Tuning of power plant controls

Before a controller can be implemented, procedures for field tuning a control system must be developed. Such a procedure should lend itself to implementation as a digital algorithm and provide a reduction in the time required for plant commissioning. It must be reliable and easy to use. This paper presents the results of a preliminary study to develop such a tuning procedure for a candidate multivariable control system that was developed for load following and frequency regulation on an Advanced Boiling Water Reactor (ABWR). The candidate multivariable controller is for normal ABWR maneuvering in the electric utility system while maintaining adequate scram margins; reactor safety is not involved. Using the tuning procedure in this paper, twelve controller parameters were simultaneously adjusted such that the nonlinear ABWR Model achieved the desired response to a command input. The tuning procedures put forth in this paper can be applied to the control systems present in any fossil fuel power plant. The ABWR was chosen as a convenient demonstration vehicle. In Section 2 a frozen parameter tuning procedure is developed. A least squares algorithm is used on a block of data to calculate the parameter updates. Section 3 presents a brief description of the nonlinear model of the ABWR. The results of the simulation study with and without process noise are presented in Section 4