Optimal Switch and Tie Line Planning in Distribution Networks: Benchmarking a Practical MILP Model With a Fast Heuristic Approach

作     者：Jooshaki, Mohammad Karimi-Arpanahi, Sahand Millar, R. John Anttila, Jaakko Lehtonen, Matti Fotuhi-Firuzabad, Mahmud 

作者机构： EspooFI-02151 Finland University of Adelaide School of Electrical and Mechanical Engineering AdelaideSA5005 Australia Aalto University Department of Electrical Engineering and Automation Espoo02150 Finland Trimble Utilities and Public Administration - Europe Espoo02600 Finland Sharif University of Technology Electrical Engineering Department Tehran11155-4363 Iran 

出 版 物：《IEEE Transactions on Power Systems》 (IEEE Trans Power Syst)

年 卷 期：2025年第40卷第4期

页      面：2860-2872页

核心收录：

学科分类：0808[工学-电气工程] 080802[工学-电力系统及其自动化] 08[工学] 0835[工学-软件工程] 0701[理学-数学] 0812[工学-计算机科学与技术（可授工学、理学学位）] 

基　　金：The work of R. John Millar was supported in part by the Department of Electrical Engineering and Automation  Aalto University  Espoo  Finland  and in part by Trimble Solutions. The work of Mahmud Fotuhi-Firuzabad was supported by Iran National Science Foundation (INSF) 

主　　题：Mixed integer linear programming 

摘      要：Optimal placement of switches and tie lines is an integral task in distribution system planning. Owing to the problem complexity and the presence of binary decision variables, using heuristic methods for obtaining a close to optimal switch and tie line plan is a common practice in industry. Efforts to employ mixed-integer linear programming (MILP) to guarantee the solution optimality also tend to sacrifice the modeling accuracy through oversimplifying assumptions to make the problem tractable. Aiming to tackle these challenges, we present an efficient, yet accurate, MILP model for optimal switch and tie line planning. The proposed model avoids common simplifying assumptions in the state-of-the-art MILP models while preserving the solving efficiency. In order to demonstrate the applicability and scalability of the proposed MILP approach, it is applied to multiple test networks, and the results are compared with those of a fast heuristic model. The outcomes not only represent the high efficiency and accuracy of the MILP model but also validate the close to optimality of the heuristic approach developed for practical applications as a module in a commercial distribution system planning toolbox. © 1969-2012 IEEE.