Multi Depot Vehicle Routing Problem (MDVRP)¶
Principles learned¶
Use lists and decision variables
Use a lambda expression to define a recursive array
Use the partition operator
Use the sum operator
Problem¶
In the Multi Depot Vehicle Routing Problem (MDVRP), which is an extension of the Capacitated Vehicle Routing Problem (CVRP), a fleet of delivery vehicles must serve customers with known demand and service time for a single commodity. Contrary to the CVRP, which treats the MDVRP problem with only one depot location, different depot locations are available, each of them having its own kind of trucks with different properties (demand capacity and service time capacity). The vehicles must end their routes at their starting depot. Each customer can only be served by one vehicle. The objective is to determine which customers should be served by each truck from each known depot by minimizing the total distance traveled by all the trucks, such that all customers are served (demand and service time), the demand for every truck must not exceed its capacity and the time it spends outside its depot must not exceed its time capacity.
Download the exampleData¶
The instances used in this example come from the Cordeau_2011 instances.
The format of the data files is as follows:
One line for: data type (2 for mdvrp), number of trucks per depot, number of customers, number of depots.
One line for each depot with: route duration capacity of trucks, quantity capacity of trucks.
One line for each customer with: customer id, coordinate x, coordinate y, service time needed, demand needed (we ignore the rest of the line which is for PVRP).
One line for each depot with: depot id, coodinate x, coordinate y.
Program¶
This Hexaly model defines a two dimensional vector customersSequences with
size nbTrucks
x nbDepots
where customersSequences[d][k]
is a list
which corresponds to the sequence of customers visited by this truck of this
depot. To ensure that all customers must be served by only one truck, all the
list variables are constrained to form a partition by using the
“partition” operator.
The quantity delivered to a customer has to meet the corresponding demand, the total quantity in a truck should not exceed its capacity and the time that the truck spends outside should not exceed its service time capacity.
The quantity served routeQuantity
by a truck is defined using a variadic sum
operator and the access to the demands
array with the
“at” operator. We can then constrain this
expression to never exceed the capacity of the truck.
In the same way, the two dimensional vector routeDistances
and especially
routeDistances[d][k]
contains the distance traveled by the truck k
of
depot d
. We define it again with a sum accessing the distanceMatrix
array with the “at” operator.
distanceMatrix
gives the distance between two customers, to compute the
total distance traveled by a truck we need to add the distances to leave and
come back to the depot stored in the distanceDepot
array.
If we add to the distance the service times needed to serve every customer this
truck should serve, we obtain the time it needs to spend outside (again with
the sum and the routeDurationCapacity
vector). This one must not exceed its
service time capacity.
The objective is to minimize the sum of the elements in routeDistances
:
totalDistance
, because it is the total distance traveled by
all the trucks.
- Execution:
- localsolver mdvrp.lsp inFileName=instances/p01 [lsTimeLimit=] [outputFileName=]
use io;
/* Read instance data */
function input() {
local usage = "Usage: localsolver mdvrp.lsp inFileName=inputFile"
+ "[solFileName=outputFile] [lsTimeLimit=timeLimit]";
if (inFileName == nil) throw usage;
// Read the data and compute matrices
readInputMdvrp();
computeDistanceMatrix();
}
/* Declare the optimization model */
function model() {
// Sequence of customers visited by each truck
customersSequences[d in 0...nbDepots][k in 0...nbTrucksPerDepot] <- list(nbCustomers);
// All customers must be visited by exactly one truck
constraint partition[d in 0...nbDepots][k in 0...nbTrucksPerDepot](customersSequences[d][k]);
for [d in 0...nbDepots][k in 0...nbTrucksPerDepot] {
local sequence <- customersSequences[d][k];
local c <- count(sequence);
// The quantity needed in each route must not exceed the truck capacity
routeQuantity <- sum(0...c, i => demands[sequence[i]]);
constraint routeQuantity <= truckCapacity[d];
// Distance traveled by truck k of depot d
routeDistances[d][k] <- sum(1...c, i => distanceMatrix[sequence[i - 1]][sequence[i]])
+ (c > 0 ? (distanceDepot[d][sequence[0]] + distanceDepot[d][sequence[c - 1]]) : 0);
// We add service Time
routeServiceTime <- sum(0...c, i => serviceTime[sequence[i]]);
routeDuration <- routeDistances[d][k] + routeServiceTime;
// The total distance of this truck should not exceed the duration capacity of the truck
// (only if we define such a capacity)
