洛阳PTA装置基于径基函数网络的4-CBA软测量

4-CBA soft sensor of Luoyang PTA plant based on Radial Basis Function Network

采用一种具有强非线性表达能力的径基函数网络方法建立洛阳PTA装置4-CBA含量软测量模型。在实际应用中,4-CBA含量软测量模型具有良好的预测精度,相对误差仅为4.3%。基于该软测量模型系统分析主要工艺操作参数对TA产品中4-CBA含量的影响,为生产操作条件的优化提供指导。

Radical basis function network, which can describe the high nonlinear system ,was employed to develop the soft sensor model of 4-CBA. In pratice application,the 4-CBA content soft sensor model has better forecast precision, relative error is only 4.3%. Further, based on the obtained soft sensor, the effect of the main processing parameters on the 4-CBA content in TA product was analyzed. It provided guidance for the optimization of processing parameters.