Distributed Optimization Strategy for Multi-Region Power Scheduling Based on Alternating Direction Method of Multipliers

Chenlin Sun,Jiang Wu,Feng Gao,Xiaohong Guan
DOI: https://doi.org/10.1109/wcica.2016.7578816
2016-01-01
Abstract:The large-scale and frequent cross-regional electricity allocations bring about the critical need to coordinate multi-region power scheduling. It is recognized that each sub-systems' taking charge of own scheduling makes the ideal centralized optimization method infeasible. In this paper, a distributed optimization strategy based on alternating direction method of multipliers (ADMM) is put forward to solve this problem. The strategy can achieve optimal solution of global power system as the ideal centralized optimization method does, by adding steps of information exchange and iterations of decision variables to subsystems' independent scheduling. The distributed bid cost minimization model with a price-wise constant bidding curve is built to describe the multi-region scheduling problem as precisely as possible. To suit the distributed model, the standard ADMM is improved, by expanding the size of sub-systems, adding inequality constraints, and considering the complex coupling relationship between sub-systems. This coordinating strategy uses the framework of augmented Lagrangian to decompose the scheduling problem into a distributed scheduling model for regional power grid, achieving global optimal solution by updating decision variables of each sub-system and Lagrange multipliers with mapping of sub-system information. Simulations on IEEE 30 system are conducted to show the computational performance and solution quality of the proposed method.
What problem does this paper attempt to address?