电工技术学报2024，Vol.39Issue(11)：3475-3485,11.DOI:10.19595/j.cnki.1000-6753.tces.230336
基于扩展卡尔曼滤波的水中放电阶段辨识方法
Method for Identifying Stages of Discharge in Water Based on Extended Kalman Filter
摘要
Abstract
High-voltage pulsed discharge in water is a complex plasma generation and collapse process involving coupled physical processes,such as electricity,light,and heat.The discharge process consists of pre-breakdown,violent,and oscillatory discharge stages.The main existing method for determine the discharge stage is by using high-speed camera photography,but this approach is not suitable for practical applications environment of pulsed discharge technology in water.Additionally,the development of the arc channel during repeated discharges exhibits strong randomness,making it difficult to classify discharge stages within a fixed time domain.Therefore,studying the identification method of the underwater discharge stage is crucial.This paper proposes a corresponding method based on the arc impedance change characteristics of the discharge in water. Firstly,the plasma change characteristics observed at each discharge stage are amalgamated to elucidate the variations in resistance.The initial plasma channel is formed during the pre-breakdown discharge stage.It continues to ionize to form a more robust initial arc channel under the action of an external electric field,with a more stable and linear decline in resistance.Subsequently,a dense arc is established once the plasma density within the channel surpasses a critical threshold,leading to a sudden surge in plasma density and a rapid decrease in resistance.Nevertheless,with the diminishing strength of the external electric field,sustaining a balanced arc discharge becomes impracticable,resulting in the extinction of the arc.The channel retains a substantial plasma that gradually diminishes over time at this stage and makes the resistance progressive increase. Secondly,the discharge process in water is mainly carried out in an inhomogeneous medium.Hence,the arc channel morphology and spatial location are random,and the resistance values within each discharge cycle in a repetitive discharge environment also vary.The Extended Kalman Filtering algorithm filters the measured resistance data against the randomness,thus obtaining a representative value of the identified resistance.Solving the slope of the resistance change at each moment and combining it with the characteristics of the changes in each discharge stage,a threshold value for the resistance change rate is derived.This threshold value serves as a criterion to determine the transition of the discharge stage. Finally,the experimental platform for pulse discharge observation in water confirmed the effectiveness of the method,and an analysis of errors was conducted.Further experimental validation analyses were carried out at the biggest distance electrode gaps of 2.04 mm and 3.01 mm based on gap distances equal to 0.62 mm,1.01 mm,and 1.38 mm.Additionally,the conductivity of the solution is increased to 4 685.13 μS/cm,6 740.45 μS/cm,7 820.62 μS/cm,12 340.68 μS/cm,and 19 319.25 μS/cm,respectively.The results indicate that the variance of the identified resistance values obtained based on the Extended Kalman Filter algorithm for underwater pulse discharge is less than 1.717 Ω2 compared to the experimentally measured resistance values.Moreover,the variances of the two type's resistance slopes are less than 1.899,with a relative error of the resistance slope being less than 8.9%.Despite variations in electrode gap distance and solution conductivity,this method maintains high accuracy.Thus,it is suitable for determining discharge stages in the underwater pulse discharge process.关键词
水中脉冲放电/时变电阻/扩展卡尔曼滤波/放电阶段辨识Key words
Pulse discharge in water/time-varying resistance/extended Kalman filtering(EKF)/discharge stage identification分类
信息技术与安全科学引用本文复制引用
高崇,邵在康,康忠健,傅雪原,侯腾飞..基于扩展卡尔曼滤波的水中放电阶段辨识方法[J].电工技术学报,2024,39(11):3475-3485,11.基金项目
国家重大科技专项(2016ZX05034004)和中国石油大学(华东)研究生创新基金(YCX2021109)资助项目. （2016ZX05034004）
