Distributed Decision-Making of Integer Problemswith Limited Information

摘要

In this paper, we develop a distributed multi-attribute decision-making method that can provide optimal solutionsrnduring collaborative planning in supply chains without disclosing private local information of the players. Variablernsplitting and Lagrangian duality create mixed-integer sub-problems for players involving local variables and copiesrnof global variables. The Lagrangian dual and sub-problems are guided towards the optimal solution using subgradientsrnand a branch and bound technique.