A model-based method for the detection and estimation of faults in dynamic systems is proposed. The method is based on the combination of the parity space approach and the modulating function framework for estimation....
详细信息
A model-based method for the detection and estimation of faults in dynamic systems is proposed. The method is based on the combination of the parity space approach and the modulating function framework for estimation. The parity space method is employed as an efficient geometric procedure determining null subspaces for annihilating unknown terms and formulating residuals. With the modulatingfunctions technique the dynamic relation from output differentiation is reformulated as an algebraic expression. This substantially reduces the noise sensitivity of the output derivatives required. The design allows for the robust fault detection and isolation also for some nonlinear systems. The robustness of the approach is demonstrated on a nonlinear model of a four-tank process. Copyright (C) 2021 The Authors.
A model-based method for the detection and estimation of faults in dynamic systems is proposed. The method is based on the combination of the parity space approach and the modulating function framework for estimation....
详细信息
A model-based method for the detection and estimation of faults in dynamic systems is proposed. The method is based on the combination of the parity space approach and the modulating function framework for estimation. The parity space method is employed as an efficient geometric procedure determining null subspaces for annihilating unknown terms and formulating residuals. With the modulatingfunctions technique the dynamic relation from output differentiation is reformulated as an algebraic expression. This substantially reduces the noise sensitivity of the output derivatives required. The design allows for the robust fault detection and isolation also for some nonlinear systems. The robustness of the approach is demonstrated on a nonlinear model of a four-tank process.
暂无评论