Dissolved oxygen concentration prediction control through multiobjective evolutionary RBF neural network

Through analyzing dissolved oxygen online control methods, a new prediction control model method was presented in this paper. The method is better than online control method in response to actual situation. In order to reduce error between actual situation and prediction result, multiobjective evolutionary RBF neural network optimized method was adopted. Real wastewater plant data was applied to the model simulation, the simulation shows that multiobjective evolutionary RBF neural network is better than other two neural network methods in certain situation control. The new method is a good way to dissolved oxgen concentration control.