Static and Dynamic Control Reconfiguration

This report describes new control reconfiguration approaches after actuator faults. A new approach that uses a static block, which is designed based on a description of the plant input/outputbehaviour using the Markov parameters, is developed and analysed for stability and robustness. This approach pursues the output trajectory recovery of the faulty plant. The dual approach for sensor faults is sketched. A new goal referring to the stationary plant behaviour is introduced and used as a basis for static reconfiguration block design. The approaches initially developed for use with a static reconfiguration block are then used to obtain a new virtual actuator design method. The generalisation of the virtual actuator motivates the entire work. The use of the Markov parameter recovery approach for designing its feedforward block is found to be superior over the previous static direct reconfiguration approach in numerous practically relevant situations. All approaches are illustrated by numerical examples and verified by experimental studies. ii