Aimed at the minimal energy consumption, the mathematical model for adjustable-blade pumps or pumping stations was established. Using the method of genetic algorithms and Matlab, the data of Jiangdu drainage and irrigation station were studied, the results showed that 3.39% energy could be save comparing with its actual energy consumed in the same year. Because the flowrate and head of pump station was changing frequently and the conventional optimal calculation method often takes long time, the author proposed artificial neural networks model to predict movement patterns of the Jiangdu first pumping station, the result of optimal operation is used as sample. By combination with the genetic algorithms and artificial neural networks, the satisfactory result can be obtained. The new method is not only simple and easy to implement, but also has high accuracy and reliability, it is an effective way for solving the optimization problem in pump station.