Hierarchical Online-Optimization for the Control of Flexible Loads and Generators in Distribution Grids

This paper proposes a hierarchical online algorithm to solve a multi-level Optimal Power Flow problem in distribution grids with a high penetration of distributed energy resources. Based on an overall utility function, the developed method aims to optimize the usage of controllable loads and generators while maintaining compliance with the defined engineering constraints. The hierarchical control architecture is composed of a superordinate controller and several regional controllers that operate on lower levels. Using an online projected gradient method, the controllers compute incremental changes for the active power setpoints and use mutual communication to capture real-time measurements of power injections, nodal voltages and line currents. The hereby created feedback loop allows to accurately monitor the grid state and detect potential limit violations. The developed algorithm is tested using time series simulations in a MATLAB/SIMULINK environment. It is shown that the hierarchical controller is suitable for controlling exemplary critical power flow scenarios, while enabling higher computational efficiency since the computations are distributed among multiple units.