2.4 LocalSolver

LocalSolver is a relatively new kind of mathematical optimization commercial solver. It provides APIs for Python, Java, C#, and C++ to model and solve your problems at hand. It also includes an powerful own proprietary modeling language. The optimizer combines state-of-the-art algorithms under the hood, it is very efficient for all kind of vehicle routing, scheduling, and other sequence problems and complements the best commercial MIP solvers, like Gurobi or Cplex. Large vehicle routing problems can be solved near optimally (see CVRP). It also can solve linear, integer, and non-linear problems without using derivatives, and contains lower bound algorithms (for minimizing problems) to get the optimality gap in the same way as MIP solver do. Models with logical constraints fits naturally into its syntax.

Its specially dedicated strength lies in its treatment of “permutation problems” (see my Permutation Paper for more information) which are ideally modeled and solved using LocalSolver. No other modeling system to my knowledge has this feature and capability. I use the classical cvrp (Capacitated Vehicle Routing Problem)

   2.4.1 Capacitated Vehicle Routing Problem (cvrp2)