Abstract: The control of 3-D temperature distribution in a utility boiler
furnace is essential for the safe, economic and clean operation of pcfired
furnace with multi-burner system. The development of the visualization of 3-D
temperature distributions in pc-fired furnaces makes it possible for a new
combustion control strategy directly with the fumacs temperature as its goal to
improve the control quality for the combustion processes. Studied in this paper
is such a new strategy that the whole furnace is divided into several parts in
the vertical direction, and the average temperature and its bias from the
center in every cross section can be extracted from the visualization results
of the 3-D temperature distributions. In the simulation stage, a computational
fluid dynamics (CFD) code served to calculate the 3-D temperature distributions
in a furnace, then a linear model was set up to relate the features of the
temperature distributions with the input of the combustion processes, such as
the flow rates of fuel and air fed into the furnaces through all the burners.
The adaptive genetic algorithm was adopted to find the optimal combination of
the whole input parameters which ensure to form an optimal 3-D temperature
field in the furnace desired for the operation of boiler. Simulation results
showed that the strategy could soon find the factors making the temperature
distribution apart from the optimal state and give correct adjusting
suggestions.