Based on the new type of wind turbine drive train, analyses of the key parts were carried out, and the mathematical models were established. Simulation models of wind speed, wind turbine aerodynamics, hydraulic torque converter and spindle speed were established. Including the neural network control model, all models were integrated to establish the whole drive train simulation model. Then, the calculation and simulation were carried out. Based on the maximum wind energy capture, considered the structure parameters of the drive train, the guide vane opening of the hydraulic torque converter was controlled by the neural network to adjust the working oil circulation flow. Then, the torque and the speed of the converter were changed. Through planetary gears, the spindle before the generator was drived by the output of the converter and the output of the wind rotor together, to make the spindle speed stabilized at the design value.
