Robust wide-range control of power plants for life extension and performance enhancement

This dissertation presents a complete methodology for life extending control system (LECS) synthesis for fossil fuel steam power plants with the objectives of performance enhancement, structural durability and life extension. The proposed LECS has a two-tier architecture. The lower tier consists of a feedforward control policy and a family of linear multivariable robust controllers which are gain-scheduled. The optimal feedforward policy is formulated on the principle of nonlinear programming. The sampled-data feedback control laws are synthesized based on an induced L{dollar}sb2{dollar}-norm technique which minimizes the worst case gain between the energy of the exogenous inputs and the energy of the regulated outputs. The supervisory controller at the upper level makes decisions based on trade-off between performance enhancement and life-extension. The supervisory controller is synthesized based on approximate reasoning embedded with rule based expert knowledge of the power plant and structural damage models. Using the fuzzy logic, the plant operation strategy is modified on-line for trade off between plant performance and structural damage in critical components. The fuzzy algorithm facilitates bumpless controller switching for gain scheduling, under wide range operation and control.; The LECS has been tested by simulation experiments on a generic fossil fuel power plant model with the generation capacity of 525 MW. The plant is represented by a 27 state, 4 input, 4 output nonlinear model. This research explores the service life of three critical components, namely the main steam header, radiant superheater and hot reheat header. Developing structural damage models for these three components is an integral part of the research. Knowledge of these models is used in feedforward optimization, robust feedback synthesis and supervisory controller development. The models are also used to estimate accumulated structural damage during plant operation.; A distinct feature of the research work reported here is that a thorough knowledge of the power plant dynamics and the mechanics of the structural materials is applied to formulate a novel controller design tool. This tool can be readily used by practicing engineers with the aid of commercially available software.