Decomposition and Parallel Computation for Operation Optimization of Medium and Long term Cascaded Hydropower Units

Abstract

Climate change and environmental pollution are global issues facing mankind. The energy transformation is currently a social issue of global concern, and the focus is on the power industry to vigorously develop clean energy. Hydropower is the most mature renewable energy source and has been in a priority position in the energy mix. In this paper, we propose a decomposition and parallel calculation method to optimize the operation of medium and long-term cascaded hydropower units. Firstly, a model for optimal scheduling of cascaded hydropower units is established. Then, the model is decoupled in time based on the augmented Lagrangian relaxation. And the diagonal quadratic approximation method is introduced to ensure the decomposability of the decoupled problem so that the subproblems can be computed in parallel. In addition, an arithmetic analysis is performed with actual data to verify the effectiveness of the proposed method.