Two-Stage Robust Planning Method For Power Grid In The Scenario Of Multi Data Center Microgrid Access

In recent years,the cloud computing industry has flourished all over the world,resulting in a sharp increase in the number of its core support system-data centers,and the corresponding energy consumption is also rising.The high cost has aroused the high attention of governments,enterprises and research institutions.Therefore,it is of great significance to fully exploit the operational flexibility of data centers,study the collaborative optimization of multiple data centers and power grid resources,and ensure the interests of each data center on privacy issues for the collaborative planning of data centers and power grids.This paper has extended from the single data center planning problem considering uncertainty to the multiple data center collaborative planning problem considering privacy,providing corresponding planning strategies for the data center operators and power grid operators.First of all,taking the data center as the core element,the model of the data center microgrid is established,including the load of the data center itself,the model of the interaction between the gas turbine,energy storage equipment,photovoltaic units and the microgrid and the power grid,and the scheduling logic of the load when the data center operates,as well as the system operation mode of the data center microgrid and the superior power grid.Then,in view of the microgrid model,considering the uncertainty of PV unit output,conventional load and data center load in the microgrid,a two-stage robust planning model of the data center microgrid is established based on the dual algorithm to plan the capacity of gas turbines and energy storage equipment in the microgrid.Considering that different data center microgrid operators may have different tolerance for uncertainty,Therefore,the system planning costs under different robust parameters are calculated separately,which provides a certain reference for operators’ planning.At the same time,the operation curves of typical intra-day systems are obtained through operation simulation.The example shows that the proposed algorithm can delay the delay load from the period of high electricity price to the period of low electricity price,and can play a positive role in reducing the microgrid operation costs and peak shaving and valley filling.Finally,considering the privacy issues involved in the operation of multiple data centers in the regional distribution network,a solution method based on the improved alternating direction multiplier method is proposed.This method can perform parallel calculation on the subjects of multiple areas after the distribution network completes the area division,and improve the public information matrix,reducing the information that each subject needs to provide at the public nodes.Compared with the original method,it can better protect the private information belonging to users of different interest subjects.At the same time,in the iterative process of the algorithm,the determination method of the optimization step length and penalty coefficient is improved.When the two variables take different values,the number of iterations required for the convergence of the algorithm is calculated.The example shows that compared with the traditional method,the number of iterations of the algorithm can be effectively reduced by adding a multiplier.