Abstract-- Mostly during wide-range operation, load-following capability and efficacy for frequency regulation of fossil-fuel power plants may be affected by interaction among the control loops, caused by the non-linear coupled plant dynamics. This paper introduces a fuzzy gain-scheduling decoupling control scheme to improve plant response under large power excursions throughout the power plant operating space. The control scheme consists of single-loop PID controllers in series with an inverse interaction compensator. Both, the controllers and compensator are gain-scheduled with fuzzy systems to embrace the entire operating space. The proposed control scheme is evaluated through simulation experiments. Results show improved wide- range operation. Index Terms-- Fossil-fuel power plants, control loop interaction, decoupling control, fuzzy PID, gain-scheduling, interaction compensation, fuzzy systems. I. INTRODUCTION OST control systems at fossil-fuel power plants consist of decentralized schemes with multiple PID-based loops. These schemes were designed under the assumption that the power plant might be separated into several single-input- single-output independent processes. The efficacy of this approach to regulate the power plant response around a fixed operating point has been abundantly proven. However, such an assumption is only valid for normal operation at base load, where the plant characteristics are nearly linear and non- interacting. These conditions do not hold throughout the power plant operating space. Hence, traditional control schemes may decrease process performance due to the nonlinear coupled plant dynamics, which change with the point of operation, and might not be acceptable for wide-range load-following operation. Also, to enhance power plant operation, decentralized control schemes have been augmented with compensation schemes to diminish the undesired effects of control loop interaction [1,2]. Nevertheless, most compensation schemes are intended for interaction compensation around definite points of operation, but for wide-range operation [3]. This paper introduces a decoupling control scheme good for wide-range operation. The scheme consists of decentralized single-loop PID controllers in series R. Garduno-Ramirez is with the Division of Control Systems, Electrical Research Institute (IIE), Cuernavaca, Morelos 62490 Mexico (e-mail: rgarduno@iie.org.mx). K. Y. Lee is with the Department of Electrical and Computer Engineering, Baylor University, Waco, TX 76798-7356 USA (e-mail: kwanglee@psu.edu). with an interaction compensator, whose parameters are updated on-line via gain-scheduling strategies based on fuzzy systems. Each decentralized PID controller is a fuzzy system that conveys a PID-based multimode scheme in the form of a gain-scheduling controller [4]. The structure of the interaction compensator is that of an inverted decoupling scheme, where compensation factors are introduced among the controllers’ output signals; preserving the control paths of the uncompensated decentralized control scheme [5]. Factors are determined off-line, from an equivalent process gain matrix, and updated on-line by fuzzy systems along any desired power-pressure operating policy. Section II introduces the configuration of the fuzzy decoupling controller. Then, Section III describes the fuzzy gain-scheduling PID controllers and Section IV the fuzzy gain- scheduling interaction compensator. Section V presents some simulation results to show the controller performance. Finally, Section VI draws a few conclusions. II. FUZZY DECOUPLING CONTROL SCHEME A. Fuzzy Gain-Scheduling Control Scheme In a fossil-fuel power plant, the turbo-generator uses the thermal energy generated and stored in the boiler to produce electric energy. The boiler and turbine controls adjust the fuel firing rate and the steam flow according to the required power. In a Coordinated Control (CC) scheme, the unit load demand is routed simultaneously to both the boiler and the turbine controls. The control signal for the fuel/air valves is provided by the load controller from the unit load demand and the measured power, while the control signal for the steam throttle valve is provided by the pressure controller from the measured throttle steam pressure and the pressure set-point, which is obtained from the unit load demand through a non-linear mapping (Fig. 1). The given non-linear mapping is called the power-pressure operating policy, which is normally set by the power plant manufacturer. The proposed fuzzy decoupling control scheme updates the conventional CC strategy in three major steps (Fig. 2). First, the typical two-loop power-pressure CC strategy is augmented with the drum water level control loop to achieve internal energy-balanced operation at any load in the entire range of operation. This is due to the fact that in a drum-type power plant, electric power and steam pressure are tightly coupled and heavily affected by the fuel/air flow and the steam flow, whilst the feedwater flow slightly affects power and pressure, Fuzzy Gain-Scheduling PID+Decoupling Control for Power Plant Wide-Range Operation R. Garduno-Ramirez, Senior Member, IEEE, and K. Y. Lee, Fellow, IEEE M The 14th International Conference on Intelligent System Applications to Power Systems, ISAP 2007 November 4 - 8, 2007, Kaohsiung, Taiwan 233