CN-121984216-A - Power transmission line state monitoring and intelligent early warning system based on edge calculation

CN121984216ACN 121984216 ACN121984216 ACN 121984216ACN-121984216-A

Abstract

The invention discloses a power transmission line state monitoring and intelligent early warning system based on edge calculation, which relates to the technical field of power transmission line monitoring and comprises a node self-checking module, a trusted checking module and a real-time early warning module, wherein the node self-checking module synchronously acquires power transmission line running state data and node working state data and generates a related data set through self-checking; the real-time early warning module screens candidate edge nodes and selects core nodes, receives data by taking the core nodes as the center, and utilizes the received data to execute regional collaborative judgment and hierarchical intelligent early warning. The system realizes the credible verification of the data source and the efficient cooperation of the region, improves the monitoring accuracy and the early warning timeliness, and adapts to the real-time monitoring requirement of the power transmission line.

Inventors

CHEN JICUI
XU FANFAN
SHAO CHANGBO
MA YIBO

Assignees

江苏合纵智慧能源有限公司

Dates

Publication Date: 20260505
Application Date: 20260211

Claims (10)

1. Power transmission line state monitoring and intelligent early warning system based on edge calculation, which is characterized by comprising: The node self-checking module synchronously collects running state data of the power transmission line and self-operating state data of the node through each edge node, carries out real-time self-checking based on the self-operating state data, generates node self-checking state information, associates the running state data, the operating state data and the self-checking state information one by one and carries out localized storage to form a node original association data set, wherein the running state data of the power transmission line comprises a wire vibration frequency and an insulator leakage current value; The reliability verification module is used for extracting continuous time sequence data of all acquisition dimensions in the running state data of the power transmission line by each edge node based on the node original association data set, comprehensively analyzing the continuous time sequence data to generate running state comprehensive scores of each edge node, carrying out association matching on the running state comprehensive scores and self-checking state information corresponding to the edge node to mark out the reliability scores of the edge node, and binding the reliability scores with the continuous time sequence data and the running state comprehensive scores corresponding to all the acquisition dimensions to form an edge node reliability diagnosis set; The real-time early warning module screens out edge nodes with reliability scores larger than a preset threshold value from the reliability diagnosis sets of all the edge nodes, divides the edge nodes into candidate edge node sets, selects core edge nodes from the candidate edge node sets, uses the core edge nodes as data receiving centers, uses nodes belonging to the candidate edge node sets in adjacent edge nodes preconfigured by the core edge nodes as data sending ends, transmits the reliability diagnosis sets of the edge nodes to the core edge nodes by the data sending ends, and extracts running state time sequence data of the edge nodes and all the sending ends by the core edge nodes so as to execute cooperative judgment of a power transmission line region and perform intelligent early warning according to judgment results.
2. The edge calculation-based power transmission line state monitoring and intelligent early warning system according to claim 1, wherein the specific operation of generating node self-checking state information is as follows: extracting V, R, T of continuous N acquisition periods by each edge node, wherein each acquisition period corresponds to a unique time stamp, and constructing a three-dimensional coordinate point cluster according to the time stamps, namely V, R, T of each acquisition period corresponds to one three-dimensional coordinate point P k (V k ,R k ,T k ), so as to form a coordinate point sequence { P 1 ,P 2 ,...,P n }, wherein k is [1, N ]; Carrying out standardization processing on each coordinate point to obtain a standardized coordinate point P k '(V k ',R k ',T k ', wherein a specific standardization expression of each coordinate point corresponding to V, R, T is ：V k '=V k /V nom ;R k '=R k /R nom ;T k '=1-T k /T max ;, V nom is the rated power supply voltage of the sensing module, R nom is the rated transmission rate of the communication module, and T max is the maximum allowable processing time delay of the computing module; based on the standardized coordinate point sequences { P1', P2',..pn ' }, respectively performing space aggregation verification and track continuity verification, and screening all abnormal coordinate points; calculating the duty ratio of the abnormal coordinate point And according to the formula Calculating a self-checking coefficient K, wherein A is a space aggregation verification score, if the space aggregation verification score is qualified, taking 1, otherwise taking 0;B as a track continuity verification score, if the space aggregation verification score is qualified, taking 1, otherwise taking 0;w 1 、w 2 as aggregation and continuity weights respectively, and satisfying w 1 ＞w 2 and w 1 +w 2 =1; Outputting a self-checking state according to K, wherein if K is more than or equal to K th , outputting a self-checking normal, otherwise outputting a self-checking abnormal, wherein K th is a coefficient threshold; Binding the self-checking coefficient and the self-checking state of each edge node as node self-checking state information.
