Two-phase Flow Measurement With Multi-sensor Data Fusion Based On Kalman Filter

Two phase flow widely exists in nature and industrial production.With The development of scientific research and industrial production,parameter measurement become a highlight spot due to the increasing demand.The complex model and the numerous parameters make it difficult.Single sensor is only available for certain types of flow patterns or can measure a part of parameters,therefore single sensor cannot obtain accurate and comprehensive description of multiphase flow.Data fusion combines data from multiple sensors to achieve improved accuracy and more specific inferences than could be achieved by the use of a single sensor.Multi-sensor data fusion is introduced into two phase flow measurement to improve the accuracy and expand the measurement range.In this study,the water hold up and flow rate of oil-water two-phase flow are measured using capacitance,conductance and ultrasonic sensors.Data fusion was used to synthesize,analyze and utilize the data.The research works is as follows:(1)Overall fusion structure of oil-water two-phase flow parameters.The characteristics and measurement principles of capacitance,conductance and ultrasonic sensors are summarized and the difference between water holdup and flow velocity measurement is analyzed.Based on multi-sensor data fusion theory,a two-layer fusion model of water holdup and flow velocity is established,(2)Water holdup fusion measurement.Capacitance conductance sensor combined with Conductance ring sensors is used to extract the water holdup characteristic information of oil-water flow.Kalman estimation theory is used to analyze the data.T Kalman centralized fusion and distributed fusion are studied and the working state of the sensor is judged by contextual information.Based on this,the data fusion measurement method of capacitance and conductance ring sensors is designed to improve accuracy and expand the measurement range.(3)Flow rate fusion measurement.Aiming at solving the problem of velocity measurement of oil-water two-phase flow,ultrasonic sensor and conductance ring sensor are used to measure the flow velocity information.The two sensors are used to correct each other's data to avoid the large error.Based on the distributed Kalman fusion method,Doppler velocity model is linearized and the further fusion of the two sensors is realized,and the aim of improving the accuracy is achieved.