Abstract: According to characteristics of wind turbine and water pump, a new scheme and its control strategy for wind power aided pumping water machine were proposed. The problems such as maximum power point tracking of wind energy and power balance between wind turbine and diesel engine need to be solved, so the system proposed is a coupling two input two output time-variable system. Based on the principle of decoupling and recurrent neural network, two Diagonal Recurrent Neural Network (DRNN) was adopted to adjust the parameters of two PID controllers online, and one nerve cell decouple compensator was adopted for decoupling the system, thus, non-model adaptive decouple PID control was implemented. Finally, the validity of the control strategy was proved via computer simulations, which can provide a foundation for further study.