Study On Soft Sensing Of Oxygen Content In Flue Gases And Optimization Of Coal Blending Based On Artificial Intelligence

Soft sensing techniques provide estimate for engineering variables which are difficult to measure by instrument directly. Its basic principle is to use variable that is easy to measure to establish the mathematical model to deduce the measured one.Based on the actual operating data in DCS system test of Huaneng power plants in 2006, a new method for measuring oxygen content in flue gases was proposed. The proposed method involves the factor analysis regressive model, partial least-square regressive model as well as BP neural network model. Fifty operating parameters of boiler, such as coal parameters, coal feed rate, air quantity, steam flow, exhaust gas temperature, burner tilt angle, etc. were treated as known parameters. Comparative analysis was made and a coupled model for predicting oxygen content in flue gases was suggested, and depress the dimension of the neural network by extracting the factors in PLSR. The simulation results were in good agreement with the experiment and the biggest generalized error is 5 %.As a kind of clean coal combustion technology suitable for China, power coal blending has gained much attention. The Coal Blending model in this article has two major sides, the author calculate the water content, volatile content as well as cost in different coal blending ratio with the assumption that the properties of the mixed coal could be obtained by weight-averaging each component. Based on these foundational work, adopt the genetic algorithms optimal model to compute the optimal coal blending ratio, comprehensive cost and coal consumption as optimization variable, object and fitness function respectively. Comparing to the traditional optimal model, the affections of properties of mixed coal to consumption have been considered in this model proposed. These simulations and analysis provided a theoretical foundation for economic performance's online monitor, consumption control, operation guide as well as production management, which is benefit to improve the production and management modernization level of power plants.