Пример #1
0
def main():
    """Solve the CVRP problem."""
    if len(sys.argv) != 7:
        print(
            'Should be called as follows: python comparison_solver_gurobi_fixedstops.py [number of waypoints] [number of drones] [range of drone] [velocity of drone] [velocity of truck] [number of fixed depots for comparison]'
        )
        return

    numnodes = int(sys.argv[1])
    numdrones = int(sys.argv[2])
    dronerange = math.ceil(float(sys.argv[3]))
    dronevel = float(sys.argv[4])
    truckvel = float(sys.argv[5])
    numdepots = int(sys.argv[6])
    bound_length = 1000

    #max number of vehicles is just one per node lol, vrp solver takes care of the rest
    numvehicles = numnodes

    x, y = np.random.randint(low=0, high=bound_length, size=(2, (numnodes)))

    x = x.tolist()
    y = y.tolist()

    vrpdata, truckdata = efficient_singledepotroad.solve_singledepotroad_fromdata(
        numnodes, numvehicles, dronerange, x, y)

    data = vrpdata['data']
    manager = vrpdata['manager']
    routing = vrpdata['routing']
    solution = vrpdata['solution']
    routedistances = truckdata['routedistances']
    usednodecoords = truckdata['usednodecoords']

    print("\nROUTEDISTANCES LENGTH:", len(routedistances))
    print("\nUSEDNODECOORDS LENGTH:", len(usednodecoords))

    depart_times, wait_times, deployments, route_times, objective, xc, yc, visit_order = truck_gurobi.tsp_truck(
        routedistances, usednodecoords, numdrones, dronevel, truckvel)

    usednodecoords = [0] + usednodecoords

    combined_route_times = sum(route_times.values())
    print("\n--------------------")
    print("\nOPTIMIZED DRONE DEPLOYMENT TIME:", objective)
    print("LOWER BOUND TIME:", combined_route_times / float(numdrones))

    complete_solver_gurobi.graph_solution(data, manager, routing, solution,
                                          depart_times, wait_times,
                                          deployments, route_times, objective,
                                          xc, yc, visit_order, usednodecoords)

    comparison_vrp_fixedstops.multi_fixed_depot_comparison(
        numnodes, numdrones, numdepots, dronerange, x, y, dronevel, truckvel)
        combined_route_times = sum(route_times.values())
        print("\n--------------------")
        print("\nOPTIMIZED DRONE DEPLOYMENT TIME:", objective)

        filename = "ortools_solution"

        complete_solver_gurobi.graph_solution(data, manager, routing, solution,
                                              depart_times, wait_times,
                                              deployments, route_times,
                                              objective, xc, yc, visit_order,
                                              usednodecoords, filename, True)

        usednodecoords.pop(0)
    else:
        depart_times, wait_times, deployments, route_times, objective, xc, yc, visit_order = truck_gurobi.tsp_truck(
            routedistances, usednodecoords, numdrones, dronevel, truckvel)

        usednodecoords = [0] + usednodecoords

        combined_route_times = sum(route_times.values())
        print("\n--------------------")
        print("\nOPTIMIZED DRONE DEPLOYMENT TIME:", objective)

        filename = "gurobi_solution"

        complete_solver_gurobi.graph_solution(data, manager, routing, solution,
                                              depart_times, wait_times,
                                              deployments, route_times,
                                              objective, xc, yc, visit_order,
                                              usednodecoords, filename, True)
def main():
    """Solve the CVRP problem."""
    if len(sys.argv) != 4:
        print(
            'Should be called as follows: python comparison_numdrones_complete.py [number of waypoints] [range of drone] [number of drones]'
        )
        return

    numnodes = int(sys.argv[1])
    dronerange = math.ceil(float(sys.argv[2]))
    numdrones = int(sys.argv[3])
    bound_length = 1000
    dronevel = 15
    truckvels = np.logspace(-1, 3, 10)

    #max number of vehicles is just one per node lol, vrp solver takes care of the rest
    numvehicles = numnodes

    x, y = np.random.randint(low=0, high=bound_length, size=(2, (numnodes)))

    x = x.tolist()
    y = y.tolist()

    vrpdata, truckdata = efficient_singledepotroad.solve_singledepotroad_fromdata(
        numnodes, numvehicles, dronerange, x, y)

    data = vrpdata['data']
    manager = vrpdata['manager']
    routing = vrpdata['routing']
    solution = vrpdata['solution']
    routedistances = truckdata['routedistances']
    usednodecoords = truckdata['usednodecoords']

    print("\nROUTEDISTANCES LENGTH:", len(routedistances))
    print("\nUSEDNODECOORDS LENGTH:", len(usednodecoords))

    finalObjectives = []

