current period; Calculating the environmental parameter variation according to the environmental parameters of the measuring area in the current period and the previous period, and inputting the environmental parameter variation into a preset environmental parameter-ground resistance influence model to obtain the interference quantity of the ground resistance increment; Subtracting the interference quantity of the grounding resistance increment from the grounding resistance increment of the measuring area in the current period to obtain a corrected grounding resistance increment; Calculating the ratio of the corrected ground resistance increment to the ground resistance value of the previous period to obtain the resistance increase rate of the measuring area in the current period; based on the same calculation method, obtaining the resistance increase rate of the measuring area in each history period; and arranging the resistance growth rates of all the historical periods and the current periods in time sequence to generate a resistance growth rate time sequence of the measuring area.
4. The method for evaluating and early warning the health state of the transformer substation grounding network based on multi-source data fusion according to claim 3, wherein the method for constructing the environmental parameter-grounding resistance influence model comprises the following steps: Selecting one or more experimental sheet areas in the grounding grid area; based on a single variable principle, changing the soil surface temperature of the experimental sheet area in an equal gradient mode, measuring the grounding resistance of the experimental sheet area at different soil surface temperatures, and obtaining data of a plurality of groups of soil surface temperature variation and corresponding grounding resistance variation, wherein the variation is a signed value; constructing a test set based on a plurality of groups of data, and utilizing a machine learning algorithm to establish a mapping relation between the soil surface layer temperature variation and the ground resistance variation; The same principle of single variable is adopted, and a mapping relation between the soil humidity variable quantity and the ground resistance variable quantity is established; And constructing an environmental parameter-ground resistance influence model according to the established mapping relation between the soil surface temperature variation and the soil humidity variation and the ground resistance variation, wherein the model takes the soil surface temperature variation and the soil humidity variation as input and takes the ground resistance accumulated variation as output.
5. The method for evaluating and early warning the health state of the transformer substation grounding grid based on multi-source data fusion according to claim 1 is characterized by comprising the following steps of: determining the center point and four corner points of the grid of the measuring area as soil sampling positions; Acquiring the buried depth of the grounding grid conductor, setting a distance upwards by taking the buried depth as a reference, and determining a corresponding depth value as the soil sampling depth; And laying a plurality of soil sampling points in each area according to the determined sampling positions and sampling depths.
6. The method for evaluating and early warning the health state of the transformer substation grounding grid based on multi-source data fusion according to claim 1, wherein the method for evaluating the soil corrosiveness score of the measuring area comprises the following steps: Based on soil corrosiveness scoring rules pre-stored in a database, acquiring corrosiveness scores corresponding to different soil resistivity intervals, different PH value intervals and different chloride ion content intervals; Matching and accumulating corresponding corrosion scores according to the measured soil resistivity, PH value and chloride ion content of the soil sample at each sampling point in the area to obtain soil corrosion scores at each sampling point; and (3) carrying out average value calculation on the soil corrosiveness scores at all sampling points in the area to obtain the soil corrosiveness score of the area.
7. The method for evaluating and early warning the health state of the transformer substation grounding grid based on multi-source data fusion according to claim 1, wherein the method for identifying whether the resistance jump and the strong corrosion area exist comprises the following steps: Judging whether each area meets the following conditions at the same time according to the resistivity growth rate time sequence and the soil corrosiveness score of each area: (1) The resistance growth rate time sequence has a continuous subsequence with the value larger than zero, and the whole subsequence has a monotonically increasing trend, wherein the length of the continuous subsequence is larger than or equal to 2; (2) The soil corrosiveness score is larger than or equal to a score threshold value corresponding to a preset strong corrosion grade; If at least one area simultaneously meets the conditions, judging that a resistance surge and strong corrosion area exists, and splicing the areas meeting the conditions to determine a corrosion suspicious area; Otherwise, judging that no resistance surge and strong corrosion area exists.
8. The method for evaluating and early warning the health state of the transformer substation grounding network based on multi-source data fusion according to claim 7, wherein the method for determining the scoring threshold value corresponding to the strong corrosion level is as follows: Extracting a reference score corresponding to the strong corrosion grade from the database; Traversing continuous subsequences with the numerical value larger than zero, and screening out peak values of the resistance growth rate in the subsequences; if the resistance increase rate peak value is smaller than or equal to a preset resistance increase rate upper limit, directly determining the reference score as a score threshold value corresponding to the strong corrosion grade; otherwise, calculating the overrun of the resistivity increase rate, calculating the downregulation of the reference score according to the preset downregulation of the strong corrosion reference score corresponding to the unit overrun, and subtracting the downregulation from the reference score to obtain the score threshold corresponding to the strong corrosion grade.
