Monitoring processes using sensor networks and an extended Kalman filter

Two main difficulties in process monitoring are lack of reliable measurements of key process variables and difficulty in defining quantitative relationships between state variables. In this study sensor networks are used to demonstrate an approach based on Kalman filtering to model the specific monitoring systems. Kalman filtering at both local nodes and fusion center are the covariance matrices of tracking errors. Performance analysis is dedicated to the distributed Kalman filtering fusion for distributed recursive state estimators of dynamic systems under consideration. Key-Words: Distributed processing, performance, sensors, Kalman filtering