Scholarlink

<p>This study investigates mothership-drone route determination using two optimization approaches: a Mixed Integer Linear Programming (MILP) model based on a discrete hexagonal grid, and a Mixed Integer Nonlinear Programming (MINLP) model operating in continuous space. Four model variants are tested for MINLP, while the MILP model is evaluated using hexagonal and micro-hexagonal grid resolutions. The models are then tested across multiple scenarios with varying grid sizes and numbers of targets. All formulations are implemented in AMPL, and parameters are kept as consistent as possible to enable fair comparison. The computational experiments employ two solvers, BARON and Gurobi, revealing notable differences in solution quality and convergence behavior. BARON often struggles to close the optimality gap within the given time limits, particularly in larger or more complex instances, whileGurobi, demonstrates more stable and efficient convergence, frequently achieving lower objective values. Overall, the results indicate that the MILP approach is more suitable for larger routing scenarios due to its computational tractability, whereas the MINLP model is better suited for generating flexible and realistic trajectories in smaller-scale problems. Future research could integrate heuristics and extend the models to 3D or dynamic environments.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Comparative study of MILP and MINLP formulations for mothership–drones routing problems

  • 期刊论文

  • 发表日期：

    2026年4月15日

查看全文
  • Astrid Wahyu Adventri Wibowo,
  • Boglárka G.-Tóth,
  • József Békési

摘要

This study investigates mothership-drone route determination using two optimization approaches: a Mixed Integer Linear Programming (MILP) model based on a discrete hexagonal grid, and a Mixed Integer Nonlinear Programming (MINLP) model operating in continuous space. Four model variants are tested for MINLP, while the MILP model is evaluated using hexagonal and micro-hexagonal grid resolutions. The models are then tested across multiple scenarios with varying grid sizes and numbers of targets. All formulations are implemented in AMPL, and parameters are kept as consistent as possible to enable fair comparison. The computational experiments employ two solvers, BARON and Gurobi, revealing notable differences in solution quality and convergence behavior. BARON often struggles to close the optimality gap within the given time limits, particularly in larger or more complex instances, whileGurobi, demonstrates more stable and efficient convergence, frequently achieving lower objective values. Overall, the results indicate that the MILP approach is more suitable for larger routing scenarios due to its computational tractability, whereas the MINLP model is better suited for generating flexible and realistic trajectories in smaller-scale problems. Future research could integrate heuristics and extend the models to 3D or dynamic environments.