Research Of Processing And Analysis Methods Of Tanδ Online Monitoring Data Environment Affection Of 110kV Capacitive Equipment

Online monitoring and fault diagnosis technologies of capacitive equipment insulation are the premise and foundation to achieve the condition-based maintenance. Online monitoring of the dielectric loss factor tanδ of capacitive equipment will be affected by temperature, humidity, contamination and other environmental factors. On the basis of an experimental investigation regarding how environmental factors influence tanδ of capacitive equipment, this paper explores the online monitoring data processing method and the data change rule analysis method under the environmental effects. The environmental factors studied in this paper have theoretical significance and practical value on the influences and data processing methods of online monitoring and fault diagnosis of capacitive equipment insulation. The main findings and research achievements are as follows:1. The influence of environmental factors on measuring tanδ of 110 kV capacitive equipment was simulated in the laboratory environment. An experiment design scheme was proposed focusing on the independent action and combined action of various environmental factors, and typical insulation problems induced by moisture of the equipment were simulated to study the changing trends and mechanisms of tanδ of a capacitive device. It has been found that interference of environmental factors such as temperature, relative humidity of atmosphere and contamination will cause the online measurement to deviate from the routine measurement of tanδ of the capacitive equipment. There are some differences in the tanδ performance of different types of capacitive devices under different environmental conditions. Besides, the more influential factors there are, the more difficult it is to extract the regularity trend, and it is hard to compensate by adhering to the uniform standards.2. This paper analyzes the impact of environmental factors on tanδ of capacitive equipment, establishes a grey relational analysis model that investigates the impact of major environmental factors(temperature, humidity and contamination) on the tanδ of capacitive devices, and gains the grey correlation degree between the prominent environmental parameters and tanδ of capacitive equipment. A strong association has validated between tanδ and environmental factors, and seasonal variation does exist among all parameter weighing factors.3. This paper also presents the data processing method about the effect of major environmental factors based on least squares support vector machine on tanδ. The environmental factors in the historical period were taken as the input data and online monitoring values of tanδ of the capacitive equipment as the output data for training. The environmental monitoring information in the target period was introduced to the training model in order to predict the historical state of tanδ within the target period. By comparing the tanδ predicted values that reflected the historical insulation condition and the real tanδ, the development trend of tanδ of capacitive equipment could be found. Further, the particle swarm optimization algorithm was employed to optimize parameters of support vector machine, which has effectively improved the correction accuracy.4. Since online monitoring is characterized by a large amount of information and information redundancy, this paper proposed a method of dielectric loss online monitoring data compression. Based on practical calculation, it can be seen that principal component analysis achieves a good processing effect on tanδ online monitoring data compression, and can effectively extract the regularity contained in the data.5. Grounded on the knowledge that lab environmental factors affect the test data and an association analysis, this paper presented the tanδ diagnostic model of least squares support vector machine underpinned by a Particle Swarm Optimization(PSO), and adopted the PSO algorithm to determine the model parameters. The online monitoring data of tanδ of the capacitive equipment were processed by using the least squares support vector machine, which can visually exhibit the changing trend of tanδ of capacitive equipment within a certain time, increase the comparability of online monitoring results, and can be used as an advanced diagnostic technique for insulation of capacitive equipment. The 220 kV substation in Xiang’.an district of Xiamen city was used for an instance analysis of the insulation online monitoring measured value, and the effectiveness of the method has been verified.