3. The edge-calculation-based power transmission line state monitoring and intelligent early warning system according to claim 2, wherein the specific steps of performing a spatial aggregation check are: Calculating a clustering center Q of the coordinate point cluster, wherein Q is a normalized coordinate point sequence mean value; And calculating Euclidean distance D k from each coordinate point P k ' to the clustering center Q, if D k of all coordinate points is less than or equal to D max , judging that the space aggregation is qualified, otherwise, judging that the space aggregation is not qualified, and marking abnormal coordinate points exceeding D max , wherein D max is a distance threshold.
4. The edge-calculation-based power transmission line state monitoring and intelligent early warning system according to claim 2, wherein the specific operation of performing the track continuity check is: Calculating Euclidean distance L k between two adjacent coordinate points P k 'and P k+1 '; Calculating the change quantity of each dimension of adjacent coordinate points, namely DeltaV ' = |V k '-V k+1 '|、△R'=|R k '-R k+1 '|、△T'=|T k '-T k+1 ' |, and a change direction sign, namely DeltaV=sgn (V k '-V k+1 ')、△R=sgn(R k '-R k+1 ')、△T=sgn(T k '-T k+1 '), wherein sgn () is a sign function; if the following two conditions are satisfied at the same time, the trace continuity is judged to be qualified: l k of all adjacent coordinate points is not more than L max , wherein L max is the maximum allowable adjacent distance; The second condition is that the following continuous rule judgment is carried out on all adjacent coordinate points, when DeltaV 'is less than or equal to DeltaV max , deltaR' is less than or equal to DeltaR max and DeltaT 'is less than or equal to DeltaT max , when DeltaV' is greater than or equal to DeltaV max , deltaR 'is less than or equal to DeltaR max ' and DeltaT 'is less than or equal to DeltaT max ', and simultaneously, deltaV= -DeltaR and DeltaV= -DeltaT are required to be satisfied; Wherein DeltaV max is the maximum allowable stable variation of V ', deltaR max is the maximum allowable stable variation of R ', deltaT max is the maximum allowable stable variation of T ', deltaR max ' is the maximum allowable variation of R ' when V ' changes, deltaT max ' is the maximum allowable variation of T ' when V ' changes; If the track continuity is judged to be disqualified, the adjacent coordinate points of L k ＞L max are screened out, if the adjacent coordinate points meet the continuous rule judgment of the second condition, the previous coordinate point of the adjacent coordinate points is marked as an abnormal coordinate point, and if the adjacent coordinate points do not meet the continuous rule judgment of the second condition, the two adjacent coordinate points are marked as abnormal coordinate points.
5. The power transmission line state monitoring and intelligent early warning system based on edge calculation according to claim 1, wherein the specific step of generating the corresponding running state comprehensive score of each edge node is as follows: Extracting a wire vibration frequency time sequence F z ={F z1 ,F z2 ,...,F zN and an insulator leakage current value time sequence I q ={I q1 ,I q2 ,...,I qN of continuous N acquisition periods from a node original association data set by each edge node, and ensuring that acquisition time stamps of the two sequences are completely corresponding; Respectively constructing a first-order time sequence dynamic differential sequence for the two time sequence sequences, wherein the differential sequence of F z is delta F z ={F z2 -F z1 ,F z3 -F z2 ,...,F zN -F z(N-1) },I q , the differential sequence of F is delta I q ={I q2 -I q1 ,I q3 -I q2 ,...,I qN -I q(N-1) , and the change symbols of each differential element are synchronously recorded to form a two-dimensional differential symbol sequence sgn (delta F z )、sgn(△I q ); Based on the two first-order sequential dynamic differential sequences DeltaF z 、△I q and the corresponding symbol sequences, a 2×2 sequential dynamic coupling matrix M DF is created, and the specific expression is: ; wherein the main diagonal element 、 Respectively absolute value average value of corresponding dimension difference sequences, and non-main diagonal elements Is the dot product mean value of the two-dimensional differential sequence, For physically coupling sign coefficients, according to the formula The number of differential pairs with inverted signs)/N-1, which represents the corresponding differential pair (sgn (Δf zk ),sgn(△I qk )) when sgn (Δf zk )+sgn(△I qk ) =0; based on the time sequence dynamic coupling matrix M DF , extracting the trace tr (M DF ) and determinant |M DF | of the matrix, and according to the formula And calculating a running state comprehensive score, wherein F epsilon (0, 1).
6. The edge-calculation-based power transmission line state monitoring and intelligent early warning system according to claim 1, wherein the specific operation of calculating the edge node reliability score is as follows: Splitting F, K according to the edge node numbers, extracting a unique number pair (F i ,K i ) for each edge node i, calculating C i '=F i ×K i , and generating a full edge node sequence C'; Carrying out ascending order on the C 'according to the numerical value to obtain C s '={C s1 ',C s2 ',...,C sm ' }, extracting a value before and after each value C si 'in the sequence to be used as a neighborhood basic set U base ={C s(i-1) ',C s(i+1) ' }, taking the adjacent next value if the value is a head node, and taking the adjacent previous value if the value is a tail node; Computing neighborhood overall trend symbols for all (F, K) pairs corresponding to neighborhood basic set U base ; Obtain the current node i (F i ,K i ), calculate And (3) with If the two are equal, obtaining a single-node trend symbol of the current node i Otherwise, the device can be used to determine whether the current, Wherein, (F Datum ,K Datum ) is a (F, K) pair corresponding to the dedicated reference node; If it is And taking the corresponding C i ' as the credibility score, otherwise, marking the credibility score as 0.
