CN-122020544-A - Intelligent maintenance device for electrified transformer and efficient cleaning method

CN122020544ACN 122020544 ACN122020544 ACN 122020544ACN-122020544-A

Abstract

The invention discloses an intelligent maintenance device for electrified transformers and a high-efficiency cleaning method, which comprise the steps of S1, multidimensional state sensing and data fusion, S2, intelligent diagnosis and risk assessment, S3, self-adaptive maintenance strategy generation, S4, dynamic resource scheduling and path planning, S5, cooperative control and accurate execution, S6, operation closed-loop feedback and model optimization, and S7, man-machine interaction and decision supervision. The intelligent transformer maintenance system based on the intelligent network has the advantages that the scheme integrating real-time monitoring, intelligent diagnosis, dynamic decision and accurate execution is constructed, the improvement of transformer live maintenance is realized, the microscopic state change of equipment and environment is perceived in real time by utilizing a multi-source sensor network, early accurate early warning of defects is realized based on cooperative analysis of edge calculation and a cloud platform, the system can dynamically generate and optimize a cleaning maintenance strategy and a resource allocation scheme by introducing an adaptive optimization algorithm, and the operation is ensured to be executed at the best opportunity, with the highest efficiency, the lowest cost and the highest risk.

Inventors

MA ZHEXUAN
NIU LIJUAN

Assignees

中环低碳节能技术(北京)有限公司

Dates

Publication Date: 20260512
Application Date: 20260129

Claims (10)

1. The intelligent maintenance device for the electrified transformer and the efficient cleaning method are characterized by comprising the following steps of: S1, multidimensional state sensing and data fusion, namely acquiring multidimensional operation state parameters and working condition data including leakage current, infrared thermal imaging temperature, partial discharge signals, environment temperature and humidity and high-pollution-removal images in real time through a multisource sensor network deployed in a transformer equipment body and a surrounding environment, and transmitting the data to an edge computing unit and a central cloud platform for synchronization and fusion processing; S2, intelligent diagnosis and risk assessment, namely calling a device health assessment model and a defect feature library in a central cloud platform based on the multidimensional data fused in the step S1, diagnosing the pollution grade, the overheated connection point position and the potential discharge defect of the transformer insulator in real time through comparative analysis and pattern recognition, and quantitatively assessing the current risk grade and evolution trend of the transformer insulator; S3, generating an adaptive maintenance strategy, namely dynamically generating an optimal charged maintenance strategy by utilizing an adaptive optimization algorithm according to the risk assessment result obtained in the step S2 and combining preset maintenance cost constraint, safety regulations and historical operation efficiency data, wherein the optimal charged maintenance strategy at least comprises a priority of maintenance operation, a target area, a recommended cleaning mode and an expected time window; S4, a step of dynamic scheduling and path planning of resources, wherein available resources are scheduled according to the maintenance strategy generated in the step S3, and comprise an autonomous mobile cleaning robot, a fixed cleaning device, an operation period and an energy quota; S5, cooperatively controlling and precisely executing, namely transmitting the detailed operation instruction and the path information determined in the step S4 to a corresponding field operation unit and a cleaning executing mechanism, driving the cleaning executing mechanism to precisely clean a target area under the cooperative control of an edge computing unit according to a preset scheme, and simultaneously monitoring the equipment state change and the cleaning effect in the operation process in real time through a sensor network; And S6, performing operation closed-loop feedback and model optimization, namely collecting and analyzing the whole process data of the operation after maintenance operation is completed, wherein the whole process data comprises actual energy consumption, time consumption, state parameter comparison before and after cleaning and target achievement degree, feeding back the analysis result to a central cloud platform, and performing iterative optimization on parameters of the equipment health evaluation model and the self-adaptive optimization algorithm.
2. The intelligent maintenance device for electrified transformers and the efficient cleaning method are characterized in that in the step S1, a redundancy configuration principle is adopted for deployment of a multi-source sensor network, cross verification is carried out on sensors with monitoring point positions deployed on at least two principles, a time stamp alignment and space coordinate matching technology is adopted for data fusion processing, and a Kalman filtering algorithm is utilized for noise reduction and estimation of dynamic parameters.
3. The intelligent maintenance device with the electric transformer and the efficient cleaning method according to claim 1 are characterized in that the equipment health degree assessment model in the step S2 is a classification and regression mixed model constructed based on a deep convolutional neural network, the equipment health degree assessment model is updated in a mode of combining offline training and online incremental learning, and the defect feature library is constructed by historical defect case data, simulation data and expert experience rules.
4. The intelligent maintenance device with the electric transformer and the efficient cleaning method according to claim 1 are characterized in that the path planning in the step S4 specifically adopts a hybrid algorithm combining an A search algorithm and an artificial potential field method, wherein the A algorithm is used for planning a global initial path in a grid map of a three-dimensional model of equipment, and the artificial potential field method is used for dynamically avoiding a risk area caused by unforeseen minor obstacles or electromagnetic field fluctuation according to sensor feedback in real-time execution.
5. The intelligent maintenance device for electrified transformers and the efficient cleaning method according to claim 1 are characterized in that the adaptive optimization algorithm in the step S3 is a multi-objective optimization algorithm, the adaptive optimization algorithm takes maximization of risk reduction benefits as a primary objective, weighted sum minimization of operation cost and time cost as a secondary objective, safety constraint conditions are set, a pareto optimal solution set is obtained by solving a multi-objective optimization problem, and a final execution strategy is selected from the solution set according to real-time resource availability, decision variables of the algorithm comprise starting time of cleaning operation, selected execution unit combination, operation strength of each execution unit and cleaning agent consumption, and the safety constraint conditions comprise that temperature rise of key connection points of equipment during operation cannot exceed a threshold value, minimum safety distance must be kept between an operation unit and an electrified body all the time, and the operation cannot cause misoperation of an equipment protection system.
6. The intelligent maintenance device with electricity and the efficient cleaning method for the transformer according to claim 5, wherein the adaptive optimization algorithm introduces current power grid load prediction information and environmental weather prediction data as boundary conditions during each decision, and dynamically adjusts parameter weights in an optimization model, so that maintenance strategies have long-term equipment health maintenance benefits and short-term power grid operation stability requirements.
7. The intelligent maintenance device with the electrified transformer and the efficient cleaning method are characterized in that cooperative control in the step S5 is achieved by adopting a master-slave control architecture and an event driving mechanism, a central cloud platform is used as a master controller to issue macroscopic instructions, an edge computing unit is used as a slave controller to take charge of microscopic motion control and emergency obstacle avoidance, when a sensor detects that equipment state is suddenly changed or an operation unit encounters an abnormality in real time, the event driving mechanism is triggered, and the edge computing unit starts a locally preset safety response program and simultaneously alarms to the central cloud platform.
8. The intelligent maintenance device with electricity and the efficient cleaning method for the transformer according to claim 7, wherein the safety response program at least comprises immediately suspending the current operation, controlling the operation unit to move to a preset safety standby point, and starting the standby monitoring sensor for rechecking confirmation.
9. The intelligent maintenance device with electricity and the efficient cleaning method for the transformer are characterized by further comprising S7, a man-machine interaction and decision supervision step, wherein the process data, the diagnosis results and the generated strategy suggestions in the steps S1 to S3 are displayed to operation and maintenance personnel in real time through a visual interface, and an interaction interface for manual confirmation, parameter fine adjustment or strategy rejection is provided.
10. The intelligent maintenance device with electricity and the efficient cleaning method for the transformer are characterized in that the central cloud platform adopts a micro service architecture to decouple and deploy data access service, model analysis service, optimization calculation service and instruction issuing service.

