Semiconductor Laser Wind Lidar for Turbine Control

This thesis describes an experimentally oriented study of continuous wave (CW) coherent Doppler lidar system design. The main application is remote wind sensing for active wind turbine control using nacelle mounted lidar systems; and the primary focus is to devise an industrial instrument that can improve the efficiency of harvesting wind energy in commercial wind farms. This work attempts to provide a complete investigation of all the necessary building blocks in a CW wind lidar, from the light source to the optical transceiver. The basic concept of Doppler lidar is introduced along with a brief historical overview within the topic of wind lidar systems. Both the potential and the challenges of an industrialized wind lidar has been addressed here. Furthermore, the basic concept behind the heterodyne detection and a brief overview of the lidar signal processing is explained; and a simple representation of the system signal-to-noise ratio (SNR), including the most common and relevant noise sources, is formulated.The impact of various system parameters, such as insertion loss, backscatter coefficient and transceiver telescope design, on the total signal power is discussed and analysed. A thorough investigation of the telescope truncation and lens aberrations is conducted, both numerically and experimentally. It is shown that these parameters dictate the spatial resolution of the lidar system, and have profound impact on the SNR. In this work, an all-semiconductor light source is used in the lidar design instead of the conventional fiber-lasers. Besides its advantage of lower cost, the relative intensity noise, which peaks around 1 MHz for fiber lasers, is inherently avoided by using a semiconductor light source. The impact of the line width increment on the SNR in the application of wind measurement has been investigated. The results hows a much less SNR penalty than expected, due to a finite signal bandwidth of the wind signal.For applications such as active yaw or pitch control, multiple lines of sight are required of the lidar system. Thus, two different beam steering methods have been investigated and demonstrated in this work. The challenge, aside from cost and compactness, is to ensure a long lifetime without regular maintenance, since the wind turbines are designed to last for 20 years. Finally, field test results of various measurement campaigns, designed to evaluate our lidar design, are presented here. Our design has been compared with both sonic anemometers and a conventional fiber laser based CW wind lidar. The results show no significant performance difference between the systems in terms of determining the wind speed of the measurement volume.