P10.2
A technique for wind turbine Radar Cross Section (RCS) characterization through electromagnetic models
Andrew Huston, Atmosphere Radar Research Center, Norman, OK; and Y. Zhang, S. Wang, F. Kong, and R. D. Palmer
Radar cross-section (RCS) modeling of wind turbine structures is one of the first steps toward predicting the impacts of wind farms on weather radar operations, as well as developing knowledge-based interference mitigation algorithms. The key challenges originate from the tremendous sizes of wind turbine structures compared to radar operating wavelengths, coupled with complex observation environments. Existing studies on wind turbine radar signatures are limited by simplified structure modeling (based on simple geometrical shapes), computing power, and empirical data interpretations. A physical scattering model will bring more insight into the impacts of detailed parts within turbine structures, and the interactions between wind turbines and terrain/atmosphere. On the other hand, the tradeoff between model accuracies and the computation loads shall be balanced.
A unique model-based approach was developed in the University of Oklahoma to address such challenges. This approach starts from fabricating a scaled model of actual wind turbine, and measuring its electromagnetic scattering parameters in a controlled laboratory environment. The scaled model contains internal controllers and sensors to track the precise dynamics (such as rotation speed and phase of the blades) of the structure in real time. Fabrication of such a wind-turbine model considers the blade aerodynamics, and thus contains higher fidelity than simple geometric models. Next, the associated CAD drawing of this model is sent to full-wave electromagnetic solvers (with pre-processing). Since the model and wavelengths have been appropriately scaled down, the full-wave solver is able to handle the size of problem. After matching between the solver's prediction and the laboratory measurement is confirmed, the results are transformed back to the real-world dimension through field/frequency scaling, near-far field transformation and interpolation. The biggest advantage of this approach is it provides a very low-cost mechanism to achieve reasonable precisions for scattering model validations. Also, since the model's orientations and laboratory environment can be controlled, valuable ‘truth data' can be obtained, which are generally difficult to obtain in actual field campaigns.
Currently, the focus has been the polarimetric RCS model of wind turbine blades and its interactions with stationary tower structure. Laboratory measurements are performed using the electromagnetic microphysics laboratory (EML) facility, involving dual-polarized T-R antennas at X band, and a microwave scatterometer system based on a vector network analyzer. It is found that for stationary blades, the existing finite-element solver is slow and computational intensive, the simple Optical Physics Approximation (OPA) has successfully matched the measurement in terms of power levels and spatial peak locations. The predicted blade signatures also fit the basic observations from actual radar data. However, the intrinsic limits of OPA prevent it from accurately predicting the multiple reflections and diffractions due to sharp edges of blades. Other accurate and efficient solvers, including Finite Difference Time Domain (FDTD) and Multilevel Fast Monopole Method (MFLMM), are also being tested and will be reported. For rotating blade structures, an X-band, polarimetric, pulsed scatterometer is being developed in parallel to continuous wave measurements, which will serve the more practical aspects of emulating pulsed Doppler radar operations and spectral signatures. The ultimate goal of the project is a telemetry based mitigation algorithm, which will use the dynamical information received directly from the wind turbine to essentially perform a signal cancellation procedure.
Poster Session 10, Advanced Radar Technologies II
Thursday, 8 October 2009, 1:30 PM-3:30 PM, President's Ballroom
Previous paper Next paper