9. The method for evaluating and early warning the health state of the grounding network of the transformer substation based on multi-source data fusion according to claim 1, wherein the method for calculating the corrosion anomaly degree of each branch comprises the following steps: D1, establishing a resistance network model corresponding to a grounding grid structure below a corrosion suspicious region according to a design drawing of the grounding grid, and identifying all nodes and branches in the model; D2, selecting a plurality of nodes connected with the grounding downlead from nodes of the resistor network model as contactable test ports; Injecting test current between the selected pairs of test ports respectively, and synchronously measuring voltage drop between each pair of ports to obtain a plurality of groups of port voltage-current measurement data; D3, taking a plurality of groups of port voltage-current measurement data as input, adopting an optimization algorithm, and iteratively adjusting the resistance parameters of each branch in the resistance network model by taking the overall error between the voltage drop predicted value between each selected port pair and the corresponding actual measured value calculated by the resistance network model as a target; Step D4, repeating the step D3 until the total error is smaller than a preset convergence threshold value or the iteration number reaches a preset upper limit; D5, outputting an optimal estimated value of each branch resistance parameter in the resistance network model after iteration convergence; And D6, calculating the percentage of the difference value between the optimal resistance estimated value and the theoretical resistance value relative to the theoretical resistance value of each branch, and taking the percentage as the corrosion anomaly degree of the branch.
10. The method for evaluating and early warning the health state of the transformer substation grounding network based on multi-source data fusion according to claim 1 is characterized in that the method for generating the branch investigation priority sequence is as follows: And sequencing all the branches according to the corrosion anomaly value of each branch from high to low to generate a branch corrosion investigation priority sequence.

Description

Substation grounding grid health state assessment and early warning method based on multi-source data fusion Technical Field The invention relates to the technical field of power equipment detection, in particular to a transformer substation grounding network health state assessment and early warning method based on multi-source data fusion. Background The grounding grid is used as a key component of the safe operation of the transformer substation, and the health state of the grounding grid is directly related to the stability of the power system and the safety of personnel and equipment. However, since the grounding grid is buried deeply, has a complex structure and is not directly observable, and belongs to a typical 'black box system', the evaluation and monitoring of the health status thereof is always a challenging technical problem. At present, the nondestructive detection method of the grounding grid mainly relies on ground measurement data to carry out mathematical inversion so as to realize evaluation under the conditions of no excavation, online or electrification. In the prior art, although a corrosion diagnosis method for a transformer substation grounding grid is provided based on a simplified network circuit model and a Taylor theorem, the method has obvious defects that firstly, the method does not perform preliminary identification and range reduction on a corrosion area before mathematical inversion, so that the calculation complexity of an inversion process is high, the solution is not unique, and the accuracy and efficiency of diagnosis are affected. Secondly, the evaluation dimension is single, inversion is carried out only by relying on electric measurement data, the change trend of the grounding resistance, soil corrosiveness and other multi-source information are not comprehensively considered, the real corrosion risk of the grounding grid cannot be comprehensively and dynamically reflected, and especially in the areas with strong soil corrosiveness and rapid resistance growth, missed judgment or false judgment easily occurs. Therefore, a method for fusing multi-source data, dynamically evaluating the health status of the grounding grid and realizing accurate early warning is needed to improve the reliability, efficiency and adaptability of the grounding grid corrosion diagnosis. Disclosure of Invention Aiming at the problems, the invention provides a transformer substation grounding network health state assessment and early warning method based on multi-source data fusion, which comprises the following specific technical scheme: S1, dividing a grounding grid area into a plurality of areas, measuring the grounding resistance and environmental parameters of each area in the current period, calling historical period measurement data, calculating the resistance growth rate based on the difference value of the grounding resistances in adjacent periods, and correcting by combining the environmental parameters to obtain the resistance growth rate time sequence of each area. And S2, arranging soil sampling points in the areas, detecting the soil resistivity, the PH value and the chloride ion content of the soil sample at the sampling points, and evaluating the soil corrosiveness scores of the areas. And S3, identifying whether a resistance sudden increase and strong corrosion area exists according to the resistance growth rate time sequence and the soil corrosiveness score of each area, judging that the ground network is suspected to be corroded if the resistance sudden increase and strong corrosion area exists, marking the corrosion suspicious area, and continuing to monitor health of the next period if the resistance sudden increase and strong corrosion area does not exist. And S4, injecting test currents into a plurality of contactable nodes of the ground network below the corrosion suspicious region, measuring node voltage response, calculating resistance parameters of the branches of the ground network by inversion of an optimization algorithm, calculating corrosion anomaly of each branch and generating a branch investigation priority sequence. Compared with the prior art, the method for evaluating and early warning the health state of the transformer substation grounding grid based on multi-source data fusion has the following beneficial effects that 1. By gridding the grounding grid area, comprehensively measuring the grounding resistance and the environmental parameter, calculating the corrected resistance growth rate time sequence, combining with the soil corrosiveness score, identifying the resistance surge and strong corrosion area, and further, before entering complex mathematical inversion, the corrosion suspicious area can be primarily defined, the subsequent refined detection range is obviously reduced, the inversion calculation difficulty and non-uniqueness are reduced, and the overall diagnosis efficiency is improved. 2. According to the invention, not only is static evalua