The characteristics of the instances are listed in Table 1. The values in the table mean the number of vehicles, dealers, warehouses, and auto-carriers for each instance. The data of instances can be found at https://drive.google.com/drive/folders/1XRFouRwvo9YDPwh5mzIE_cOOEFYNyVul? usp=sharing.
Strategies for Solving the Mathematical Model
Lexicographic OptimizationThe lexicographic optimization strategy first solves the problem with the primary objective.Then, after the termination of the solving process, it solves the problem that considers the next objective and includes the constraints imposed by the previous objective values. Solving the problem with each objective stops if the optimal solution is obtained or if the runtime of CPLEX exceeds two hours. The optimization process calls four repetitions of CPLEX and stops if the problem with the quaternary objective is solved. As shown in the above iterations, when solving the problem with one objective function, the constraints on the previous objective function values are added to the model. Therefore, if CPLEX solves the four problems to optimality, then the optimal solution to our problem is obtained based on the dominant relationship between solutions. Its details are described in Algorithm A.1.
Weighted Sum OptimizationThe four objectives have different optimization directions and scales. The weighted sum objective function is presented as f all = w 1 • − f urgent + w 2 • (− f vehicles ) + w 3 • f visits + w 4 • f distance , where