A New LMP-Sensitivity-Based Heterogeneous Decomposition for Transmission and Distribution Coordinated Economic Dispatch.

Zhengshuo Li,Qinglai Guo,Hongbin Sun,Jianhui Wang
DOI: https://doi.org/10.1109/tsg.2016.2571848
2017-01-01
Abstract:Transmission and distribution coordinated economic dispatch (TDCED) will be necessary to settle the problems caused by high penetrations of distributed generators in future grids. An efficient heterogeneous decomposition algorithm (HGD) was proposed to solve the TDCED, but it may diverge occasionally. To overcome this shortcoming, a new HGD (N-HGD) with enhanced convergency is proposed and studied in this paper. Compared with the previous HGD, the N-HGD similarly decomposes the TDCED into transmission ED and distribution ED subproblems, but it considers the sensitivity of the boundary locational marginal price (LMP) to load injections to enhance convergency and avoid divergence. Two kinds of sensitivity computing methods, Mc and Mp, are studied. Mc is based on a centralized sensitivity equation, executed on the transmission side. Mp is based on a probing mechanism and can be executed distributedly on the distribution side. The reason for the N-HGD's improved convergency is theoretically analyzed, and the optimality is proven. Numerical tests show that N-HGD performs better than the original HGD, as well as a penalty-based modified HGD that needs a careful parametertuning process. Furthermore, compared with Mc, an Mp-based N-HGD is preferable because it has a much lower computational cost and no additional communicational cost.
What problem does this paper attempt to address?