摘要
介绍了某聚酯生产过程酯化工艺建立的过程质量指标酯化率的软测量模型。提出一种利用减法聚类产生初始的T-S模糊模型,通过粗调与细调聚类半径优化模糊模型的方法。建模前选择或计算出辅助变量,对样本数据进行了误差剔除与归一化处理。仿真结果表明,该方法建模速度快,模型泛化性能良好,为酯化率的估计提供了一种有效方法。
In this paper, a soft-sensing model of ester rate is established through certain PET process in a polyester plant. Ester rate is the quality index of PET process. A modeling method is presented, in which subtractive clustering is used to generate an initial T- S fuzzy model, then the fuzzy model is optimized by coarse-tuning and fine-tuning the radius of clustering. Before modeling, the supplementary variables are selected or calculated, and the error of data sample is eliminated and the data is normalized, The simulation results show that the model can be built fast and offers perfect generalization performing, it is an effective method to estimate ester rate.
出处
《自动化仪表》
CAS
2005年第10期13-16,共4页
Process Automation Instrumentation
