Research on the prediction of bauxite slurry composition based on deep learning

The preparation of bauxite slurry plays an important role in the alumina production process based on the Bayer process. Whether or not a slurry that meets the requirements of alumina production can be prepared will directly affect the technical and economic indicators of alumina dissolution rate, sedimentation, seed decomposition, and productivity. Taking into account the lag of the raw ore pulp composition test, in order to improve the output and quality of alumina and accurately predict the composition of the bauxite slurry, the author builds a gray radial basis basis (GM-RBF) neural network based on the principle of error compensation based on the batching mechanism. Composition prediction model. Take the alumina production data of Chinalco Guangxi Branch in the first quarter of 2020 as an example to verify the validity of the model. The results show that the relative error of the prediction using the GM model alone is 20.39%, and the relative error of the prediction using the GM-RBF model is 0.12%. The GM-RBF model improves the prediction accuracy.