PSO优化灰色建模在工艺参数预测中的应用  

PSO Optimization Grey Modeling and Its Application in Forecast of Manufacturing Process Parameters

作　　者：张宇飞[1,2] 宋超[2] 党志燕[2] 

机构地区：[1]天津大学自动化学院,天津300071 [2]天津航海仪器研究所,天津300131

出　　处：《自动化与仪表》2009年第5期28-31,共4页Automation & Instrumentation

摘　　要：在对波动幅度较大的工艺参数建立灰色预测模型时,采用常规的背景值构造会有较大的预测误差。针对该问题,文中提出利用粒子群算法(PSO)进行自适应灰色预测建模的方法,该方法通过对预测模型背景值的在线优化,能有效提高预测精度。将该方法应用于中药滴制剂工艺参数的预测中,预测结果表明了方法的有效性。When grey modeling method is applied for forecast of large oscillation process parameters, the forecast error will be large if the standard background value is adopted. So an adaptive grey forecast modeling method is supposed, which optimizes background value by particle swarm optimization （PSO） and which can improve forecast precision. The method is effective when it is applied for modeling of chinese materia medica pilule manufacturing process parameters forecast modeling.

关 键 词：灰色预测 粒子群优化算法 背景值优化 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]