Water content ratio measurement with neural network based on simulated annealing

On-line density method is an effective measurement method of water content of crude oil, but the measurement results are vulnerable to uncertainties factors in practice. In order to improve the accuracy of water content ratio and stability, this paper proposes the use of back-propagation neural network annealing algorithm in computation. The error prediction is thus amended in the calculated value of density method model. Experimental results show that, this method effectively improves the on-line rapid determination accurate results of the water content ratio by the training of the off-line experimental data.