Wind Farm Consensus Yaw Control Using Distributed Reinforcement Learning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Wind turbines in a wind farm typically operate independently, relying on nacelle-based wind measurements, which can be noisy and unreliable, leading to dynamic yaw misalignment and inefficient power production due to the lack of consideration for nearby turbines' wind measurements.
Innovation Solution
A system that uses an augmented Lagrangian method, specifically alternating direction method of multipliers (ADMM), to combine wind measurements from multiple turbines, calculate a consensus wind estimate, and adjust operating parameters to improve yaw alignment and power production.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If wind turbines operate independently using only local nacelle-based wind measurements, then each turbine can maintain simple individual control, but the wind direction estimation becomes noisy and unreliable leading to dynamic yaw misalignment
Solution Approach 1:
The patent combines wind measurement data from multiple turbines into a consensus estimate using an augmented Lagrangian method. Each turbine shares its local wind measurements with neighbors, and the consensus algorithm integrates these measurements to produce a more accurate wind direction estimate for each turbine, reducing the noise and unreliability of individual nacelle-based measurements.
Solution Approach 2:
The patent introduces a consensus algorithm as an intermediary processing layer between individual turbine measurements and final control decisions. This mediator aggregates and reconciles measurements from multiple turbines, producing a consensus wind estimate that serves as a more reliable input for yaw control than individual measurements alone.
2Productivity
If wind turbines use only local wind measurements for yaw control, then the control system remains simple, but power production efficiency decreases due to yaw misalignment
Solution Approach 1:
The patent merges wind measurement information from multiple turbines to create a consensus estimate that contains more complete and accurate wind direction information than any single turbine's local measurement. This enriched information directly improves power production efficiency by enabling more accurate yaw alignment.
3Reliability
If wind turbines operate independently without considering nearby turbines' measurements, then the system architecture remains simple, but dynamic yaw misalignment increases reducing overall performance
Solution Approach 1:
The consensus algorithm acts as an intermediary that processes and reconciles measurements from multiple turbines, producing a stable and reliable wind direction estimate. This mediator handles the complexity of integrating multiple data sources, allowing individual turbines to maintain simple control logic while achieving improved yaw alignment stability through the consensus estimate.
Data Source
AI summary
Disclosed herein are methods, systems, and devices for utilizing distributed reinforcement learning and consensus control to most effectively generate and utilize energy. In some embodiments, individual turbines within a wind farm may communicate to reach a consensus as to the desired yaw angle based on the wind conditions.