Abstract:

This paper proposes a hybrid genetic algorithm （GA） to optimize the operation of pulverizing system in admixing combustion conditions. In this algorithm, a fuzzy neural network is adopted to evaluate the slag property of admixed coal, and a Radial Basis Function neural network is used to predict the fineness of pulverized coal and the outlet temperature of the mill. Then, the pulverizing system is optimized, with the mill safety and the furnace com- bustion stability as the restraints, and with the lowest blended coal price and pulverizing unit cost as the targets. Moreover, a kind of hierarchical gene is designed to fasten the searching process. Simulated results indicate that the proposed algorithm is practicable and feasible in optimizing the operation of pulverizing system in admixing combus- tion conditions.

Key words: pulverizing system, hybrid genetic algorithm, fuzzy neural network