首页 | 本学科首页   官方微博 | 高级检索  
     

基于BP网络模型的水合肼合成收率预测及参数优化
引用本文:邱添,王煤,许峰,薄向利,付永宽.基于BP网络模型的水合肼合成收率预测及参数优化[J].四川化工,2007,10(1):20-23.
作者姓名:邱添  王煤  许峰  薄向利  付永宽
作者单位:1. 四川大学化工学院,成都,610065
2. 宜宾天原化工有限公司,四川宜宾,644000
摘    要:采用BP算法建立了水合肼合成收率的神经网络模型,所建网络有良好的回想和泛化性能,对水合肼合成收率预测的平均相对误差小于0.5%。运用网络模型考察了反应温度、物料配比、次氯酸钠流量等操作参数对水合肼合成收率的影响,同时确定了获得较高合成收率时各操作参数的优化值。

关 键 词:神经网络  BP算法  水合肼  合成  建模  收率  参数优化

Analysis of Yield and Optimization of Parameters Based on BP Neural Network for Hydrazine hydrate Synthesis
Qiu Tian,Wang Mei,Xu Feng,Bo Xiang-li,Fu Yong-kuan.Analysis of Yield and Optimization of Parameters Based on BP Neural Network for Hydrazine hydrate Synthesis[J].sichuan chemical industry,2007,10(1):20-23.
Authors:Qiu Tian  Wang Mei  Xu Feng  Bo Xiang-li  Fu Yong-kuan
Abstract:A back-propagation(BP) neural network model is established to predict the yield and optimise the operation parametersfor Hydrazine hydratesynthesis.The results show that the model based on operation data can predict the yield accurately with the average relative error is less than 0.5%.The effects of reaction temperature and mixture ratios on the yield are discussed and the optimum values of operation parameters for high yield of Hydrazine hydratesynthesis are obtained.
Keywords:neural network  BP algorithm  Hydrazine hydrate  modeling  
本文献已被 CNKI 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号