Currently, environmental and economic concerns have pressed the power generation industry to develop more efficient and clean ways of generating electricity power. Facing such environmental and economic pressures, advanced control technology plays key role for both fossil fueled power plants and renewable energy systems. In this work, nonlinear control problems are studied for a boiler-turbine unit and wind turbine systems.
Advanced control is crucial for safe and efficient operations of power plants, especially in the presence of fast and large load changes. We study the control problems of load changes for a 160MW boiler-turbine unit. Two schemes are proposed to effectively control the boiler-turbine unit that has various constraints on the system states, outputs, control input and rate of control signals. By taking advantage of the nonlinearities of the boiler-turbine unit, we design a nonlinear state feedback controller. With a careful selection of the controller gains, the states and control inputs can be guaranteed to be within the required physical constraints. To further incorporate the constraints on the system outputs as well as control-input rates into feedback design, we utilize the moving horizon control strategy. The nonminimum phase behavior can be compensated by using a relatively long but computationally-affordable horizon length. The simulations of two control methods demonstrate the well-controlled performances of the boiler-turbine unit under large and fast load changes.
Modeling and variable speed control strategies for wind turbines are studied in order to capture maximum wind power. Wind turbines are modeled as two-mass drive-train system. Based on the obtained wind turbine models, variable speed control schemes are investigated for region 2. We designed nonlinear tracking controllers to achieve asymptotic tracking control for given rotor speed reference signals so as to yield maximum wind power capture. Due to the difficulty of torsional angle measurement, we design observed based control with using rotor speed information only to fulfill asymptotic tracking control. The proposed schemes are shown to be able to achieve smooth and asymptotic tracking and illustrated by the given simulation results.