This second part addresses the problem of controlling a continuous-time linear time-invariant plant with external point delays using a multi-estimation scheme. The plant is supposed to posses parametric uncertainty while the delay is supposed to be unknown but belonging to a known compact subset. Firstly, the continuous-time plant is discretized by using a ZOH. The discretization step allows converting a real-valued delay uncertainty in a discrete-type one. Thus, the estimation scheme is composed of a bank of estimation algorithms running in parallel, being each one associated to a discretization of the plant with a different nominal value of the discrete time delay. Then, a high-level supervisor selects the most appropriate model to be used to parameterize the adaptive controller at each time interval. In this way, a conceptually simple control design for the adaptive control of unknown plants with unknown external point delays is presented. Furthermore, the scheme has revealed to be potentially capable of estimating an approximation of the actual delay of the plant as simulation examples have pointed out. This is an important novelty since there is not effective method to estimate simultaneously the delay of the continuous plant and its parameters. ICIC International © 2008.
|Journal||International Journal of Innovative Computing, Information and Control|
|Publication status||Published - 1 Jun 2008|
- Adaptive control
- Digital control
- Time-delayed systems