Get all the updates for this publication
The objective of this paper is first to design an Adaptive Linear Kalman Filter (ALKF) to estimate nonlinear process states and to compare the performance of the ALKF with the Extended Kalman Filter (EKF). The designed ALKF is next used to detect sensor and actuator biases which may occur either sequentially or simultaneously using a Multi Model ALKF (MMALKF). Finally the Multi Model Adaptive Linear Hâ Filter (MMALHâF) is designed to increase the robustness of bias Detection in the presence of unknown noise statistics and unmodeled dynamics. The proposed estimator is demonstrated on the variable area tank process and Continuously Stirred Tank Reactor (CSTR) process to show its efficacy.
Journal | Data powered by TypesetComputers & Electrical Engineering |
---|---|
Publisher | Data powered by TypesetElsevier BV |
ISSN | 0045-7906 |
Open Access | 0 |