CN-122017505-A - Insulation on-line state monitoring method for power transmission cable sheath

CN122017505ACN 122017505 ACN122017505 ACN 122017505ACN-122017505-A

Abstract

The application discloses a method for monitoring the insulation on-line state of a power transmission cable sheath, which mainly relates to the technical field of on-line state monitoring and is used for solving the problems that in the existing scheme, a multi-field coupling mechanism which is difficult to reflect insulation degradation in single-parameter monitoring, infrared temperature measurement is influenced by environmental radiation and surface emissivity, off-line detection needs power failure operation and response is lagged. The method comprises the steps of taking a diagram structure state flow as an input, driving a physical simulation engine ‌ based on ‌ electromagnetic-thermal-current three-field coupling, simulating a transient charge migration path of partial discharge pulse in an XLPE sheath, generating ‌ transient electromagnetic fingerprints corresponding to each node by solving a modified Maxwell equation set, inputting ‌ insulation degradation electromagnetic fingerprints into a multi-agent game model ‌ based on information entropy flow, regarding each cable node as an agent with state evaluation capability, and outputting ‌ insulation health state dynamic decision signals.

Inventors

YANG JUNSHI
DING JINYONG
ZHAO JINHUI
LI YONGXU
WU WENBING
YAN ZHENLEI
Zheng jinxin
ZHANG WANYUE
YU CHUANWEI

Assignees

国网山东省电力公司电力科学研究院
国网山东省电力公司聊城供电公司

Dates

Publication Date: 20260512
Application Date: 20260414

Claims (8)

