Real-time Heliostat Field Aiming Strategy Optimization Based on Reinforcement Learning

Zhichen Zeng,Dong Ni,Gang Xiao

DOI: https://doi.org/10.1016/j.apenergy.2021.118224

IF: 11.2

2022-01-01

Applied Energy

Abstract:The heliostat field aiming strategy optimization is a crucial topic in solar power tower (SPT) plants. The optimal aiming strategy attempts to maximize the thermal power output while preserving certain operational limits. While existing methods could produce high-optimality results, they cannot be applied to real-time optimization for ultra large scale problems due to high computational complexity. In this work, an effective and scalable optimization method is proposed to achieve real-time optimization for ultra large scale problems, enabling efficient and robust SPT plant operations under varying solar conditions (i.e., cloud shadowing variations). The real-time optimization problem is reformulated as a reinforcement learning problem, capturing the learning from the intrinsic sequential decision-making process, to minimize unnecessary repetitions of computations. A Crescent Dunes-like SPT plant is studied, with the allowable flux density as the optimization constraint, and the results show that the proposed method provides better or comparable performance compared to heuristic optimization methods with an order of magnitude less computation time.