聚合物混合质量的智能控制

2002年8月Aug., 2(X12维晋资讯 2002年8月Aug., 2(X122002年8月Aug., 2(X12第6卷第：J期Vol.6 No.3运筹学学报OR TrmsactimsThe Intellect Control of thm Polymer Mixtuie QualityQin LIAO Deiiua HE(Dept, of Applied Math, South China (Jniversity of Tehnoioj Guangzhou, 5J060, Chtna)Abstrac tThe filler mixed quality in polymer directly influences the polymer product quality. That thv filler mixed quality should be controlled and optimized online have breonic the quality key of polymer product, n this paper a filler quality grade recognition mod(?J based on quotas of the size and locMion distribution of Bllet granules area is huiJt. The information of the filler mixing quality in polymer are obtained by means of the processing techniques of computer and digital image, and the optimuni control model is established about the parameter s of process mixed. The results of simulation test show the recognition and control model of polymer filler are reasonable and better than other the instruments used at recently.Key words： polymer, mixed quality, neural network, image process. inteJligcnt recognition聚合物混合质量的智能控制2002年8月Aug., 2(X122002年8月Aug., 2(X12（华南理工大学应用数学系，广州，510640）M 要：聚合物与填料的混合质議，点接影响蠻合物成品质混合质量的在线控制 并优化控制参数已成为保证聚合物曲品质童的关键.本文首先痙立基于填料颗粒面积与位 輕分布为判断指标的填料尊级识别模型.并借助计算机与数字图形处理技术获取填料混合 质量的信息，在此基础上建立了聚合物混合参数的优化控制榄型.仿真检验的结果表明. 模型趣合理的且与投人使用的同类仪器相比具有更满意的结果.关词：集合物，混合质量，神经两络,.图像处理.智能识别.1. IntfoductioiiThe physics mechanical performance of polymer product is improved generally by means of filling material such as carbon black, CaCQj in polymer in the real application. The mixture quality influences directly the polymer product quality. One of ttie indices forlleccived; November 5, 2001 -