if (routeDurationCapacity[d] > 0) {
constraint routeDuration <= routeDurationCapacity[d];
}
}
// Total distance traveled
totalDistance <- sum[d in 0...nbDepots][k in 0...nbTrucksPerDepot](routeDistances[d][k]);
// Objective: minimize the total distance traveled
minimize totalDistance;
}
function param() {
if (lsTimeLimit == nil) lsTimeLimit = 20;
}
// Write the solution in a file with the following format:
// - instance, time_limit, total distance
// - for each depot and for each truck in this depot, the customers visited
function output() {
if (solFileName == nil) return;
local resfile = io.openWrite(solFileName);
resfile.println("Instance: "+ inFileName + " ; time_limit: " +
lsTimeLimit + " ; Objective value: " + totalDistance.value);
for [d in 0...nbDepots] {
nbTrucksUsed = 0;
for [k in 0...nbTrucksPerDepot] {
if (customersSequences[d][k].value.count() > 0) {
nbTrucksUsed += 1;
}
}
if (nbTrucksUsed > 0){
n = 1;
resfile.println("Depot " + (d + 1));
for [k in 0...nbTrucksPerDepot] {
if (customersSequences[d][k].value.count() > 0) {
resfile.print("Truck " + n + " : ");
for [customer in customersSequences[d][k].value] {
resfile.print(customer + 1, " ");
}
n += 1;
resfile.println();
}
}
resfile.println();
}
}
}
// Input files following "Cordeau"'s format
function readInputMdvrp() {
local inFile = io.openRead(inFileName);
inFile.readInt(); // We ignore the first int of the instance
// Numbers of trucks per depot, customers and depots
nbTrucksPerDepot = inFile.readInt();
nbCustomers = inFile.readInt();
nbDepots = inFile.readInt();
for [d in 0...nbDepots] {
routeDurationCapacity[d] = inFile.readInt(); // Time capacity for every type of truck from every depot
truckCapacity[d] = inFile.readInt(); // Capacity for every type of truck from every depot
}
for [n in 0...nbCustomers] {
inLine = inFile.readln();
line = inLine.split();
// Coordinates X and Y, service time and demand for customers
nodesX[n] = line[1].toDouble();
nodesY[n] = line[2].toDouble();
serviceTime[n] = line[3].toInt();
demands[n] = line[4].toInt();
}
for [d in 0...nbDepots] {
inLine = inFile.readln();
line = inLine.split();
// Coordinates X and Y for depots
depotX[d] = line[1].toDouble();
depotY[d] = line[2].toDouble();
}
}
// Compute the distance matrix for customers and for depots/warehouses
function computeDistanceMatrix() {
for [i in 0...nbCustomers] {
distanceMatrix[i][i] = 0;
for [j in (i + 1)...nbCustomers] {
local localDistance = computeDist(i, j);
distanceMatrix[j][i] = localDistance;
distanceMatrix[i][j] = localDistance;
}
}
for [i in 0...nbCustomers][d in 0...nbDepots] {
local localDistance = computeDistDepot(d, i);
distanceDepot[d][i] = localDistance;
}
}
// Compute the distance between two points
function computeDist(i, j) {
return sqrt(pow(nodesX[i] - nodesX[j], 2) + pow(nodesY[i] - nodesY[j], 2));
}
function computeDistDepot(d, j) {
return sqrt(pow(depotX[d] - nodesX[j], 2) + pow(depotY[d] - nodesY[j], 2));
}
- Execution (Windows)
- set PYTHONPATH=%LS_HOME%\bin\pythonpython mdvrp.py instances/p01
- Execution (Linux)
- export PYTHONPATH=/opt/localsolver_13_0/bin/pythonpython mdvrp.py instances/p01
import localsolver
import sys
import math
def main(instance_file, str_time_limit, output_file):
#
# Read instance data
#
nb_trucks_per_depot, nb_customers, nb_depots, route_duration_capacity_data, \
truck_capacity_data, demands_data, service_time_data, \
distance_matrix_customers_data, distance_warehouse_data = read_input_mdvrp(instance_file)
with localsolver.LocalSolver() as ls:
#
# Declare the optimization model
#
model = ls.model
# Sequence of customers visited by each truck
customer_sequences = [[model.list(nb_customers) for _ in range(nb_trucks_per_depot)] for _ in range(nb_depots)]
# Vectorization for partition constraint
customer_sequences_constraint = [customer_sequences[d][k]
for d in range(nb_depots) for k in range(nb_trucks_per_depot)]
# All customers must be visited by exactly one truck
model.constraint(model.partition(customer_sequences_constraint))
# Create LocalSolver arrays to be able to access them with an "at" operator
demands = model.array(demands_data)
service_time = model.array(service_time_data)
dist_customers = model.array(distance_matrix_customers_data)
# Distances traveled by each truck from each depot
route_distances = [[None for _ in range(nb_trucks_per_depot)] for _ in range(nb_depots)]
# Total distance traveled
total_distance = model.sum()
for d in range(nb_depots):
dist_depot = model.array(distance_warehouse_data[d])
for k in range(nb_trucks_per_depot):
sequence = customer_sequences[d][k]
c = model.count(sequence)
# The quantity needed in each route must not exceed the truck capacity
demand_lambda = model.lambda_function(lambda j: demands[j])
route_quantity = model.sum(sequence, demand_lambda)
model.constraint(route_quantity <= truck_capacity_data[d])
# Distance traveled by truck k of depot d
dist_lambda = model.lambda_function(lambda i: model.at(dist_customers, sequence[i - 1], sequence[i]))
