Distributed optimal scheduling of VPP based on EST: an ADMM algorithm based on historical data online transfer

Abstract

In order to realize the precise control of virtual power plant (VPP) over its internal demand-side resource (DR) clusters with dispersed locations and huge numbers, and realize the VPP's rapid and effective participation in demand response, in the VPP hierarchical control framework, the multi-edge service terminal (MEST) is introduced as the distributed processor of the VPP to decompose and execute the control task. Firstly, the pre-learning method is used to enrich the historical data, and the historical data is clustered based on the spectral clustering. The distributed cooperative control strategy of MEST and the alternating direction multiplier method (ADMM) of historical data online transfer are proposed, which greatly reduces the number of iteration steps. Then, based on the results of the power adjustment task of the MEST, the hierarchical relationship of control priorities is divided, such as the control mode, comfort tolerance value, switch controllable amount and tolerance value, which can reduce the power adjustment cost of VPP and improve user's comfort. Finally, the effectiveness of the proposed algorithm and control strategy in executing distributed optimization tasks is proved by simulation experiments. The convergence time of the proposed algorithm is only 32.7% of that of the ADMM and the consensus algorithm.