Selection of sensors and actuators based on a sufficient condition for robust performance

A new method for selecting sensors and actuators for linear control system applications is proposed. The goal is to eliminate actuator/sensor combinations for which a controller achieving a desired robust stability or robust performance level cannot be designed. To account for the structured uncertainty block, the plants for all candidate combinations are extended with certain scaling filters. Each combination is subjected to six conditions for existence of a stabilizing controller meeting a required H_∞ norm bound. However, combinations may be incorrectly rejected, since optimality of the scaling filters is not guaranteed. An active suspension control problem is used as an illustrative example.