route_distances[d][k] = model.sum(model.range(1, c), dist_lambda) \
+ model.iif(c > 0, dist_depot[sequence[0]] + dist_depot[sequence[c - 1]], 0)
# We add service Time
service_lambda = model.lambda_function(lambda j: service_time[j])
route_service_time = model.sum(sequence, service_lambda)
total_distance.add_operand(route_distances[d][k])
# The total distance should not exceed the duration capacity of the truck
# (only if we define such a capacity)
if (route_duration_capacity_data[d] > 0):
model.constraint(route_distances[d][k] + route_service_time <= route_duration_capacity_data[d])
# Objective: minimize the total distance traveled
model.minimize(total_distance)
model.close()
# Parameterize the solver
ls.param.time_limit = int(str_time_limit)
ls.solve()
#
# Write the solution in a file with the following format:
# - instance, time_limit, total distance
# - for each depot and for each truck in this depot, the customers visited
#
if output_file is not None:
with open(output_file, 'w') as f:
f.write("Instance: " + instance_file + " ; " + "time_limit: " + str_time_limit + " ; " +
"Objective value: " + str(total_distance.value))
f.write("\n")
for d in range(nb_depots):
trucks_used = []
for k in range(nb_trucks_per_depot):
if (len(customer_sequences[d][k].value) > 0):
trucks_used.append(k)
if len(trucks_used) > 0:
f.write("Depot " + str(d + 1) + "\n")
for k in range(len(trucks_used)):
f.write("Truck " + str(k + 1) + " : ")
customers_collection = customer_sequences[d][trucks_used[k]].value
for p in range(len(customers_collection)):
f.write(str(customers_collection[p] + 1) + " ")
f.write("\n")
f.write("\n")
# Input files following "Cordeau"'s format
def read_input_mdvrp(filename):
with open(filename) as f:
instance = f.readlines()
nb_line = 0
datas = instance[nb_line].split()
# Numbers of trucks per depot, customers and depots
nb_trucks_per_depot = int(datas[1])
nb_customers = int(datas[2])
nb_depots = int(datas[3])
route_duration_capacity = [None]*nb_depots # Time capacity for every type of truck from every depot
truck_capacity = [None]*nb_depots # Capacity for every type of truck from every depot
for d in range(nb_depots):
nb_line += 1
capacities = instance[nb_line].split()
route_duration_capacity[d] = int(capacities[0])
truck_capacity[d] = int(capacities[1])
# Coordinates X and Y, service time and demand for customers
nodes_xy = [[None, None]] * nb_customers
service_time = [None] * nb_customers
demands = [None] * nb_customers
for n in range(nb_customers):
nb_line += 1
customer = instance[nb_line].split()
nodes_xy[n] = [float(customer[1]), float(customer[2])]
service_time[n] = int(customer[3])
demands[n] = int(customer[4])
# Coordinates X and Y of every depot
depot_xy = [None] * nb_depots
for d in range(nb_depots):
nb_line += 1
depot = instance[nb_line].split()
depot_xy[d] = [float(depot[1]), float(depot[2])]
# Compute the distance matrices
distance_matrix_customers = compute_distance_matrix_customers(nodes_xy)
distance_warehouse = compute_distance_warehouse(depot_xy, nodes_xy)
return nb_trucks_per_depot, nb_customers, nb_depots, route_duration_capacity, \
truck_capacity, demands, service_time, distance_matrix_customers, distance_warehouse
# Compute the distance matrix for customers
def compute_distance_matrix_customers(nodes_xy):
nb_customers = len(nodes_xy)
distance_matrix = [[0 for _ in range(nb_customers)] for _ in range(nb_customers)]
for i in range(nb_customers):
for j in range(i+1, nb_customers):
distij = compute_dist(nodes_xy[i], nodes_xy[j])
distance_matrix[i][j] = distij
distance_matrix[j][i] = distij
return distance_matrix
# Compute the distance matrix for warehouses/depots
def compute_distance_warehouse(depot_xy, nodes_xy):
nb_customers = len(nodes_xy)
nb_depots = len(depot_xy)
distance_warehouse = [[0 for _ in range(nb_customers)] for _ in range(nb_depots)]
for i in range(nb_customers):
for d in range(nb_depots):
distance_warehouse[d][i] = compute_dist(depot_xy[d], nodes_xy[i])
return distance_warehouse
# Compute the distance between two points
def compute_dist(p, q):
return math.sqrt(math.pow(p[0] - q[0], 2) + math.pow(p[1] - q[1], 2))
if __name__ == '__main__':
if len(sys.argv) < 2:
print("Usage: python mdvrp.py input_file [output_file] [time_limit]")
sys.exit(1)
instance_file = sys.argv[1]
output_file = sys.argv[2] if len(sys.argv) > 2 else None
str_time_limit = sys.argv[3] if len(sys.argv) > 3 else "20"
main(instance_file, str_time_limit, output_file)
- Compilation / Execution (Windows)
- cl /EHsc mdvrp.cpp -I%LS_HOME%\include /link %LS_HOME%\bin\localsolver130.libmdvrp instances/p01.dat
- Compilation / Execution (Linux)
- g++ mdvrp.cpp -I/opt/localsolver_13_0/include -llocalsolver130 -lpthread -o mdvrp./mdvrp instances/p01
#include "localsolver.h"
#include <cmath>
#include <cstring>
#include <fstream>
#include <iostream>
#include <vector>
using namespace localsolver;
