机构地区: 东莞市清洁生产科技中心广东东莞523808
出 处: 《节能技术》 2009年第3期209-211,共3页
摘 要: 准确地预测聚丙烯的产品质量,对加强产品质量的控制和节省能源,具有十分重要的现实意义。聚丙烯产品质量的控制,是受反应时间和升温速率等多因素共同影响的非线性问题。本文模型将聚丙烯聚合前期的升温速率或釜内压力变化速率,作为神经网络输入参数,研究了聚合前期温度(压力)与熔融指数的关系。仿真结果表明,建立的小波神经网络网络模型推广能力好,效果比较令人满意,将为下一步对生产装置的改造和控制提供指导。 Precisely predicting the quality of polypropylene(PP), which can guide production, enhance quality control and save energy, is of great importance in the polymerization of propylene. The control of PP quality is a kind of nonlinear problem, related to many factors such as polymerization time, rising temperate rate and rising pressure rate, which could be the input of wavelet neural network or BP neural network to study correlation of melt index with temperature or pressure. Simulation demonstrates the temperature behaves better predicting performance than the pressure as the network input. The model can provide a good guide for the following industrial control and equipment update.
关 键 词: 液相本体法 升温速率 升压速率 小波神经网络 熔融指数
领 域: [电气工程]