Description

Intelligent maintenance device for electrified transformer and efficient cleaning method Technical Field The invention relates to the technical field of operation and maintenance of power equipment, in particular to an intelligent maintenance device for electrified transformers and a high-efficiency cleaning method. Background Currently, for charged maintenance and cleaning of large-scale power equipment such as transformers, the traditional method mainly relies on manual periodic inspection and a preset fixed procedure. The method has the obvious defects that firstly, the real-time and multidimensional sensing capability on the running state of equipment is lacking, potential defects such as insulator pollution, connection point heating and the like cannot be accurately identified and positioned, secondly, maintenance decisions depend on experience, the time, frequency and resource allocation of cleaning operation are lack of scientific optimization, the failure or excessive maintenance caused by insufficient maintenance is easy to cause resource waste, thirdly, the automation degree of the operation process is low, the personnel safety risk is high, and the method is difficult to adapt to complex and changeable environments and equipment working conditions. With the development of smart grid construction, an operation and maintenance management method capable of realizing full-flow intellectualization, self-adaptive optimization and efficient coordination is needed. Disclosure of Invention Therefore, the invention provides the intelligent maintenance device with the electricity and the efficient cleaning method for the transformer, which are used for solving the problems that the prior art lacks real-time and multidimensional sensing capability on the running state of equipment and cannot accurately identify and position potential defects such as insulator pollution, connection point heating and the like. In order to achieve the above object, the present invention provides the following technical solutions: The intelligent maintenance device for the electrified transformer and the efficient cleaning method thereof comprise the following steps: S1, multidimensional state sensing and data fusion, namely acquiring multidimensional operation state parameters and working condition data including leakage current, infrared thermal imaging temperature, partial discharge signals, environment temperature and humidity and high-pollution-removal images in real time through a multisource sensor network deployed in a transformer equipment body and a surrounding environment, and transmitting the data to an edge computing unit and a central cloud platform for synchronization and fusion processing; S2, intelligent diagnosis and risk assessment, namely calling a device health assessment model and a defect feature library in a central cloud platform based on the multidimensional data fused in the step S1, diagnosing the pollution grade, the overheated connection point position and the potential discharge defect of the transformer insulator in real time through comparative analysis and pattern recognition, and quantitatively assessing the current risk grade and evolution trend of the transformer insulator; S3, generating an adaptive maintenance strategy, namely dynamically generating an optimal charged maintenance strategy by utilizing an adaptive optimization algorithm according to the risk assessment result obtained in the step S2 and combining preset maintenance cost constraint, safety regulations and historical operation efficiency data, wherein the optimal charged maintenance strategy at least comprises a priority of maintenance operation, a target area, a recommended cleaning mode and an expected time window; S4, a step of dynamic scheduling and path planning of resources, wherein available resources are scheduled according to the maintenance strategy generated in the step S3, and comprise an autonomous mobile cleaning robot, a fixed cleaning device, an operation period and an energy quota; S5, cooperatively controlling and precisely executing, namely transmitting the detailed operation instruction and the path information determined in the step S4 to a corresponding field operation unit and a cleaning executing mechanism, driving the cleaning executing mechanism to precisely clean a target area under the cooperative control of an edge computing unit according to a preset scheme, and simultaneously monitoring the equipment state change and the cleaning effect in the operation process in real time through a sensor network; And S6, performing operation closed-loop feedback and model optimization, namely collecting and analyzing the whole process data of the operation after maintenance operation is completed, wherein the whole process data comprises actual energy consumption, time consumption, state parameter comparison before and after cleaning and target achievement degree, feeding back the analysis result to a central