using namespace std;
class Mdvrp {
public:
LocalSolver localsolver;
// Number of customers
int nbCustomers;
// Number of depots/warehouses
int nbDepots;
// Number of trucks per depot
int nbTrucksPerDepot;
// Capacity of the trucks per depot
vector<int> truckCapacity;
// Duration capacity of the trucks per depot
vector<int> routeDurationCapacity;
// Service time per customer
vector<int> serviceTimeData;
// Demand per customer
vector<int> demandsData;
// Distance matrix between customers
vector<vector<double>> distanceMatrixCustomersData;
// Distances between customers and depots
vector<vector<double>> distanceWarehouseData;
// Decision variables
vector<vector<LSExpression>> customersSequences;
// Distance traveled by all the trucks
LSExpression totalDistance;
/* Read instance data */
void readInstance(const string& fileName) { readInputMdvrp(fileName); }
void solve(int limit) {
// Declare the optimization model
LSModel model = localsolver.getModel();
// Sequence of customers visited by each truck
customersSequences.resize(nbDepots);
// Vectorization for partition constraint
vector<LSExpression> customersSequencesConstraint(nbDepots * nbTrucksPerDepot);
for (int d = 0; d < nbDepots; ++d) {
customersSequences[d].resize(nbTrucksPerDepot);
for (int k = 0; k < nbTrucksPerDepot; ++k) {
customersSequences[d][k] = model.listVar(nbCustomers);
customersSequencesConstraint[d * nbTrucksPerDepot + k] = customersSequences[d][k];
}
}
// All customers must be visited by exactly one truck
model.constraint(model.partition(customersSequencesConstraint.begin(), customersSequencesConstraint.end()));
// Create LocalSolver arrays to be able to access them with an "at" operator
LSExpression demands = model.array(demandsData.begin(), demandsData.end());
LSExpression serviceTime = model.array(serviceTimeData.begin(), serviceTimeData.end());
LSExpression distMatrix = model.array();
for (int n = 0; n < nbCustomers; ++n) {
distMatrix.addOperand(
model.array(distanceMatrixCustomersData[n].begin(), distanceMatrixCustomersData[n].end()));
}
// Distances traveled by each truck from each depot
vector<vector<LSExpression>> routeDistances;
routeDistances.resize(nbDepots);
// Total distance traveled
totalDistance = model.sum();
for (int d = 0; d < nbDepots; ++d) {
routeDistances[d].resize(nbTrucksPerDepot);
LSExpression distDepot = model.array(distanceWarehouseData[d].begin(), distanceWarehouseData[d].end());
for (int k = 0; k < nbTrucksPerDepot; ++k) {
LSExpression sequence = customersSequences[d][k];
LSExpression c = model.count(sequence);
// The quantity needed in each route must not exceed the truck capacity
LSExpression demandLambda = model.createLambdaFunction([&](LSExpression j) { return demands[j]; });
LSExpression routeQuantity = model.sum(sequence, demandLambda);
model.constraint(routeQuantity <= truckCapacity[d]);
// Distance traveled by truck k of depot d
LSExpression distLambda = model.createLambdaFunction(
[&](LSExpression i) { return model.at(distMatrix, sequence[i - 1], sequence[i]); });
routeDistances[d][k] = model.sum(model.range(1, c), distLambda) +
model.iif(c > 0, distDepot[sequence[0]] + distDepot[sequence[c - 1]], 0);
totalDistance.addOperand(routeDistances[d][k]);
// We add service time
LSExpression serviceLambda = model.createLambdaFunction([&](LSExpression j) { return serviceTime[j]; });
LSExpression routeServiceTime = model.sum(sequence, serviceLambda);
// The total distance should not exceed the duration capacity of the truck
// (only if we define such a capacity)
if (routeDurationCapacity[d] > 0) {
model.constraint(routeDistances[d][k] + routeServiceTime <= routeDurationCapacity[d]);
}
}
}
// Objective: minimize the total distance traveled
model.minimize(totalDistance);
model.close();
// Parametrize the solver
localsolver.getParam().setTimeLimit(limit);
localsolver.solve();
}
// Write the solution in a file with the following format:
// - instance, time_limit, total distance
// - for each depot and for each truck in this depot, the customers visited
void writeSolution(const string& fileName, const string& instanceFile, const string& timeLimit) {
ofstream outfile;
outfile.open(fileName.c_str());
outfile << "Instance: " << instanceFile << " ; time_limit: " << timeLimit
<< " ; Objective value: " << totalDistance.getDoubleValue() << endl;
for (int d = 0; d < nbDepots; ++d) {
vector<int> trucksUsed;
for (int k = 0; k < nbTrucksPerDepot; ++k) {
if (customersSequences[d][k].getCollectionValue().count() > 0) {
trucksUsed.push_back(k);
}
}
if (trucksUsed.size() > 0) {
outfile << "Depot " << d + 1 << endl;
for (int k = 0; k < trucksUsed.size(); ++k) {
outfile << "Truck " << (k + 1) << " : ";
LSCollection customersCollection = customersSequences[d][trucksUsed[k]].getCollectionValue();
for (int p = 0; p < customersCollection.count(); ++p) {
outfile << customersCollection[p] + 1 << " ";
}
outfile << endl;
}
outfile << endl;
}
}
}
private:
// Input files following "Cordeau"'s format
