Field Testing LIDAR-Based Feed-Forward Controls on the NREL Controls Advanced Research Turbine

Abstract

Researchers at the National Renewable Energy Laboratory (NREL) and the University ofStuttgart are designing, implementing, and testing advanced feedback and feed-forward controls for multimegawatt wind turbines that will help reduce the cost of wind energy. Past wind turbine controllers have depended on turbine feedback measurements to determine the controller pitch commands. In this setup, wind speed disturbances can only be corrected after their effects have been detected in the turbine's loads and dynamic response, which causes a delayed control response due to turbine and pitch actuator dynamics. LIght Detection And Ranging (LIDAR) systems can provide information regarding the approaching wind field to the controller in advance, thereby increasing the controller's available reaction time and allowing pitch actuation to occur in advance to mitigate wind disturbance effects. Feed-forward control algorithms that use these "look ahead" wind speed measurements can improve load mitigation and controller performance compared to feedback only controllers. This paper describes the development and field testing of a feed-forward collective pitch control algorithm to show its effects on speed regulation in above-rated wind speeds. The controller is implemented and field tested on one of the Controls Advanced Research Turbines (CARTs) at NREL. The wind speed measurements to the feed-forward controller are provided by BlueScout Technologies' Optical Control System (OCS) LIDAR mounted on the nacelle of the CART3. Results show that inclusion of the LIDAR measurement into the control system leads to further rejection of the wind disturbance at low frequencies when compared to feedback alone. This in turn provides confidence that LIDAR technology could be used to obtain load reductions with wind turbine controls.

Direct Speed Control using LIDAR and turbine data

“…There have been numerous ongoing studies into the means by which LIDAR can improve wind turbine control performance. LIDAR-assisted collective pitch control for load reduction is one option and has actually been demonstrated recently in field testing [1], [2]. Another potential beneficial use of LIDAR is to increase energy yield by LIDAR-assisted yaw control [3], [4].…”

Adaptive Vorsteuerung für Windenergieanlagen

“…Da wie in dieser Arbeit zwar die mittlere Windgeschwindigkeit in der Adaption des Filters beachtet wurde, die Wellenzahl jedoch nicht nachgeführt wurde, erhöhte sich in den Abschnitten mit schlechterer Korrelation durch einen Messfehler die Standardabweichung der Generatordrehzahl um circa 30 %. Mit einem kommerziellen System mit drei festen Messpositionen auf der CART3 konnte nur eine maximale kohärente Wellenzahl von 0,03 rad/m erreicht werden [8]. Damit konnte die Standardabweichung der Generatordrehzahl um circa 10 % reduziert werden (Bild 13 unten).…”

“…Damit konnte die Standardabweichung der Generatordrehzahl um circa 10 % reduziert werden (Bild 13 unten). Diese Ergebnisse bestätigen, dass mit der vorgeschlagenen Vorsteuerung auch unter realen Bedingungen eine Verbesserung in der Regelung erzielt werden kann, die unterschiedliche Erfassung der rotorBild 13 Spektren der Generatordrehzahl von Feldtests auf einer Forschungsanlage mit 2 Rotorblättern (oben, [7]) und mit 3 Rotorblättern (unten, [8]): Ohne Vorsteuerung (dunkelgrau); mit LIDAR basierter adaptiver Vorsteuerung (schwarz).…”

A Lidar Assisted Feedforward Preview Controller for Load Mitigation in Wind Turbines

“…LIDARs provide information about the future wind inflow seconds ahead, which can be used by novel control systems to promote better speed regulation and load alleviation. In recent years a number of studies related to LIDAR-enabled control has been reported [7] and feedforward control has already been tested on experimental turbines [13,14]. The approach used in the field tests [13,14] was to augment an existing feedback controller with a feedforward term.…”

“…In recent years a number of studies related to LIDAR-enabled control has been reported [7] and feedforward control has already been tested on experimental turbines [13,14]. The approach used in the field tests [13,14] was to augment an existing feedback controller with a feedforward term. This technique has 2 also been investigated in [2] with various design methods, such as model-inverse, optimized FIR filter and shaped compensators, preview control, non-causal series expansion, zerophase-error tracking and adaptive methods.…”