ABSTRACT
This chapter reviews and discusses the main methods currently used to improve thermal power plant boiler efficiency. Using Neural Networks (NN), model predictive control, and direct search technologies, NeuCo’s CombustionOpt can determine the optimal fuel type and air set points for the specific goals and constraints and then make the necessary adjustments to fuel and air variables in real time. NN-based optimization technologies have been applied in United States power plant optimization demonstration projects to improve the fossil fuel power plant combustion process. Although NN-based technologies are employed to successfully solve problems in the power-generation industry, some problems still seriously impair the efficiency of heat transfer, degrading the performance of the boiler in a power plant. The NN algorithm learns the relationships between operating conditions, emissions, and performance parameters by training process and develops a highly complex nonlinear function which maps the system inputs to the corresponding outputs. The chapter also discusses existing thermal power plant efficiency problems.
