ABSTRACT
The emulsion matrix viscosity determines the quality of an emulsion explosive. However, in the production process of the emulsion explosive, the emulsion matrix viscosity cannot be measured directly. In order to solve this problem, a soft measuring method based on a BP neural network is proposed in this paper. However, the simple BP neural network cannot accurately measure the viscosity of the emulsion matrix. So, a Genetic Algorithm (GA) is used to optimise the network and then train the network prediction model to get the optimal solution. The predicted results with the predictions of the conventional BP neural network prediction model are compared. The BP neural network optimised by the GA can obtain more accurate results in a shorter period of time.
