.png){kind=icon}
返回
Q-Learning Algorithm and CMAC Approximation Based Robust Optimal Control for Renewable Energy Management Systems
.png){kind=icon}
PRE
AI
DOI：.png)
.png){kind=icon}
原文求助.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
This paper investigates a robust optimal control algorithm for a renewable energy management system. The algorithm is obtained by developing a novel method based on the zero-sum games (ZSG) theory for H(infinity )control and the well-known Q-learning algorithm. Firstly, the H-infinity performance index function is formed via real-time parameters including electricity price, load demand, solar energy, and battery lifetime. Secondly, a self-learning and control algorithm is established, and the value function solution is approximated by the cerebellar model articulation controller (CMAC). Finally, the algorithm guarantees that the disturbance compensation policy, optimal value function, and the optimal control strategy converge to the near-optimal values. Comparison with other methods in a numerical experiment using practically measured data is implemented to evaluate the effectiveness of the designed algorithm.
Keyword:
Renewable energy
storage system
robust optimal control
Q-learning
CMAC
neural network