void readInputMdvrp(const string& fileName) {
ifstream infile;
infile.open(fileName.c_str());
infile.ignore(); // We ignore the first int of the instance
// Numbers of trucks per depot, customers and depots
infile >> nbTrucksPerDepot;
infile >> nbCustomers;
infile >> nbDepots;
routeDurationCapacity.resize(nbDepots);
truckCapacity.resize(nbDepots);
for (int d = 0; d < nbDepots; ++d) {
infile >> routeDurationCapacity[d];
infile >> truckCapacity[d];
}
// Coordinates X and Y, service time and demand for customers
vector<double> nodesX(nbCustomers);
vector<double> nodesY(nbCustomers);
serviceTimeData.resize(nbCustomers);
demandsData.resize(nbCustomers);
for (int n = 0; n < nbCustomers; ++n) {
int id;
int bin;
infile >> id;
infile >> nodesX[id - 1];
infile >> nodesY[id - 1];
infile >> serviceTimeData[id - 1];
infile >> demandsData[id - 1];
// Ignore the end of the line
infile.ignore(numeric_limits<streamsize>::max(), '\n');
}
// Coordinates X and Y for depots
vector<double> depotX(nbDepots);
vector<double> depotY(nbDepots);
for (int d = 0; d < nbDepots; ++d) {
int id;
int bin;
infile >> id;
infile >> depotX[id - nbCustomers - 1];
infile >> depotY[id - nbCustomers - 1];
// Ignore the end of the line
infile.ignore(numeric_limits<streamsize>::max(), '\n');
}
// Compute the distance matrices
computeDistanceMatrixCustomers(nodesX, nodesY);
computeDistanceWarehouse(depotX, depotY, nodesX, nodesY);
infile.close();
}
// Compute the distance matrix for customers
void computeDistanceMatrixCustomers(const vector<double>& nodesX, const vector<double>& nodesY) {
distanceMatrixCustomersData.resize(nbCustomers);
for (int i = 0; i < nbCustomers; ++i) {
distanceMatrixCustomersData[i].resize(nbCustomers);
}
for (int i = 0; i < nbCustomers; ++i) {
distanceMatrixCustomersData[i][i] = 0;
for (int j = i + 1; j < nbCustomers; ++j) {
double distance = computeDist(nodesX[i], nodesX[j], nodesY[i], nodesY[j]);
distanceMatrixCustomersData[i][j] = distance;
distanceMatrixCustomersData[j][i] = distance;
}
}
}
// Compute the distance matrix for warehouses
void computeDistanceWarehouse(const vector<double>& depotX, const vector<double>& depotY,
const vector<double>& nodesX, const vector<double>& nodesY) {
distanceWarehouseData.resize(nbDepots);
for (int d = 0; d < nbDepots; ++d) {
distanceWarehouseData[d].resize(nbCustomers);
for (int i = 0; i < nbCustomers; ++i) {
distanceWarehouseData[d][i] = computeDist(nodesX[i], depotX[d], nodesY[i], depotY[d]);
}
}
}
// Compute the distance between two points
double computeDist(double xi, double xj, double yi, double yj) { return sqrt(pow(xi - xj, 2) + pow(yi - yj, 2)); }
};
int main(int argc, char** argv) {
if (argc < 2) {
cerr << "Usage: mdvrp inputFile [outputFile] [timeLimit]" << endl;
return 1;
}
const char* instanceFile = argv[1];
const char* outputFile = argc > 2 ? argv[2] : NULL;
const char* strTimeLimit = argc > 3 ? argv[3] : "20";
try {
Mdvrp model;
model.readInstance(instanceFile);
model.solve(atoi(strTimeLimit));
// If we want to write the solution
if (outputFile != NULL)
model.writeSolution(outputFile, instanceFile, strTimeLimit);
} catch (const exception& e) {
cerr << "An error occured: " << e.what() << endl;
}
return 0;
}
- Compilation / Execution (Windows)
- copy %LS_HOME%\bin\localsolvernet.dll .csc Mdvrp.cs /reference:localsolvernet.dllmdvrp instances\p01
using localsolver;
using System;
using System.IO;
using System.Collections.Generic;
public class Mdvrp : IDisposable
{
// LocalSolver
LocalSolver localsolver;
// Number of customers
int nbCustomers;
// Number of depots/warehouses
int nbDepots;
//Number of trucks per depot
int nbTrucksPerDepot;
// Capacity of the trucks per depot
long[] truckCapacity;
// Duration capacity of the trucks per depot
long[] routeDurationCapacity;
// Service time per customer
long[] serviceTimeData;
// Demand per customer
long[] demandsData;
// Distance matrix between customers
double[][] distanceMatrixCustomersData;
// Distances between customers and depots
double[][] distanceWarehouseData;
// Decision variable
LSExpression[][] customersSequences;
// Distance traveled by all the trucks
LSExpression totalDistance;
public Mdvrp()
{
localsolver = new LocalSolver();
}
void ReadInstance(string fileName)
{
readInputMdvrp(fileName);
}
public void Dispose()
{
if (localsolver != null)
localsolver.Dispose();
}
void Solve(int limit)
{
// Declare the optimization model
LSModel model = localsolver.GetModel();
// Sequence of customers visited by each truck
customersSequences = new LSExpression[nbDepots][];
// Vectorization for partition constraint
LSExpression[] customersSequencesConstraint = new LSExpression[nbTrucksPerDepot * nbDepots];
for (int d = 0; d < nbDepots; ++d)
{
customersSequences[d] = new LSExpression[nbTrucksPerDepot];
for (int k = 0; k < nbTrucksPerDepot; ++k)
{
customersSequences[d][k] = model.List(nbCustomers);
customersSequencesConstraint[d * nbTrucksPerDepot + k] = customersSequences[d][k];
