In this thesis an adaptive observer is designed and a pole assignment technique is applied in order to accomplish a satisfactory automatic control of steam generators from zero to full power.
The change of the water level of a steam generator is caused by three effects; mass capacity, swelling and shrinking, and mechanical oscillations. The knowledge of the water level contributions caused by these effects will be helpful in controlling the water level. The state observer is designed in order to use state feedback. The obvious result of introducing state feedback is to change the undesired open-loop system into the desired overall closed-loop system. Since a nuclear steam generator is controllable and observable, it is possible to design a state observer. Also, since a steam generator is subject to parameter variation according to the change of operating conditions, an adaptive observer must be used. The adaptive observer estimates the parameters and states of the steam generator simultaneously. A fourth-order linear model is presented and, on the basis of this model, an adaptive observer is designed.
Since an implicit-type adaptive observer is applied, a state reconstruction process and a parameter adaptation one are separated and system inputs and outputs are unnecessary to be bounded for the stability. The time-varying problem of the steam generator is resolved by estimating at every time step the parameters which change according to the operating conditions. A pole assignment controller is derived on the basis of the adaptive observer. The characteristics of the overall closed-loop control system can be expressed in terms of its assigned poles. The troublesome tuning procedure of the conventional P-I controller is reduced to the determination of the desired poles only. The proposed algorithm is compared with the P-I controller through numerical simulations.
Also, the adaptive pole assignment controller is studied experimentally by implementing it to the ...