Research On Comprehensive Evaluation Method Of High Voltage Switchgear State

The operation state of electric power equipment is a dynamic change process.The establishment of the operation andmaintenance strategy based on the running state of equipment is the premise of realizing the precision operation and maintenance management of equipment.And it is a necessary part of intelligent power grid.The substation as a whole,in order to improve the reliability of network operation,it is necessary to improve the reliability of each equipment component.Maintenance strategies based on equipment running state have been developed in the power grid for many years.For a long time,the hot spots of state evaluation mainly focus on a few large expemsive equipment such as generator,transformer and so on,while there are few researches on the large number of widely used high voltage switchgear equipment.With the improvement of the reliability of power supply in recent years,the condition evaluation and maintenance of this kind of equipment have gradually attracted wide attention.This paper takes the high voltage switchgear as the research object,studies the multidimensional information state evaluation model and algorithm for the high voltage switchgear.The key problems to be solved include: qualitative and quantitative analysis between detection data and equipment insulation deterioration;accurate extraction of detection data under electromagnetic interference environment;Establishment of multi-dimensional state parametric high voltage switchgear evaluation model;Research on high voltage switchgear state evaluation algorithm based on information fusion.The main work and research results of this paper include:Combining the electromagnetic wave propagation characteristics,the quantitative relationship between the transient earth voltage detection data and the insulation degradation intensity in the switch cabinet is given by theoretical derivation.Ansoft electromagnetic simulation software was used to carry out three-dimensional modeling of the switchgear,and high-frequency current-carrying coil was used to simulate the internal local discharge.The simulation obtained the influence partial discharge position,amplitude and frequency changes of the switchgear on the TEV detection value,providing theoretical support for determining the state parameters of the high voltage switchgear state evaluation model.Based on the analysis of the characteristics of existing power equipment detection methods,combining with the installation and operation characteristics of high-voltage switchgear,through the comparison and analysis of the field detection data,the electric-acoustic-optical joint detection scheme of high-voltage switchgear suitable for the field is obtained.It provides data support for the selection of state parameters in the high voltage switchgear state evaluation model and the division of primary and secondary relations among the parameters.In order to reduce the influence of electromagnetic environment on detection data in substation,an anti-jamming algorithm for live detection data is proposed.By deviation of mutation of detection data,the influence of local electromagnetic interference on the detection value is eliminated.On this basis,the modified formula is used to further modify the data to eliminate the adverse impact brought by the increase of detection value due to the insulation deterioration in the adjacent equipment.It improves the accuracy of the detection data and lays a foundation for the accuracy of further equipment state evaluation.A multi-dimensional state parameter evaluation model for high-voltage switchgear is established.In order to better reflect the development trend of equipment insulation deterioration,the fuzzy quantity state evaluation factor is adopted in the evaluation model,and the comprehensive evaluation vector based on fuzzy quantity is obtained.At the same time,the five-level system is adopted to evaluate the status of equipment,and the evaluation vector is divided into corresponding evaluation levels according to the membership function.The state parameters of the model cover online detection data,environmental information,historical evaluation information,equipment operation information,preventive test data and routine inspection information,including multiple aspects of data information.Aiming at high-voltage switchgear equipment,a fuzzy hierarchical state evaluation algorithm based on entropy weight and information integrity is proposed.The algorithm reduced the dependence on subjective weight by entropy weight and incorporated the influence of information quantity on the weight.By introducing the concept of "information integrity" associated with the credibility of the evaluation results,the equipment state evaluation results of the algorithm in the condition where the state parameters are missing have more practical significance.In order to further eliminate the influence of subjective weights,a research on state evaluation algorithm based on catastrophe series is proposed,which can evaluate the state change of equipment from the perspective of catastrophe of equipment state.In order to reflect the distribution of equipment state more intuitively from the perspective of clustering,an evaluation algorithm based on variable clustering center modified FCM is proposed.The traditional variable clustering center FCM algorithm is modified by using the clustering center fixed value,so that it can be better applied to the actual situation of high voltage switchgear state evaluation.Based on the analysis and comparison of the three evaluation algorithms applicable to high voltage switchgear,a new evaluation algorithm based on information fusion is proposed,which can give full play to the advantages of the three evaluation algorithms and enhance the fault tolerance and effectiveness of the algorithm.Fuzzy stratification theory,catastrophe series algorithm and variable center modified clustering algorithm are used as sub-attributes of the evaluation algorithm,and bayesian fusion algorithm is adopted at the decision-making level.In order to reflect the sub-attribute characteristics more objectively and increase the confidence and fault tolerance of the evaluation system,a bayesian classification mode probability calculation method based on data information similarity is proposed.In the process of evaluation system bayes fusion decision makers,on the basis of the similarity between the decisions attribute vector of each subsystem is probability model.The feasibility and effectiveness of the fusion algorithm are verified by using the field test information,which proves the superiority of the fusion evaluation algorithm.