Robust Decoupled P-Q Controller for Doubly Fed Induction Generator

Researchers at McGill University have developed a novel controller for Doubly Fed Induction Generators (DFIG) which are widely used in wind turbines. This controller allows for a robust decoupled sensorless control. This approach will enable continually optimal performance of the DFIG without the need for the operator to update the parameters of the unit. This approach can be easily implemented with DSPs and FGPAs and may also be applied to motor controls. Applications: As world energy consumption grows nations are faced with difficult and costly decisions to satisfy its needs. With the success in Denmark where nearly 20% of electricity is generated by wind turbines many countries are now planning to deploy large scale wind farms. The current cost of these systems is approximately $1500 Canadian per kilowatt of generating capacity with a 2 Megawatt wind turbine costing in the order of $3 Million Canadian. The resulting electricity would be produced at a cost between $0.05 and $0.10 per kilowatt-hour depending on site conditions and amortization of the installation costs a figure higher than current production methods. However as fossil fuel costs rise and as feasible hydro electric projects are on the decline wind energy is increasingly attractive and will encourage additional development driving the cost of production even lower. In addition to the increasingly competitive cost of wind energy countries are investing in wind farms to reduce dependency on energy sources from politically unstable regions. To achieve decoupled P-Q control requires the rotor position to be known precisely. The position can be measured by an absolute position encoder which besides being expensive requires mechanical maintenance. Further the existing approaches require the parameters of the DFIGs to be known. The parameters have to be measured or obtained from the manufacturer. Even when the parameters are known their characteristics drift over time or are affected by operating conditions for example electrical resistance changes due to temperature fluctuations or magnetic saturation. The proposed solution applies a sensorless approach by which the position can be estimated by electrical measurements of voltages and currents and does not require prior knowledge of the parameters of the DFIG. By eliminating the need for measurements in a noisy environment more accurate control can be achieved. Further as conditions change over time the operation of the controller would be continually optimized without the need for an expensive field visit to re-set the parameters.