Analysis And Study Of The Composition Of Superheated Decomposition Of Generator Insulation Material

The on-line monitoring device for generator failure is an important device for monitoring generator failure at present.The monitoring device represented by the domestic FJR series has the advantages of wide range of application,long-term continuous online monitoring,and high accuracy of generator failure warning.However,after the rapid detection of overheating fault,how to determine the fault site and the degree of overheating has become a major problem,the collection of on-site samples need to be sent to the corresponding manufacturers for chemical analysis.In order to solve this engineering problem,this project is based on the idea of pushing back the generator state by quickly analyzing the organic gas samples volatile by the generator during overheating failure,with the goal of accurately determining the overheating site of the generator and the degree of overheating.In this paper,the gas chromatography-mass spectrometry technology is used to establish a database of overheated volatile organic gas products of insulation materials,and then analyze the similarities and differences of volatile gases in each material,and establish a method system to determine the hot spot and the degree of overheating.First of all,through chromatography-mass spectrometry report and other direct data comparison analysis of the similarities and differences of the products of the library,to get a comparative library,the establishment of an in-depth identification system.Mainly includes the GC-MS test results of statistical analysis,summarized the combustion of various insulation materials volatile common substances,unique material types of the library.Based on GC-MS qualitative results,the methods of preliminary judgment,correspondence analysis and in-depth analysis,as well as retention time comparison analysis,are discussed.This paper explores the idea of judging the degree of overheating based on the analysis of relative content,quality loss data,overheating temperature and the number of species produced in the three dimensions of GC-MS quantitative results.Then try to use the chemical measurement method to analyze the retention time and peak area information which can represent the composition information in the spectrum,and establish a simple and qualitative discrimination system.The analysis process is to analyze the amount of map information by the entropy calculation method of Xiannon,and set a tone for the analysis effect.The correlation analysis is then carried out to discuss the correlation of the retention time data for each group.Finally,cluster analysis and discrimination analysis are carried out to obtain the similarity and difference of each component.The feasibility of this method is confirmed by comparison with the map.