文献类型：期刊文献

中文题名：基于改进生物地理学算法的甲醇合成转化率软测量

英文题名：Soft Sensor of Conversion Rate in Methanol Synthesis Based on Improved Biogeography-Based Optimization

作者：陈绍武[1];宋淑群[2];张凌波[1];顾幸生[1]

机构：[1]华东理工大学化工过程先进控制与优化技术教育部重点实验室,上海200237;[2]兖矿集团国宏化工有限责任公司,山东邹城273500

年份：2014

卷号：26

期号：1

起止页码：17

中文期刊名：系统仿真学报

外文期刊名：Journal of System Simulation

收录：CSTPCD;;Scopus;北大核心:【北大核心2011】；CSCD:【CSCD2013_2014】；

基金：上海市科委基础研究重点项目(12JC1403400)

语种：中文

中文关键词：生物地理学优化算法;粒子群算法;甲醇转化率;BP神经网络

外文关键词：BBO; PSO; methanol conversion rate; BP neural network

摘要：甲醇是一种重要的工业原料,甲醇合成反应是甲醇生产中重要的环节,甲醇合成转化率是甲醇合成反应中的关键指标,以甲醇合成转化率为对象建立准确的软测量模型可以有效指导甲醇合成过程的控制。生物地理学优化(BBO)算法是一种新型智能优化算法,结合粒子群优化(PSO)算法提出一种改进的生物地理学优化算法,称为BBOPSO。在14个基准测试函数上,对改进的算法、基本BBO算法、基本PSO算法及其他两种算法进行比较,结果表明改进的算法优于所比较的其他算法。最后,用改进的算法优化BP神经网络的参数,获得了甲醇合成转化率模型,实验结果表明,基于改进算法的神经网络模型能够较好的预测甲醇合成转化率。
Methanol is a kind of important industrial raw material, and methanol synthesis reaction also has been an important part in the production of methanol. As a key indicator of methanol synthesis reaction, an accurate soft sensor model of conversion rate in methanol synthesis could guide the control of the methanol synthesis process effectively. Biogeography-Based Optimization （BBO） was a new intelligent algorithm, an improved BBO combined with Particle Swarm Optimization （PSO） was proposed, called it BBOPSO. Among 14 benchmark test functions, the improved algorithm, the original BBO, the original PSO and other two approaches were compared, and the proposed approach was proved to be better than the other approaches. Lastly, the improved algorithm was employed to optimize BP neural network parameters and obtain the soft sensor model for methanol conversion rate. The result indicates that BBOPSO-based neural network model can predict the methanol conversion rate well.