Soft sensor modeling of AlCl3·6H2O content based on Powell-BP

The production period of the crystalline aluminium chloride is considerably long. However, the offline assay of AlCl
₃
·6H
₂
O content has large time delay. Thus soft sensor modeling is needed to analyze its content, and estimate the value to improve the product quality. The conventional back-propagation (BP) neural network training is easily trapped to the local minimum, To overcome this embarrassment, the Powell method is adopted to optimize the training algorithm, improve the method of extremum computing, and reduce the difficulty of network realization. Finally, the soft sensor model of AlCl
₃
·6H
₂
O is established. Test results show that, compared with conventional neural network algorithms, the prediction of Powell-BP is more accurate and the convergence time is shorter.