Research On Variable Pitch Control Strategies Of Large-scale Wind Turbine

In order to achieve sustainable development, clean energy is being vigorously developed in China. Thus, wind power industry is developed rapidly. Wind power technology is the core of the wind power industry, which power control and reducing the imbalance loads are the key technologies of large scale wind turbine. In this PhD thesis, variable pitch control strategies of the domestic large scale wind turbine (SUT-3000) are studied. On the basis of analysis of the basic laws of pitch control, the strategies and implementation methods of collective pitch control (CPC) for power control, and individual pitch control (IPC) for reducing imbalance loads are proposed.First of all, for the problem of the output power stability of wind turbine system is affected by the random uncertainties of wind speed, the self-tuning PID collective pitch control based on radial basis function (RBF) neural network is proposed. The mathematical models of the wind turbine and pitch mechanism are established. In the case of the wind speed is over its rated value, the parameters PID variable pitch controller are adjusted in real time by the RBF neural network. The blade pitch angles are adjusted according to the changes of wind speed. Accordingly, the generator output power is adjusted so that the output power of wind turbines is stable at the neighborhood of its rated value. Simulation results show that, Compared with the conventional PID variable pitch control, the self-tuning PID variable pitch control based on RBF neural network make the output power more stable with small overshoot and good robustness.Secondly, for the problems of the traditional individual pitch control can not reduce the load at high frequency, and the complexity of modeling of wind speed, multivariable LQG optimal individual pitch control strategy is proposed based on feedforward-feedback structure. The multivariable controller in this thesis is composed by the optimal linear quadratic Gaussian (LQG) controller and the feedforward disturbance compensation controller based on the wind speed signal estimation. The feedforward controller focuses on compensation for the impact of the low frequency components of wind speed to wheel torque. A simple and effective random walk model is used to estimate wind speed, to avoid the establishment of complex wind speed model. The tilt and yaw direction components of effective wind speed are approximated by the random walk model. This control method is compared with the traditional individual pitch control methods with different models by simulation. The simulation results shows that, two individual pitch control can reduce the impact of imbalance loads. But the multiple LQG optimal individual pitch control strategy is better than the other one and it can reduce the loads with high-frequency. It is more suitable for the pitch control of large-scale wind turbine.Finally, in order to avoid the technology, cost and safety issues brought by the real machine experiment, the design and experiment of 3 MW electric variable pitch loop digital comprehensive simulation platform are accomplished in this thesis. In loop digital comprehensive simulation platform, the wind turbine operation control system, electric variable pitch actuator, sensor system are real machine, while others like the wind wheel, wind tower, generators and so on are simulated by simulation software. Thus the whole loop simulation platform is equivalent to a variable pitch wind turbine. The effectiveness of collective pitch control and individual pitch control strategies is verified on the loop simulation platform. Experimental results show that the designed control scheme can meet the control requirements; the theory and simulation are correct, feasible and effective.