Abstract
A well-known two-stage ac to dc converter composed of a diode bridge rectifier and a dc converter is widely used in the existing small wind power generation system (WPGS). However, due to the nonlinearity of the diode bridge rectifier, there will be significant total harmonics distortion (THD) in the generator currents which will not only reduce the efficiency but also result in additional mechanical stress and undesired acoustic noise. As to the maximum power point tracking (MPPT) controller, the widely adopted optimal torque control algorithm only concerns about the steady state characteristics of the wind turbine. However, the poor dynamic response due to the mechanical inertia effect will result in reduction of the wind turbine efficiency under rapidly changing wind speed situations. To overcome the above disadvantages, a novel sensorless dynamic MPPT controller and a high efficiency single-stage ac to dc converter are proposed in this dissertation as a compromise of both performance and cost for a small WPGS.Basically, the major contributions of this dissertation can be briefly outlined as follows. First, a novel sensorless dynamic optimal torque MPPT control with adjustable virtual inertia technique is proposed to improve the wind turbine dynamic response and increase the output power. In addition, since the proposed novel MPPT control contains no mechanical sensors, both the reliability and cost performance index can be further improved. Second, a single-stage converter with three active switches is proposed to replace the conventional two-stage converter for improving the power quality and efficiency. The generator current THD is greatly reduced to around 5% which can reduce the corresponding mechanical stress and acoustic noise as well. Third, a novel quasi-synchronous rectification technique is proposed to further reduce the conduction losses of the body diodes. Moreover, a hybrid control composed of partial CCM, partial DCM and QSR techniques is proposed for the corresponding single-stage converter to reduce the losses. From the experimental results, one can see that the total efficiency of the whole system can be increased by 12% to 15% and the payback period is reduced by 11% to 13% approximately.