Selectivity index and separation efficiency prediction in industrial magnetic separation process using a hybrid neural genetic algorithm

Abstract It is essential to know the process efficiency in industrial magnetic separation under different operating conditions because it required control parameters optimize efficiency. To our knowledge, there no information about using artificial intelligence for modeling process. Hence, finding a robust and more accurate estimation method predicting selectivity index still necessary. In this regard, feed-forward neural network was developed predict index. This model trained present predictive based on percentage of iron, iron oxide sulfur mill feed cobber feed, 80% passing size plant capacity. Therefore, work aims develop an intelligent technique hybrid neural-genetic algorithm concentration Results indicated that values mean square error coefficient determination testing phase were obtained 0.635 0.86 4.646 0.84 efficiency, respectively. order improve performance network, genetic used weights biases network. The results with GA-ANN achieved by 0.276 0.95 1.782 0.92 other statistical criteria better than those ANN model.