Solution Approach and Implementation
摘要
In this chapter, we outline our solution approach and its implementation. We implement a Benders decomposition algorithm using a branch-and-Benders-cut (B&BC) strategy, leveraging the capabilities of the Gurobi solver to efficiently handle model decompositions developed previously. The chapter details the initialization process, where we set up the Benders master problem (RMP) and corresponding subproblems (SPs), as well as the optimization process that includes various techniques for improving computation time, such as a construction heuristic and an in-out method for dual bound enhancement. We also describe how Benders cuts are separated during optimization, emphasizing the use of parallel subprocesses to streamline calculations and enhance overall time efficiency.