Abstract:

Two models respectively describing the pulverizing unit cost and the coal fineness of the middle-storage pulverizing system are established by means of the least-square support vector machine , and the combinational genetic algorithm is adopted to ,seek the optimal operation parameters for the minimum unit consumption in different operating modes. The proposed model of coal fineness is then adopted to predict the coal fineness based on the optimized operation parameters , and the predicted results are fed back to control the optimization of the pulverizing unit cost. The hot test at a 50 MW unit in a power plant indicates that the proposed optimization control system based on the feedback control of coal fineness is of high dependability and practicability , which provides guidance for the optimization of the pulverizing system , improves the operation security and saves the cost.

Key words: pulverizing unit cost, coal fineness, least-square support vector machine, pulverizing system, optimization