}
}
// All customers must be visited by exactly one truck
model.Constraint(model.Partition(customersSequencesConstraint));
// Create LocalSolver arrays to be able to access them with an "at" operator
LSExpression demands = model.Array(demandsData);
LSExpression serviceTime = model.Array(serviceTimeData);
LSExpression distMatrix = model.Array(distanceMatrixCustomersData);
//Distances for each truck from each depot
LSExpression[][] routeDistances = new LSExpression[nbDepots][];
// Total distance traveled
totalDistance = model.Sum();
for (int d = 0; d < nbDepots; ++d)
{
routeDistances[d] = new LSExpression[nbTrucksPerDepot];
LSExpression distDepot = model.Array(distanceWarehouseData[d]);
for (int k = 0; k < nbTrucksPerDepot; ++k)
{
LSExpression sequence = customersSequences[d][k];
LSExpression c = model.Count(sequence);
// The quantity needed in each route must not exceed the truck capacity
LSExpression demandLambda = model.LambdaFunction(j => demands[j]);
LSExpression routeQuantity = model.Sum(sequence, demandLambda);
model.Constraint(routeQuantity <= truckCapacity[d]);
// Distance traveled by truck k of depot d
LSExpression distLambda = model.LambdaFunction(
i => distMatrix[sequence[i - 1], sequence[i]]
);
routeDistances[d][k] =
model.Sum(model.Range(1, c), distLambda)
+ model.If(c > 0, distDepot[sequence[0]] + distDepot[sequence[c - 1]], 0);
totalDistance.AddOperand(routeDistances[d][k]);
// We add service time
LSExpression serviceLambda = model.LambdaFunction(j => serviceTime[j]);
LSExpression routeServiceTime = model.Sum(sequence, serviceLambda);
// The total distance should not exceed the duration capacity of the truck
// (only if we define such a capacity)
if (routeDurationCapacity[d] > 0)
{
model.Constraint(routeDistances[d][k] + routeServiceTime <= routeDurationCapacity[d]);
}
}
}
// Objective: minimize the distance traveled
model.Minimize(totalDistance);
model.Close();
// Parametrize the solver
localsolver.GetParam().SetTimeLimit(limit);
localsolver.Solve();
}
// Write the solution in a file with the following format:
// - instance, time_limit, total distance
// - for each depot and for each truck in this depot, the customers visited
void WriteSolution(string fileName, string instanceFile, string timeLimit)
{
using (StreamWriter outfile = new StreamWriter(fileName))
{
outfile.WriteLine("Instance: " + instanceFile + " ; time_limit: " + timeLimit +
" ; Objective value: " + totalDistance.GetDoubleValue());
for (int d = 0; d < nbDepots; ++d)
{
List<int> trucksUsed = new List<int>();
for (int k = 0; k < nbTrucksPerDepot; ++k)
{
if (customersSequences[d][k].GetCollectionValue().Count() > 0)
{
trucksUsed.Add(k);
}
}
if (trucksUsed.Count > 0)
{
outfile.WriteLine("Depot " + (d + 1));
for (int k = 0; k < trucksUsed.Count; ++k)
{
outfile.Write("Truck " + (k + 1) + " : ");
LSCollection customersCollection = customersSequences[d][trucksUsed[k]].GetCollectionValue();
for (int p = 0; p < customersCollection.Count(); ++p)
outfile.Write((customersCollection[p] + 1) + " ");
outfile.WriteLine();
}
outfile.WriteLine();
}
}
}
}
// Input files following "Cordeau"'s format
private void readInputMdvrp(string fileName)
{
using (StreamReader input = new StreamReader(fileName))
{
string[] splitted;
splitted = input
.ReadLine()
.Split((char[])null, StringSplitOptions.RemoveEmptyEntries);
// Numbers of trucks per depot, customers and depots
nbTrucksPerDepot = int.Parse(splitted[1]);
nbCustomers = int.Parse(splitted[2]);
nbDepots = int.Parse(splitted[3]);
routeDurationCapacity = new long[nbDepots];
truckCapacity = new long[nbDepots];
for (int d = 0; d < nbDepots; ++d)
{
splitted = input
.ReadLine()
.Split((char[])null, StringSplitOptions.RemoveEmptyEntries);
routeDurationCapacity[d] = int.Parse(splitted[0]);
truckCapacity[d] = int.Parse(splitted[1]);
}
// Coordinates X and Y, service time and demand for customers
double[] nodesX = new double[nbCustomers];
double[] nodesY = new double[nbCustomers];
serviceTimeData = new long[nbCustomers];
demandsData = new long[nbCustomers];
for (int n = 0; n < nbCustomers; ++n)
{
splitted = input
.ReadLine()
.Split((char[])null, StringSplitOptions.RemoveEmptyEntries);
nodesX[n] = double.Parse(splitted[1]);
nodesY[n] = double.Parse(splitted[2]);
serviceTimeData[n] = int.Parse(splitted[3]);
demandsData[n] = int.Parse(splitted[4]);
}
// Coordinates X and Y for depots
double[] DepotX = new double[nbDepots];
double[] DepotY = new double[nbDepots];
for (int d = 0; d < nbDepots; ++d)
{
splitted = input
.ReadLine()
.Split((char[])null, StringSplitOptions.RemoveEmptyEntries);
DepotX[d] = double.Parse(splitted[1]);
DepotY[d] = double.Parse(splitted[2]);
}
// Compute the distance matrices
ComputeDistanceMatrixCustomers(nodesX, nodesY);
ComputeDistanceWarehouse(DepotX, DepotY, nodesX, nodesY);
}
}
// Compute the distance matrix for customers
private void ComputeDistanceMatrixCustomers(double[] nodesX, double[] nodesY)