1. A method for monitoring insulation on-line state of a power transmission cable sheath, the method comprising: determining a specific acquisition node according to the insulation length of the transmission cable sheath, and acquiring the current, the surface temperature and the environmental temperature of the grounding wire of the acquisition node; Taking all acquisition nodes along the cable as graph nodes, taking a physical distance and electrical coupling relation as edge weights, obtaining a multi-dimensional sensing data graph, taking the multi-dimensional sensing data graph as input, constructing ‌ a dynamic graph neural network, and outputting ‌ graph structure state flow; Using the state flow of the graph structure as input, driving a physical simulation engine ‌ based on ‌ electromagnetic-thermal-current three-field coupling, simulating a transient charge migration path of partial discharge pulse in an XLPE sheath, and generating ‌ transient electromagnetic fingerprints corresponding to each node by solving and correcting a Maxwell equation set; The insulation degradation electromagnetic fingerprint is input ‌ to a multi-agent game model ‌ based on information entropy flow, each cable node is regarded as an agent with state evaluation capability, and a dynamic decision signal of the insulation health state is output ‌.
2. The method for monitoring the insulation on-line state of a power transmission cable sheath according to claim 1, wherein the specific acquisition node is determined according to the insulation length of the power transmission cable sheath, and specifically comprises: obtaining ‌ total length ‌ information of transmission cable sheath insulation to be monitored; dividing the total length of the cable by the preset node spacing to calculate the theoretical ‌ acquisition node number, and if the result is not an integer, carrying out upward rounding; Starting from the initial end of the cable, determining ‌ specific physical layout positions of each acquisition node along the length direction of the cable according to preset node intervals; After theoretical calculation of uniform distribution points is completed, acquisition nodes are additionally arranged at ‌ joints, turning positions, previously known vulnerable points and preset key monitoring segment positions of the cable.
3. The method for monitoring the insulation on-line state of the power transmission cable sheath according to claim 1, wherein the method for acquiring the ground wire current, the surface temperature and the environment temperature of the acquisition node comprises the following steps: The ground wire current of the acquisition node is obtained through a clamp-on current transformer in a non-contact mode, the surface temperature of the acquisition node is monitored through an integrated temperature sensor, and the environment temperature corresponding to the acquisition node is obtained through a preset environment sensor.
4. The method for monitoring the insulation on-line state of a power transmission cable sheath according to claim 1, wherein each acquisition node along the cable is regarded as a graph node, and a multidimensional sensing data graph is obtained by taking a physical distance and an electrical coupling relation as side weights, and the method specifically comprises the following steps: Regarding each ‌ data acquisition node ‌ deployed along an insulated line of a power transmission cable sheath as one ‌ node in a graph structure, wherein an attribute vector of each node at least comprises ‌ ground wire current, ‌ surface temperature and ‌ ambient temperature; Calculating ‌ actual physical distances between any two acquisition nodes on the cable along the line, wherein the physical distances are used as first weights of edges; acquiring an electrical topological structure of the power transmission cable, giving a second weight to the first ladder by edges between nodes which are positioned in the same power supply loop, phase or have direct electrical connection, and giving a second weight to the second ladder by edges which are not positioned in the same power supply loop, phase or have direct electrical connection; Carrying out ‌ normalization processing on the first weight and the second weight respectively, inputting the first weight and the second weight subjected to normalization processing into a preset fusion function, and obtaining ‌ final weight value; All nodes, the connection relation between the multidimensional sensing attributes and all edges and ‌ final weight values are organized together into a complete multidimensional sensing data graph.
5. The method for monitoring the insulation on-line state of the power transmission cable sheath according to claim 1, wherein the steps of constructing ‌ a dynamic graph neural network and outputting ‌ a structural state flow by taking the multidimensional sensing data graph as input comprise the following steps: Setting the network layer number, node updating iteration number and characteristic propagation step length of the dynamic graph neural network based on the node number, node attribute dimension and edge weight range of the multidimensional sensing data graph; The ground wire current, the surface temperature and the environment temperature corresponding to each acquisition node are used as node state characteristics; extracting local state characteristics and integral topological characteristics of all-line nodes of the cable layer by layer through multi-layer network iterative computation, and fusing sensing data changes in time dimension to form a continuous dynamic characteristic sequence; and integrating the extracted and integrated node state characteristics and time sequence characteristics into a unified graph structure state flow for output.
6. The method for monitoring the insulation on-line state of the transmission cable sheath according to claim 1, wherein the physical simulation engine ‌ based on ‌ electromagnetic-thermal-current three-field coupling is driven by taking the state flow of the graph structure as an input, and the method for monitoring the insulation on-line state of the transmission cable sheath simulates a transient charge migration path of a partial discharge pulse in the XLPE sheath, and specifically comprises the following steps: Based on the material parameters, the geometric structure and the electrical characteristics of the XLPE sheath of the cable, a ‌ three-dimensional refined physical simulation model of the cable section is established, a ‌ multi-physical field coupling simulation engine integrating ‌ electromagnetic field, thermal field and current field ‌ is established; Converting node state features in the graph structure state flow into ‌ boundary conditions ‌ and ‌ excitation source parameters which can be identified by a simulation engine, and mapping ‌ node state features ‌ onto the spatial positions of the physical simulation model; in the simulation model, taking a cable insulation preset potential defect point as a local discharge source, ‌ simulating a preset transient high-voltage pulse sent by the local discharge source; Based on electromagnetic-thermal coupling, ‌ local joule heat ‌ generated by pulse discharge is calculated and used as input for thermal field update; Based on the thermal-electric coupling, ‌ updates the electric current field according to the influence of the material temperature change on the electric conductivity, and based on the electric current-electromagnetic coupling, ‌ forms closed loop iterative calculation by taking the updated electric current field as an excitation source of the electromagnetic field; solving a Maxwell equation set ‌ corrected by actual physical conditions through ‌, driving the whole coupling system, and calculating the propagation of electromagnetic waves in XLPE insulation media containing defects and inhomogeneities; Through a plurality of time step simulations, the whole transient propagation and evolution path of the transient electromagnetic disturbance ‌ generated by the partial discharge pulse in the cable insulation sheath is finally depicted.
7. The method for monitoring the insulation on-line state of the transmission cable sheath according to claim 1, wherein the generating of ‌ transient electromagnetic fingerprints corresponding to each node by solving the modified maxwell's equations comprises: Through a plurality of time step simulations, the whole transient propagation and evolution path of the transient electromagnetic disturbance ‌ generated by the partial discharge pulse in the cable insulation sheath is depicted; According to the path calculation, a preset transient electromagnetic response signal ‌ represented by each simulation node on the cable ‌ is output and used as a transient electromagnetic fingerprint reflecting the state of the insulation local microstructure of the point.
8. The method for monitoring the insulation on-line state of the transmission cable sheath according to claim 1, wherein the input ‌ of the insulation degradation electromagnetic fingerprint is based on the multi-agent game model ‌ of the information entropy flow, each cable node is regarded as an agent with state evaluation capability, and the output ‌ of the dynamic decision signal of the insulation health state specifically comprises: defining each data acquisition node of the transmission cable, along which a sensor is arranged, as a ‌ independent intelligent body; The ‌ insulation degradation electromagnetic fingerprint ‌ corresponding to each node agent is used as ‌ initial state information ‌ of the agent to be input into a model; inputting ‌ information entropy flows for each agent, wherein ‌ information entropy flows are information flows describing the chaos of a system ‌, and transient electromagnetic fingerprints are shared among the agents; Performing ‌ game interaction by each node agent based on the electromagnetic fingerprint state of each node agent and the input ‌ information entropy flow, wherein the goal of each agent is to evaluate ‌ insulation health state of a cable section represented by each agent; and through interaction of game and electromagnetic fingerprint states of preset rounds, ‌ quantitative evaluation signals, which are finally output by each node intelligent agent, about the insulation health state of the corresponding cable segment are obtained.

