Soft Sensor Modeling for Oxygen-Content in Flue Gasses in 1000MW Ultra-superficial Units

Abstract

Ultra-supercritical unit, which can implement clean coal combustion and improve energy efficiency, is an important trend of thermal power plants in China. Aiming to the measurement of oxygen-content in flue gasses in Ultra-supercritical unit in a power plant, this paper discusses a soft-sensing model method based on Gaussian process regression (GPR). Then GPR based soft sensor is applied to estimate the Oxygen-content in Flue Gasses in 1000MW Ultra-superficial Units. The experiment results show that the method of soft-sensing based on Gaussian process regression is not only easy to implement, but also has small predicted error and uncertainty.