| Summary: | Small-scale Wind Energy Conversion Systems (WECS) are becoming an attractive option for distributed and renewable energy generation. In order to be affordable, WECS must have low capital and maintenance costs. This leads to the increasing penetration of Permanent Magnet Synchronous Generators (PMSG) operating at variable frequency with connections to the power grid through a rectifier, and grid-tie inverter. Because PMSGs lack brushes and can be directly coupled to wind turbines, the capital and maintenance costs are greatly reduced. A direct connection to the grid further reduces system costs by removing the requirement of large battery banks.
The loading produced by grid-tie inverters on the DC bus is different than more typical constant-current or constant-power loads. They are characterized by large input ripple currents at twice the inverter's grid frequency. These ripple currents are reflected through the DC bus into the PMSG causing increased heating in the stator, and ripple torques which lead to premature bearing failure and increased maintenance costs. To mitigate this problem, manufacturers typically add large amounts of capacitance on the DC bus to partially absorb these ripples at the expense of system size, cost, and reliability.
In this work, the effects of the grid-tie inverter load are explored using system behavioural models which provide insight into the low frequency behaviour of the PMSG, rectifier, DC bus, and inverter. The swinging bus concept is presented and analysed in the time and frequency domains. A control philosophy is developed which allows the DC bus to swing, thus removing the effects of the grid-tie inverter on the PMSG while keeping the DC bus capacitor small. A solution consisting of a Moving Average Filter (MAF) is presented as an integral part of the control strategy.
Full simulations of a complete system are developed and investigated to verify the ripple torque reduction technique. Finally, a prototype is developed and experimental results are presented for a 2.5kW PMSG turbine generator. The simulation and experimental results are compared to a traditional controller showing tangible improvements in ripple current and torque in the PMSG, while improving the dynamic response of the system. === Applied Science, Faculty of === Electrical and Computer Engineering, Department of === Graduate
|
|---|