Description

Insulation on-line state monitoring method for power transmission cable sheath Technical Field The application relates to the technical field of online state monitoring, in particular to a method for monitoring the insulation online state of a power transmission cable sheath. Background Current insulation monitoring of transmission cable jackets generally relies on conventional means such as ground wire current measurement, surface temperature detection, infrared thermal imaging, partial discharge off-line testing, and the like. These methods are usually based on single physical quantity collection, such as measuring sheath ground circulation by a clamp ammeter or obtaining cable surface temperature distribution by using an infrared thermometer, and the local discharge is usually detected at fixed points under the power failure state by adopting a pulse current method or an ultrasonic method. The monitoring nodes are distributed sparsely, synchronous sensing capability on the whole cable electrical-thermal-mechanical coupling state is lacking, and most systems can only provide static or periodic data and cannot realize continuous online evaluation. The prior art has the limitations that a multi-field coupling mechanism of insulation degradation is difficult to reflect by single-parameter monitoring, the ground current is easy to be influenced by soil resistivity, stray current and multi-point ground interference, evaluation distortion is caused, partial discharge detection is difficult to capture nanosecond transient pulses due to insufficient sampling frequency and weak noise resistance, positioning errors generally exceed 1 meter, infrared temperature measurement is influenced by environmental radiation and surface emissivity, internal defects cannot be identified by penetrating a sheath, offline detection needs power failure operation, response is delayed, and state maintenance requirements cannot be supported. Disclosure of Invention The application provides a method for monitoring the insulation on-line state of a power transmission cable sheath, which aims to solve the problems that in the existing scheme, a multi-field coupling mechanism which is difficult to reflect insulation degradation in single-parameter monitoring, infrared temperature measurement is influenced by environmental radiation and surface emissivity, offline detection needs power failure operation, and response is lagged. In a first aspect, the present application provides a method for monitoring insulation on-line state of a power transmission cable sheath, the method comprising: determining a specific acquisition node according to the insulation length of the transmission cable sheath, and acquiring the current, the surface temperature and the environmental temperature of the grounding wire of the acquisition node; Taking all acquisition nodes along the cable as graph nodes, taking a physical distance and electrical coupling relation as edge weights, obtaining a multi-dimensional sensing data graph, taking the multi-dimensional sensing data graph as input, constructing ‌ a dynamic graph neural network, and outputting ‌ graph structure state flow; Using the state flow of the graph structure as input, driving a physical simulation engine ‌ based on ‌ electromagnetic-thermal-current three-field coupling, simulating a transient charge migration path of partial discharge pulse in an XLPE sheath, and generating ‌ transient electromagnetic fingerprints corresponding to each node by solving and correcting a Maxwell equation set; The insulation degradation electromagnetic fingerprint is input ‌ to a multi-agent game model ‌ based on information entropy flow, each cable node is regarded as an agent with state evaluation capability, and a dynamic decision signal of the insulation health state is output ‌. In one implementation of the present application, determining a specific acquisition node according to a length of insulation of a power transmission cable sheath specifically includes: obtaining ‌ total length ‌ information of transmission cable sheath insulation to be monitored; dividing the total length of the cable by the preset node spacing to calculate the theoretical ‌ acquisition node number, and if the result is not an integer, carrying out upward rounding; Starting from the initial end of the cable, determining ‌ specific physical layout positions of each acquisition node along the length direction of the cable according to preset node intervals; After theoretical calculation of uniform distribution points is completed, acquisition nodes are additionally arranged at ‌ joints, turning positions, previously known vulnerable points and preset key monitoring segment positions of the cable. In one implementation manner of the application, the acquisition of the ground line current, the surface temperature and the ambient temperature of the acquisition node specifically comprises: The ground wire current of the acquisition node is obtain