It is necessary to identify the flow pattern of CO2 in order to accurately measure its flow in the pipeline under carbon capture and storage(CCS)conditions.A flow pattern identification model of CO2 gas-liquid two-phase flow is established based on recurrence quantitative analysis and fuzzy C-mean(FCM)clustering algorithm.The method for analyzing structure characteristics of recurrence plots of capacitive signal of CO2 gas-liquid two-phase flow in multiple scales using wavelet transforms is proposed and the recurrence plot at each scale is quantitatively analyzed to extract the recurrence characteristic parameters of capacitive sensors for CO2 flow in different patterns.Using the FCM clustering algorithm to cluster characteristic parameters by flow pattern,identification of CO2 gas-liquid two-phase flow patterns,i.e.stratified flow,bubble flow,and mist flow,are achieved.
万方数据