Data-driven optimal control of wind turbines using reinforcement learning with function approximation

Data-driven optimal control of wind turbines using reinforcement learning with function approximation
复制标题

使用函数逼近强化学习的数据驱动风力涡轮机优化控制

DOI:
10.1016/j.cie.2022.108934
复制
发表时间:
2023
影响因子:
7.9
通讯作者:
Feng, Qianmei
Feng, Qianmei
中科院分区:
工程技术2区
文献类型:
--
作者:
Peng, Shenglin;Feng, Qianmei

文献摘要

参考文献

被引文献

相似文献

我们提出了一种具有函数逼近的强化学习方法,用于最大化风力涡轮机(WT)的功率输出。风力发电机的优化控制主要使用最大功率点跟踪(MPPT)策略进行顺序决策,可以建模为马尔可夫决策过程(MDP)。在文献中,连续控制变量通常被离散化,以应对传统动态规划方法中的维数灾难。为了提供更准确的预测,我们利用强化学习中的函数逼近将问题表述为具有连续状态和动作空间的 MDP。选择常用的桨距角作为我们关心的控制变量,将其视为系统状态以及其他一些被证明会影响功率输出的可控和不可控变量。对实际数据的计算研究表明,所提出的方法在获得最佳功率输出方面优于文献中的现有方法。
We propose a reinforcement learning approach with function approximation for maximizing the power output of wind turbines (WTs). The optimal control of wind turbines majorly uses the maximum power point tracking (MPPT) strategy for sequential decision-making that can be modeled as a Markov decision process (MDP). In the literature, the continuous control variables are typically discretized to cope with the curse of dimensionality in traditional dynamic programming methods. To provide a more accurate prediction, we formulate the problem into an MDP with continuous state and action spaces by utilizing the function approximation in reinforcement learning. The commonly used pitch angle is selected as a control variable we are concerned with, which is regarded as the system state along with some other controllable and uncontrollable variables proven to affect the power output. Computational studies of real data are conducted to demonstrate that the proposed method outperforms the existing methods in the literature in obtaining the optimal power output.
DOI: 10.1155/2012/736586
发表时间: 2012
影响因子: 1.7
作者:
T. Li;A. Feng;L. Zhao
通讯作者: T. Li;A. Feng;L. Zhao
一种快速高效的风能转换系统最大功率点跟踪的新算法
DOI: --
发表时间: 2008
期刊:
影响因子: --
作者:
K.S.M. Raza;H. Goto;H. J. Guo;O. Ichinokura
通讯作者: O. Ichinokura
DOI: 10.1016/j.cie.2021.107321
发表时间: 2021-04
期刊: Comput. Ind. Eng.
影响因子: --
作者:
Shenglin Peng;Q. Feng
通讯作者: Shenglin Peng;Q. Feng