{
distanceMatrixCustomersData = new double[nbCustomers][];
for (int i = 0; i < nbCustomers; ++i)
distanceMatrixCustomersData[i] = new double[nbCustomers];
for (int i = 0; i < nbCustomers; ++i)
{
distanceMatrixCustomersData[i][i] = 0;
for (int j = i + 1; j < nbCustomers; ++j)
{
double dist = ComputeDist(nodesX[i], nodesX[j], nodesY[i], nodesY[j]);
distanceMatrixCustomersData[i][j] = dist;
distanceMatrixCustomersData[j][i] = dist;
}
}
}
// Compute the distance matrix for depots/warehouses
private void ComputeDistanceWarehouse(double[] DepotX, double[] DepotY, double[] nodesX, double[] nodesY)
{
distanceWarehouseData = new double[nbDepots][];
for (int d = 0; d < nbDepots; ++d)
{
distanceWarehouseData[d] = new double[nbCustomers];
for (int i = 0; i < nbCustomers; ++i)
{
distanceWarehouseData[d][i] = ComputeDist(nodesX[i], DepotX[d], nodesY[i], DepotY[d]);
}
}
}
private double ComputeDist(double xi, double xj, double yi, double yj)
{
return Math.Sqrt(Math.Pow(xi - xj, 2) + Math.Pow(yi - yj, 2));
}
public static void Main(string[] args)
{
if (args.Length < 1)
{
Console.WriteLine("Usage: Mdvrp inputFile [outputFile] [timeLimit]");
Environment.Exit(1);
}
string instanceFile = args[0];
string outputFile = args.Length > 1 ? args[1] : null;
string strTimeLimit = args.Length > 2 ? args[2] : "20";
using (Mdvrp model = new Mdvrp())
{
model.ReadInstance(instanceFile);
model.Solve(int.Parse(strTimeLimit));
if (outputFile != null)
model.WriteSolution(outputFile, instanceFile, strTimeLimit);
}
}
}
- Compilation / Execution (Windows)
- javac Mdvrp.java -cp %LS_HOME%\bin\localsolver.jarjava -cp %LS_HOME%\bin\localsolver.jar;. mdvrp instances\p01
- Compilation / Execution (Linux)
- javac Mdvrp.java -cp /opt/localsolver_13_0/bin/localsolver.jarjava -cp /opt/localsolver_13_0/bin/localsolver.jar:. mdvrp instances/p01
import java.util.*;
import java.io.*;
import localsolver.*;
public class Mdvrp {
// LocalSolver
private final LocalSolver localsolver;
// Number of customers
int nbCustomers;
// Number of depots/warehouses
int nbDepots;
// Number of trucks per depot
private int nbTrucksPerDepot;
// Capacity of the trucks per depot
private long[] truckCapacity;
// Duration capacity of the trucks per depot
private long[] routeDurationCapacity;
// Service time per customer
private long[] serviceTimeData;
// Demand per customer
private long[] demandsData;
// Distance matrix between customers
private double[][] distanceMatrixCustomersData;
// Distance matrix between customers and depots
private double[][] distanceWarehouseData;
// Decision variable
private LSExpression[][] customersSequences;
// Distance traveled by all the trucks
private LSExpression totalDistance;
private Mdvrp(LocalSolver localsolver) {
this.localsolver = localsolver;
}
private void readInstance(String fileName) throws IOException {
readInputMdvrp(fileName);
}
private void solve(int limit) {
// Declare the optimization model
LSModel model = localsolver.getModel();
// Sequence of customers visited by each truck
customersSequences = new LSExpression[nbDepots][nbTrucksPerDepot];
// Vectorization for partition constraint
LSExpression[] customersSequencesConstraint = new LSExpression[nbDepots * nbTrucksPerDepot];
// Sequence of customers visited by each truck
for (int d = 0; d < nbDepots; ++d) {
for (int k = 0; k < nbTrucksPerDepot; ++k) {
customersSequences[d][k] = model.listVar(nbCustomers);
customersSequencesConstraint[d * nbTrucksPerDepot + k] = customersSequences[d][k];
}
}
// All customers must be visited by exactly one truck
model.constraint(model.partition(customersSequencesConstraint));
// Create LocalSolver arrays to be able to access them with an "at" operator
LSExpression demands = model.array(demandsData);
LSExpression serviceTime = model.array(serviceTimeData);
LSExpression distMatrix = model.array(distanceMatrixCustomersData);
// Distances for each truck from each depot
LSExpression[][] routeDistances = new LSExpression[nbDepots][nbTrucksPerDepot];
// Total distance traveled
totalDistance = model.sum();
for (int d = 0; d < nbDepots; ++d) {
LSExpression distDepot = model.array(distanceWarehouseData[d]);
for (int k = 0; k < nbTrucksPerDepot; ++k) {
LSExpression sequence = customersSequences[d][k];
LSExpression c = model.count(sequence);
// The quantity needed in each route must not exceed the truck capacity
LSExpression demandLambda = model.lambdaFunction(j -> model.at(demands, j));
LSExpression routeQuantity = model.sum(sequence, demandLambda);
model.constraint(model.leq(routeQuantity, truckCapacity[d]));
// Distance traveled by truck k of depot d
LSExpression distLambda = model
.lambdaFunction(
i -> model.at(distMatrix, model.at(sequence, model.sub(i, 1)), model.at(sequence, i)));
routeDistances[d][k] = model.sum(model.sum(model.range(1, c), distLambda),
model.iif(model.gt(c, 0), model.sum(model.at(distDepot, model.at(sequence, 0)),
model.at(distDepot, model.at(sequence, model.sub(c, 1)))), 0));