7. The edge-calculation-based transmission line state monitoring and intelligent early warning system according to claim 6, wherein the neighborhood overall trend sign of all (F, K) pairs corresponding to the neighborhood basic set Ubase is calculated The specific steps of (a) are as follows: If U base contains 2 values, namely corresponding to 2 neighborhood nodes, respectively calculating the operation and inspection trend coefficients of the 2 neighborhood nodes 、 If τ1=τ2, then the neighborhood overall trend sign Otherwise, taking the operation and detection trend coefficient corresponding to the neighborhood node with large C' value from the 2 neighborhood nodes as the operation and detection trend coefficient ; If U base contains 1 value, then calculate the neighborhood node and the current edge node i But needs to meet at the same time If not, assign sgn (K Adjacent to -K i ) to 。
8. The power transmission line state monitoring and intelligent early warning system based on edge calculation according to claim 1, wherein the specific rule of selecting core edge nodes is that a preconfigured neighboring node list of each edge node is extracted from a candidate edge node set, the number of nodes belonging to the candidate edge node set in the list is counted, the number is recorded as an effective neighboring candidate number G of the edge node, an edge node with the effective neighboring candidate number G as a maximum value is selected as a core edge node, and if a plurality of candidate nodes G as the maximum value exist, an edge node with the reliability score as the maximum value is selected as the core edge node.
9. The edge calculation-based power transmission line state monitoring and intelligent early warning system according to claim 1, wherein the specific operation of performing power transmission line region cooperative determination is: The core edge node receives reliability diagnostic sets of all data sending ends, extracts two-dimensional running state time sequence data of the self and corresponding wire vibration frequency F z and insulator leakage current I q of the data sending ends, aligns all data according to time stamps, and meanwhile, constructs an area running state time sequence matrix M R1 by taking edge nodes as rows and acquisition dimensions as columns; For two columns of vectors of the regional operational state time sequence matrix M R1 , respectively calculating a two-dimensional time sequence coupling coefficient rho i of each edge node; The total number of edge nodes in the statistical region is v, the two-dimensional time sequence coupling coefficient of the ith edge node is ρ i , and the region two-dimensional coupling characteristic value is The maximum absolute difference delta R =max{|ρ i -ρ R | } between ρ i and ρ R is calculated, and the average absolute deviation of all edge nodes ρ i relative to ρ R is calculated as Delta R corresponds to an abnormal edge node, and max () is a maximum function; calculating the ratio gamma of delta R to epsilon R , and dividing the operation state grade of a power transmission line area according to gamma, wherein if gamma is less than or equal to gamma min , the operation state of the power transmission line area is judged to be normal, if gamma min ＜γ≤γ max , the operation state of the power transmission line area is judged to be slightly abnormal, and if gamma is more than gamma max , the operation state of the power transmission line area is judged to be heavy abnormal, wherein gamma min 、γ max is the lower limit and the upper limit of the ratio; Executing intelligent early warning according to the running state grade of the power transmission line area, and specifically comprising the following steps: if the running state is normal, no early warning is generated, only the area coupling state normal report is sent to the power transmission line monitoring master station, if the running state is slightly abnormal, the slight coupling abnormal early warning is generated and only pushed to the monitoring master station and the local operation and maintenance terminal where the abnormal node belongs, no on-site linkage measures are needed, if the running state is severely abnormal, the severe coupling abnormal early warning is generated, and the severe coupling abnormal early warning is synchronously pushed to the monitoring master station and the local operation and maintenance terminals of all the area edge nodes, and the on-site acousto-optic early warning of the abnormal edge node and the core edge node is triggered until the operation and maintenance personnel complete on-site confirmation.
10. The edge calculation-based power transmission line state monitoring and intelligent early warning system according to claim 9, wherein the specific steps of calculating the two-dimensional time sequence coupling coefficient ρ i of each edge node are as follows: Acquiring a wire vibration frequency time sequence F zi ={f i1 ,f i2 ,...,f iV of an ith edge node, wherein the insulator leakage current time sequence is I qi ={I q1 ,I q2 ,...,I qV , and V is the total number of acquisition points of a two-dimensional time sequence of the edge node; Calculating a discrete cross-correlation function R FI (k) of F zi and I qi , traversing all hysteresis orders k to obtain a cross-correlation function value set { R FI (k) |k epsilon [ - (V-1), V-1] }, recording the cross-correlation function value set as a cross-correlation sequence, screening the cross-correlation function value with the largest absolute value from the cross-correlation sequence, and taking the cross-correlation function value as a cross-correlation absolute peak value R FI,max ; The autocorrelation function R FF (0)、R II (0) of F zi 、I qi is calculated separately, the hysteresis order k of which is taken to be 0, and the autocorrelation peak is obtained And according to the formula A two-dimensional time-series coupling coefficient ρ i is calculated.

