Robust AC-DC Distribution Network Planning Considering N-1 Security Criterion

As the accommodation of renewable energy and the proportion of DC load in distribution network increase,the advantages of using DC distribution system in terms of power supply capacity and quality are becoming increasingly prominent.On the one hand,renewable energy such as wind power and photovoltaics can be connected to the DC distribution network without DC/AC inverters,which reduces the difficulty of grid connection and improves the reliability of the system.On the other hand,the DC load demand can be satisfied by the use of DC distribution network without AC/DC converters,which reduces the line loss and improves the rationality of investment.Therefore,AC-DC hybrid distribution network has become a hotspot in academia.Based on the idea of “modeling?solution?application”,this paper provides a robust planning model for AC-DC hybrid distribution network considering N-1 security criterion.Specifically,the main work of this paper is as follows:Firstly,based on the min-max-min structure,this paper proposes a two-stage robust programming model for AC/DC hybrid distribution network considering N-1 security criterion.The core idea is to find the most economical planning solution while satisfying the N-1 security criterion.The first stage is the planning and operating model for optimizing the AC/DC distribution network planning scheme under normal working conditions.The second stage is the contingency and recovery model,which is used to find the worst N-1 contingency and its corresponding optimal recovery scheme under any given planning solution.Under the min-max-min structure,the two-stage robust optimization model minimizes the investment and operation costs of the planning model and the load-shedding cost of the worst N-1 contingency under any given planning solution to achieve the global optimium of the AC/DC hybrid distribution network planning.Then,an efficient solution strategy based on Benders algorithm and column-and-constraint generation(C&CG)algorithm is designed to solve the proposed two-stage robust optimization model.The procedure is implemented in a master-subproblem framework.The Benders algorithm is used to solve the planning model in the first stage,while the C&CG algorithm is used to generate the security constraints under the worst N-1 contingency in the second stage.The optimal planning decisions determined in the first stage are employed in the second to examine the N-1 security criterion.The security constraints will then be sent back to the first stage for correctness.In particular,this paper proposes a modern Benders algorithm based on the Branchand-cut structure for solving the slow convergence of the traditional Benders algorithm.The modern Benders algorithm realizes a deep fusion of the Benders algorithm and the modern commercial solver Gurobi,which significantly improves the efficiency of the algorithm.Finally,based on the real parameter of distribution network in Ninghai County,Ningbo City,Zhejiang Province,this paper analyzes the grid scale,power balance,grid structure,operating status and renewable energy distribution of Ninghai Power Grid.Four pre-planning work including voltage level determination,medium-and long-term load forecasting,scenario generation and reduction,and renewable energy location and capacity determination.Based on the real parameters of Ninghai Power Grid,an AC-DC hybrid distribution network planning model considering N-1 security criterion is developed and solved by using the proposed solution strategy based on Benders and C&CG algorithm.Finally,the three-stage AC-DC planning solution for Ninghai Power Grid from 2019 to 2025 is given,and the planning solution is evaluated from three perspectives of economy,reliability and power flow distribution.