totalDistance.addOperand(routeDistances[d][k]);
// We add service time
LSExpression serviceLambda = model.lambdaFunction(j -> model.at(serviceTime, j));
LSExpression routeServiceTime = model.sum(sequence, serviceLambda);
// The total distance should not exceed the duration capacity of the truck (only
// if we define such a capacity)
if (routeDurationCapacity[d] > 0) {
model.constraint(
model.leq(model.sum(routeDistances[d][k], routeServiceTime), routeDurationCapacity[d]));
}
}
}
// Objective: minimize the total distance traveled
model.minimize(totalDistance);
model.close();
// Parametrize the solver
localsolver.getParam().setTimeLimit(limit);
localsolver.solve();
}
// Write the solution in a file with the following format:
// - instance, time_limit, total distance
// - for each depot and for each truck in this depot, the customers visited
private void writeSolution(String fileName, String instanceFile, String timeLimit) throws IOException {
try (PrintWriter outfile = new PrintWriter(fileName)) {
outfile.println("Instance: " + instanceFile + " ; time_limit: " + timeLimit + " ; Objective value: "
+ totalDistance.getDoubleValue());
for (int d = 0; d < nbDepots; ++d) {
ArrayList<Integer> trucksUsed = new ArrayList<Integer>();
for (int k = 0; k < nbTrucksPerDepot; k++) {
if (customersSequences[d][k].getCollectionValue().count() > 0) {
trucksUsed.add(k);
}
}
if (trucksUsed.size() > 0) {
outfile.println("Depot " + (d + 1));
for (int k = 0; k < trucksUsed.size(); ++k) {
outfile.print("Truck " + (k + 1) + " : ");
LSCollection customersCollection = customersSequences[d][trucksUsed.get(k)]
.getCollectionValue();
for (int p = 0; p < customersCollection.count(); ++p)
outfile.print((customersCollection.get(p) + 1) + " ");
outfile.println();
}
outfile.println();
}
}
}
}
// Input files following "Cordeau"'s format
private void readInputMdvrp(String fileName) throws IOException {
try (Scanner input = new Scanner(new File(fileName))) {
String[] splitted;
splitted = input.nextLine().split(" ");
// Numbers of trucks per depot, customers and depots
nbTrucksPerDepot = Integer.parseInt(splitted[1].trim());
nbCustomers = Integer.parseInt(splitted[2].trim());
nbDepots = Integer.parseInt(splitted[3].trim());
routeDurationCapacity = new long[nbDepots];
truckCapacity = new long[nbDepots];
for (int d = 0; d < nbDepots; ++d) {
routeDurationCapacity[d] = input.nextInt();
truckCapacity[d] = input.nextInt();
}
// Coordinates X and Y, service time and demand for customers
double[] nodesX = new double[nbCustomers];
double[] nodesY = new double[nbCustomers];
serviceTimeData = new long[nbCustomers];
demandsData = new long[nbCustomers];
for (int n = 0; n < nbCustomers; ++n) {
input.nextLine();
input.nextInt();
nodesX[n] = input.nextDouble();
nodesY[n] = input.nextDouble();
serviceTimeData[n] = input.nextInt();
demandsData[n] = input.nextInt();
}
// Coordinates X and Y for depots
double[] DepotX = new double[nbDepots];
double[] DepotY = new double[nbDepots];
for (int d = 0; d < nbDepots; ++d) {
input.nextLine();
input.nextInt();
DepotX[d] = input.nextDouble();
DepotY[d] = input.nextDouble();
}
// Compute the distance matrices
ComputeDistanceMatrixCustomers(nodesX, nodesY);
ComputeDistanceWarehouse(DepotX, DepotY, nodesX, nodesY);
}
}
// Compute the distance matrix for customers
private void ComputeDistanceMatrixCustomers(double[] nodesX, double[] nodesY) {
distanceMatrixCustomersData = new double[nbCustomers][nbCustomers];
for (int i = 0; i < nbCustomers; ++i) {
distanceMatrixCustomersData[i][i] = 0;
for (int j = i + 1; j < nbCustomers; ++j) {
double dist = computeDist(nodesX[i], nodesX[j], nodesY[i], nodesY[j]);
distanceMatrixCustomersData[i][j] = dist;
distanceMatrixCustomersData[j][i] = dist;
}
}
}
// Compute the distance matrix for depots/warehouses
private void ComputeDistanceWarehouse(double[] DepotX, double[] DepotY, double[] nodesX, double[] nodesY) {
distanceWarehouseData = new double[nbDepots][nbCustomers];
for (int d = 0; d < nbDepots; ++d) {
for (int i = 0; i < nbCustomers; ++i) {
distanceWarehouseData[d][i] = computeDist(nodesX[i], DepotX[d], nodesY[i], DepotY[d]);
}
}
}
private double computeDist(double xi, double xj, double yi, double yj) {
return Math.sqrt(Math.pow(xi - xj, 2) + Math.pow(yi - yj, 2));
}
public static void main(String[] args) {
if (args.length < 1) {
System.err.println("Usage: java Mdvrp inputFile [outputFile] [timeLimit]");
System.exit(1);
}
try (LocalSolver localsolver = new LocalSolver()) {
String instanceFile = args[0];
String outputFile = args.length > 1 ? args[1] : null;
String strTimeLimit = args.length > 2 ? args[2] : "20";
Mdvrp model = new Mdvrp(localsolver);
model.readInstance(instanceFile);
model.solve(Integer.parseInt(strTimeLimit));
if (outputFile != null) {
model.writeSolution(outputFile, instanceFile, strTimeLimit);
}
} catch (Exception ex) {
System.err.println(ex);
ex.printStackTrace();
System.exit(1);
}
}
}