Abstract:The quantity and quality prediction model for lead-zinc sintering process was proposed, the corresponding mathematic optimization model was built, and the hybrid particle swarm optimization (PSO) was applied to optimize the mathematic model. The quantity optimization problem with quality-constraint was converted into a two-objective optimization problem, one of which is the objective function of quantity, the other is the degree function of constraint violation of quality. To perform the parallel optimization, the comparison used in tournament selection was employed to compare the searched solutions of the PSO. When the PSO stagnates, the line searching algorithm is executed with a certain probability, which executes the line searching in the direction of the difference between the current global optimum and the last global optimum, and the better searched solution is taken as the current global optimum. Thus it keeps the swarm particle algorithm alive. In order to decrease the fluctuation of sintering process, expert rules were established to modify the calculated optimization parameters, and the practical result shows that the optimization method can increase the quantity to a certain extent and ensure the quality of sinter.

Key words: lead; zinc; sinter; quality prediction model; hybrid particle swarm optimization algorithm; tournament selection