Description

Power transmission line state monitoring and intelligent early warning system based on edge calculation Technical Field The invention relates to the technical field of power transmission line monitoring, in particular to a power transmission line state monitoring and intelligent early warning system based on edge calculation. Background In the field of power transmission line state monitoring and intelligent early warning, although the application of edge computing technology is gradually popularized, the existing scheme still has the following defects that the monitoring requirement of safe and stable operation of a power transmission line is difficult to meet: firstly, the prior art mostly focuses on the acquisition and analysis of the running state data of the power transmission line, but ignores the direct influence of the reliability of the working state of the edge node on the quality of the data, and does not establish a correlation checking mechanism of the node self-checking and the running state data, so that the acquired data is distorted under hardware faults or abnormal working conditions, and early warning misjudgment is caused; secondly, most of the existing running state analysis selects single-dimensional time sequence data for independent processing, so that consideration of inherent physical coupling characteristics among multiple dimensions is lacked, and the overall running state of a circuit is difficult to reflect comprehensively; Thirdly, the existing scheme lacks a judgment method for quantifying the reliability of the edge nodes, cannot accurately screen high-reliability data, and is mixed with invalid data during the collaborative analysis of the areas to influence judgment accuracy, meanwhile, the collaborative judgment of the areas depends on centralized calculation, the distributed advantages of the edge nodes are not fully exerted, the problem that local abnormality is difficult to quickly locate exists, and the timeliness of final early warning response is insufficient; Therefore, there is a need for an edge-calculation-based transmission line status monitoring and intelligent early warning system. Disclosure of Invention Aiming at the defects of the prior art, the invention provides a power transmission line state monitoring and intelligent early warning system based on edge calculation, which solves the problems of unreliable data obtained by the existing power transmission line monitoring, and low efficiency of analysis of one-sided and regional collaborative early warning. In order to achieve the purpose, the invention is realized by the following technical scheme that the power transmission line state monitoring and intelligent early warning system based on edge calculation comprises: The node self-checking module synchronously collects running state data of the power transmission line and self-operating state data of the node through each edge node, carries out real-time self-checking based on the self-operating state data, generates node self-checking state information, associates the running state data, the operating state data and the self-checking state information one by one and carries out localized storage to form a node original association data set, wherein the running state data of the power transmission line comprises a wire vibration frequency and an insulator leakage current value; The reliability verification module is used for extracting continuous time sequence data of all acquisition dimensions in the running state data of the power transmission line by each edge node based on the node original association data set, comprehensively analyzing the continuous time sequence data to generate running state comprehensive scores of each edge node, carrying out association matching on the running state comprehensive scores and self-checking state information corresponding to the edge node to mark out the reliability scores of the edge node, and binding the reliability scores with the continuous time sequence data and the running state comprehensive scores corresponding to all the acquisition dimensions to form an edge node reliability diagnosis set; The real-time early warning module screens out edge nodes with reliability scores larger than a preset threshold value from the reliability diagnosis sets of all the edge nodes, divides the edge nodes into candidate edge node sets, selects core edge nodes from the candidate edge node sets, uses the core edge nodes as data receiving centers, uses nodes belonging to the candidate edge node sets in adjacent edge nodes preconfigured by the core edge nodes as data sending ends, transmits the reliability diagnosis sets of the edge nodes to the core edge nodes by the data sending ends, and extracts running state time sequence data of the edge nodes and all the sending ends by the core edge nodes so as to execute cooperative judgment of a power transmission line region and perform intelligent early warning according to judgm