废旧聚丙烯制板工艺预测模型的研究  

Research on Prediction Model Used to Board Preparation from Waste PP Material

作　　者：姚正军[1] 李超[1] 邹戈[1] 李莉平[1] 

机构地区：[1]南京航空航天大学材料科学与技术学院,江苏南京210016

出　　处：《高分子材料科学与工程》2007年第6期24-27,共4页Polymer Materials Science & Engineering

基　　金：校企横向预研课题(1006-264006);江苏省自然科学基金资助项目(BK2005128)

摘　　要：在研究基本误差前向反馈(BP)算法和模拟退火(SA)算法的基础上,提出了基于模拟退火优化的Resilient Back Propagation(RPROP)算法,并利用该算法建立了废旧聚丙烯制备板材重要工艺参数与成品板材的压制成形性能、力学性能之间的模拟及预测模型。与单纯的BP算法比较,数值计算结果表明,RPROP-SA混合优化策略具有较高的速度和精度。Based on the research of Back Propagation algorithm and Simulated Annealing algorithm, this paper proposed the Resilient Back propagation （RPROP） algorithm which is optimized by Simulated Annealing algorithm. Then, it was applied to establish the simulation and forecast model which is based on the relationship between the technological parameters of Board Preparation from Waste PP Material and the shaping property, mechanical properties of the board. Compared with single BP algorithm, the value calculation result demonstrates that RPROP-SA mixing and optimizing strategy has a higher speed and accuracy.

关 键 词：废旧聚丙烯 板材 预测模型 RPROP算法 SA算法 

分 类 号：TQ319[化学工程—高聚物工业]