基于群算法的过程参量聚类研究

Clustering research of process parameters based on particle swarm optimization

针对复杂过程的参量聚类问题,提出一种基于粒子群优化算法的聚类方法,阐述了聚类算法的基本思路。通过对过程煅烧温度和煅烧转速二维数据的聚类仿真研究,证明该算法在类似过程参量聚类中的实用性能。对粒子群优化算法的聚类特性及参数设置进行了详细的分析,并将其与前期人工免疫聚类结果进行对比,提出了算法的改进方案。

The paper adopts the Particle Swarm Optimization （PSO） to solve the parameters clustering problem of complex processes. The basic mechanism of PSO is presented in the paper. The clustering simulation on tempera- tures and rotation speeds of the calcination process verifies the practicability of PSO in parameters clustering of sim- ilar complex processes. The clustering features and parameters setting of PSO are discussed in detail. Combined with artificial immune, some improved methods are brought forward to achieve better performances.