A simulated annealing algorithm to solve the log-truck scheduling problem

In this paper, we present a simulated annealing (SA) approach for solving the log-truck scheduling problem, which is an extension of the timber transport vehicle routing problem with time windows (TTVRPTW). The problem is characterised by a heterogeneous fleet of trucks that depart from their corresponding depots and must perform a number of daily transport tasks to move log products between wood pickup (harvest areas) locations and industrial customers, such as pulpmills and sawmills. Commencing at their depots, each truck must perform a number of successive transport tasks, each one charaterised by a trip between a designated harvest area and an industrial customer, until the number of daily working hours is completed (Karanta 2000, Audy et al. 2012). The objective of the model is to minimise total empty travel time for the whole fleet as well as waiting times at wood pickup locations. In addition, time windows and accessibility to harvest areas and customers, must be taken into account.