Optimal Control for Aluminum Hydroxide Calcinations Based on Fuzzy Neural Networks

A logic analysis model for aluminum hydroxide calcinations process was proposed, which includes four kinds of parameters and two units, namely, feeding, operation, structure and index parameters, process control and combustion unit. Integrated fuzzy algorithm, artificial neural networks and genetic algorithms, a kind of fuzzy neural network was constructed and optimized. A strategy for calcinations adjustment that consists of five fuzzy neural network models and a fuzzy linguistic inference model was schemed, whose primary functions include feeding computing, fuel flux computing, the ratio of air to fuel computing, and flux tracking of air and fuel. The application of optimal control indicates that temperature error is ±5°C. Optimal control is beneficial to practical operation and energy-saving.