A2/O process is one of the major processes in municipal waste water treatment

but many parameters affect the operation effect of A2/O process. And these parameters interact with each other

affecting the efficiency of the process. In order to make up the insufficience of single variable control method or orthogonal designing method

it establishes the neural network model (GA-ANN model) based on genetic algorithm. The model has been applied to an urban waste water treatment plant by A2/O process optimization. During the commissioning operation of the plant

it has obtained 154 effective monitoring data

and 2/3 of the data has been randomly selected for the GA-ANN model

and 1/3 of the data has been used for the model test. The process parameters have been optimized and get the best combination of operating parameters. The results show that it is feasible to establish the neural network model based on the genetic algorithm for the optimization of the A2/O process operation parameters. It can provide the theoretical reference for setting operation parameter of the waste water treatment plant. And it is also helpful to the practical production and application for adjustment and improvement of the operation efficiency.