欢迎您！东篱公司
退出
-
申报数据库
-
立项数据库
-
成果数据库
-
项目获奖数据库
中文摘要：
针对田纳西—伊斯曼（Tennessee-Eastman,TE）过程具有的高度非线性、时变及多个操作模式等特征,为在线预测该过程产品流道中各种成分的含量,提出一种基于局部加权偏最小二乘的多模型建模方法。多模型建模方法首先要进行子模型的划分,将TE过程各种操作模式下的训练数据放入不同数据库中,利用贝叶斯分类器对在线测得的数据进行分类;然后采用即时（just-in-time,JIT）建模思想,基于局部加权偏最小二乘建立相应的在线局部模型;最后,将贝叶斯分类器得到的测试数据属于各个数据库的后验概率作为加权系数,对得到的局部模型的预测结果进行融合输出。基于TE化工过程仿真平台,采用该方法来预测产品流道中成分G和H的含量与真值基本一致,证明提出的基于局部加权偏最小二乘的在线多模型建模方法具有良好的预测效果。
英文摘要：
For the characteristics of high nonlinearity,time-varying and multiple operating modes in the Tennessee Eastman process,this paper adopted the multi-model modeling method based on locally weighted partial least squares to predict the content in product stream. The first step was to determine the local models under the frame work of multi-model method. Then this paper divided the training data from the TE process into different sub-databases and assigned the new coming data by Bayesian classifier. In the following step,it employed locally weighted partial least squares to establish the online local models using the idea of just-in-time. Finally,the output estimation was the combination of each local model with respect to the posterior probabilities. The simulation results based on the TE process show that the predicted value of component G and H are approximate to the true value. It illustrates the feasibility and efficiency of the proposed soft sensor.
同期刊论文项目
同项目期刊论文
Effectof temperature on Chinese rice wine brewing with high concentration pre-steamedwhole sticky ri
期刊信息
位置：