Prediction model of pellets quality based on BP neural network optimized by genetic algorithm

As the technics of Grate-Kiln is complicated, for which it's difficult to establish accurate model, a black-box prediction model of quality is established in this paper with compression strength, drum index and screening index being the output based on the method of BP Neural Network optimized by Genetic Algorithm. And then the parameters of model are identified by MATLAB using the real data of some pellets factory. And the result of simulating indicates that the model converges quickly and the hit-ratio is increased.