In this paper, we propose a data-driven algorithm that uses historical data, advanced optimization techniques, and machine learning methods to design local controls that emulate the optimal behavior ...