    for i in range(len(truckvels)):
        print("starting truck velocity ", truckvels[i], "solution")
        depart_times, wait_times, deployments, route_times, objective, xc, yc, visit_order = truck_gurobi.tsp_truck(
            routedistances, usednodecoords, numdrones, dronevel, truckvels[i])

        usednodecoords = [0] + usednodecoords

        combined_route_times = sum(route_times.values())
        print("\n--------------------")
        print("\nOPTIMIZED DRONE DEPLOYMENT TIME:", objective)
        print("LOWER BOUND TIME:", combined_route_times / float(i))
        finalObjectives.append(objective)

        filename = str(
            i) + "_index_truck_velocity_solution_complete_comparison"
        """ to_graph = True
        if(truckvels[i]/dronevel < 0.05):
            to_graph = False """

        complete_solver_gurobi.graph_solution(data, manager, routing, solution,
                                              depart_times, wait_times,
                                              deployments, route_times,
                                              objective, xc, yc, visit_order,
                                              usednodecoords, filename, False)

        usednodecoords.pop(0)

    print("TRUCK VELS: ", truckvels)
    print("FINAL OBJECTIVES: ", finalObjectives)
Пример #4
0
def main():
    """Solve the CVRP problem."""
    if len(sys.argv) != 3:
        print(
            'Should be called as follows: python comparison_numdrones_complete.py [number of waypoints] [range of drone]'
        )
        return

    numnodes = int(sys.argv[1])
    dronerange = math.ceil(float(sys.argv[2]))
    bound_length = 1000

    #max number of vehicles is just one per node lol, vrp solver takes care of the rest
    numvehicles = numnodes
    dronevel = 15
    truckvel = 5

    #overallObjectives = []

    #for k in range(3):
    """ if(k == 0):
            truckvel = 5
        
        if(k == 1):
            truckvel = 15
        
        if(k == 2):
            truckvel = 45 """

    velocityObjectives = []

    for j in range(10):
        x, y = np.random.randint(low=0,
                                 high=bound_length,
                                 size=(2, (numnodes)))

        x = x.tolist()
        y = y.tolist()

        vrpdata, truckdata = efficient_singledepotroad.solve_singledepotroad_fromdata(
            numnodes, numvehicles, dronerange, x, y)

        data = vrpdata['data']
        manager = vrpdata['manager']
        routing = vrpdata['routing']
        solution = vrpdata['solution']
        routedistances = truckdata['routedistances']
        usednodecoords = truckdata['usednodecoords']

        print("\nROUTEDISTANCES LENGTH:", len(routedistances))
        print("\nUSEDNODECOORDS LENGTH:", len(usednodecoords))

        if (len(routedistances) > 10):
            x, y = np.random.randint(low=0,
                                     high=bound_length,
                                     size=(2, (numnodes)))

            x = x.tolist()
            y = y.tolist()

            vrpdata, truckdata = efficient_singledepotroad.solve_singledepotroad_fromdata(
                numnodes, numvehicles, dronerange, x, y)

            data = vrpdata['data']
            manager = vrpdata['manager']
            routing = vrpdata['routing']
            solution = vrpdata['solution']
            routedistances = truckdata['routedistances']
            usednodecoords = truckdata['usednodecoords']

            print("\nROUTEDISTANCES LENGTH:", len(routedistances))
            print("\nUSEDNODECOORDS LENGTH:", len(usednodecoords))

        finalObjectives = []

        for i in range(1, len(routedistances) + 2):
            print("starting ", i, " drone solution")
            depart_times, wait_times, deployments, route_times, objective, xc, yc, visit_order = truck_gurobi.tsp_truck(
                routedistances, usednodecoords, i, dronevel, truckvel)

            usednodecoords = [0] + usednodecoords

            combined_route_times = sum(route_times.values())
            print("\n--------------------")
            print("\nOPTIMIZED DRONE DEPLOYMENT TIME:", objective)
            print("LOWER BOUND TIME:", combined_route_times / float(i))
            finalObjectives.append(objective)

            filename = str(
                i) + "_drone_solution_complete_comparison_truckvel_" + str(
                    int(truckvel)) + "_iter_" + str(j) + "_numroutes_" + str(
                        len(routedistances))

            if (j == 0):
                complete_solver_gurobi.graph_solution(
                    data, manager, routing, solution, depart_times, wait_times,
                    deployments, route_times, objective, xc, yc, visit_order,
                    usednodecoords, filename, True)
            else:
                complete_solver_gurobi.graph_solution(
                    data, manager, routing, solution, depart_times, wait_times,
                    deployments, route_times, objective, xc, yc, visit_order,
                    usednodecoords, filename, False)

            usednodecoords.pop(0)

        velocityObjectives.append(finalObjectives)

        print("FINAL OBJECTIVES: ", finalObjectives)

    #overallObjectives.append(velocityObjectives)
    print("VELOCITY OBJECTIVES: ", velocityObjectives)