Nonlinear multivariable control of nuclear power plants based on the unknown-but-bounded disturbance model

A nonlinear control system that is based entirely on the time-domain representation of dynamic systems is proposed for the control of a simplified pressurized-water-reactor (PWR) nuclear power plant model. The initial stage consists of designing several linear control systems using plant models linearized at preselected operating points. A set-theoretic algorithm is used that explicitly treats the control, control rate, and state constraints. The control objective is to design a system that uses only the available control, at the available rate, in the presence of an input disturbance, without violating the prespecified state constraints. The final stage of the design process uses the gain-scheduling technique to implement the linear control systems a global nonlinear control system. The final design is evaluated through transient response simulations using a simplified nonlinear model of a PWR-type nuclear power plant, with encouraging results