Abstract: Melt index (MI) is considered as one of the important quality variables of propylene (PP) polymerization, which determines the products specifications. Thus, a reliable estimation of MI is crucial in quality control. Addressing the deficiency of the particle swarm optimization (PSO) algorithm, whose particles are easy to sink into premature convergence and run into local optimization in the iterative process,this article introduces the selection strategy,the immune clone PSO(ICPSO) algorithm,which makes the particles of ICPSO maintain the diversity during the iterative process so as to overcome the premature convergence of PSO. The ICPSO was used to optimize the parameters of weighted least support vector machine (WLSSVM) to predict the melt index of polypropylene,so the optimized model ICPSO_WLSSVM was obtained. Researches on the optimized model were illustrated with the real plant of propylene polymerization, and the results showed that the proposed approach had great prediction accuracy and validity.

Key words: immune clone particle swarm optimization, diversity, support vector machine, melt index prediction, parameters optimization

摘要： 熔融指数（MI）是聚丙烯生产的重要指标，建立可靠的熔融指数预报模型非常重要。针对标准粒子群算法(PSO)在迭代过程中易出现粒子过早收敛而陷入局部最优的缺陷，通过引入免疫系统的抗体选择机制，构造了一种基于免疫机制的免疫粒子群优化算法(ICPSO)，来保持更新粒子的多样性，从而克服标准粒子群算法过早收敛的缺陷；然后利用ICPSO方法对鲁棒最小二乘支持向量机预报模型(WLSSVM)进行参数寻优，得到最优的ICPSO_WLSSVM预报模型。以实际聚丙烯生产的熔融指数预报作为实例进行研究，结果表明所提出的ICPSO_WLSSVM模型的有效性和良好的预报精度。

关键词: 免疫粒子群优化, 多样性, 支持向量机, 熔融指数预报, 参数寻优