[6448] | 1 | #region License Information
|
---|
| 2 | /* HeuristicLab
|
---|
| 3 | * Copyright (C) 2002-2011 Heuristic and Evolutionary Algorithms Laboratory (HEAL)
|
---|
| 4 | *
|
---|
| 5 | * This file is part of HeuristicLab.
|
---|
| 6 | *
|
---|
| 7 | * HeuristicLab is free software: you can redistribute it and/or modify
|
---|
| 8 | * it under the terms of the GNU General Public License as published by
|
---|
| 9 | * the Free Software Foundation, either version 3 of the License, or
|
---|
| 10 | * (at your option) any later version.
|
---|
| 11 | *
|
---|
| 12 | * HeuristicLab is distributed in the hope that it will be useful,
|
---|
| 13 | * but WITHOUT ANY WARRANTY; without even the implied warranty of
|
---|
| 14 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
|
---|
| 15 | * GNU General Public License for more details.
|
---|
| 16 | *
|
---|
| 17 | * You should have received a copy of the GNU General Public License
|
---|
| 18 | * along with HeuristicLab. If not, see <http://www.gnu.org/licenses/>.
|
---|
| 19 | */
|
---|
| 20 | #endregion
|
---|
| 21 |
|
---|
| 22 | using HeuristicLab.Common;
|
---|
| 23 | using HeuristicLab.Core;
|
---|
| 24 | using HeuristicLab.Persistence.Default.CompositeSerializers.Storable;
|
---|
| 25 | using System.Collections.Generic;
|
---|
| 26 | using HeuristicLab.Data;
|
---|
| 27 | using System;
|
---|
[6455] | 28 | using HeuristicLab.Parameters;
|
---|
[6448] | 29 |
|
---|
| 30 | namespace HeuristicLab.Problems.VehicleRouting.Encodings.Potvin {
|
---|
[6455] | 31 | [Item("PotvinInsertionBasedCrossover", "The IBX crossover for VRP representations. It is implemented as described in Berger, J and Solois, M and Begin, R (1998). A hybrid genetic algorithm for the vehicle routing problem with time windows. LNCS 1418. Springer, London 114-127.")]
|
---|
[6448] | 32 | [StorableClass]
|
---|
| 33 | public sealed class PotvinInsertionBasedCrossover : PotvinCrossover {
|
---|
[6455] | 34 | public IValueParameter<IntValue> Length {
|
---|
| 35 | get { return (IValueParameter<IntValue>)Parameters["Length"]; }
|
---|
| 36 | }
|
---|
| 37 |
|
---|
[6448] | 38 | [StorableConstructor]
|
---|
| 39 | private PotvinInsertionBasedCrossover(bool deserializing) : base(deserializing) { }
|
---|
| 40 | private PotvinInsertionBasedCrossover(PotvinInsertionBasedCrossover original, Cloner cloner)
|
---|
| 41 | : base(original, cloner) {
|
---|
| 42 | }
|
---|
| 43 | public override IDeepCloneable Clone(Cloner cloner) {
|
---|
| 44 | return new PotvinInsertionBasedCrossover(this, cloner);
|
---|
| 45 | }
|
---|
| 46 | public PotvinInsertionBasedCrossover()
|
---|
[6455] | 47 | : base() {
|
---|
| 48 | Parameters.Add(new ValueParameter<IntValue>("Length", "The maximum length of the replaced route.", new IntValue(1)));
|
---|
| 49 | }
|
---|
[6448] | 50 |
|
---|
| 51 | protected static int SelectRandomTourBiasedByLength(IRandom random, PotvinEncoding individual) {
|
---|
| 52 | int tourIndex = -1;
|
---|
| 53 |
|
---|
| 54 | double sum = 0.0;
|
---|
| 55 | double[] probabilities = new double[individual.Tours.Count];
|
---|
| 56 | for (int i = 0; i < individual.Tours.Count; i++) {
|
---|
| 57 | probabilities[i] = 1.0 / ((double)individual.Tours[i].Cities.Count / (double)individual.Cities);
|
---|
| 58 | sum += probabilities[i];
|
---|
| 59 | }
|
---|
| 60 |
|
---|
| 61 | for (int i = 0; i < probabilities.Length; i++)
|
---|
| 62 | probabilities[i] = probabilities[i] / sum;
|
---|
| 63 |
|
---|
| 64 | double rand = random.NextDouble();
|
---|
| 65 | double cumulatedProbabilities = 0.0;
|
---|
| 66 | int index = 0;
|
---|
| 67 | while (tourIndex == -1 && index < probabilities.Length) {
|
---|
| 68 | if (cumulatedProbabilities <= rand && rand <= cumulatedProbabilities + probabilities[index])
|
---|
| 69 | tourIndex = index;
|
---|
| 70 |
|
---|
| 71 | cumulatedProbabilities += probabilities[index];
|
---|
| 72 | index++;
|
---|
| 73 | }
|
---|
| 74 |
|
---|
| 75 | return tourIndex;
|
---|
| 76 | }
|
---|
| 77 |
|
---|
| 78 | private double CalculateCentroidDistance(Tour t1, Tour t2, DoubleMatrix coordinates) {
|
---|
| 79 | double xSum = 0;
|
---|
| 80 | double ySum = 0;
|
---|
| 81 | double c1X, c1Y, c2X, c2Y;
|
---|
| 82 |
|
---|
| 83 | for (int i = 0; i < t1.Cities.Count; i++) {
|
---|
| 84 | xSum += coordinates[t1.Cities[i], 0];
|
---|
| 85 | ySum += coordinates[t1.Cities[i], 0];
|
---|
| 86 | }
|
---|
| 87 | c1X = xSum / t1.Cities.Count;
|
---|
| 88 | c1Y = ySum / t1.Cities.Count;
|
---|
| 89 |
|
---|
| 90 | for (int i = 0; i < t2.Cities.Count; i++) {
|
---|
| 91 | xSum += coordinates[t2.Cities[i], 0];
|
---|
| 92 | ySum += coordinates[t2.Cities[i], 0];
|
---|
| 93 | }
|
---|
| 94 | c2X = xSum / t1.Cities.Count;
|
---|
| 95 | c2Y = ySum / t1.Cities.Count;
|
---|
| 96 |
|
---|
| 97 | return Math.Sqrt(
|
---|
| 98 | Math.Pow(c1X - c2X, 2) +
|
---|
| 99 | Math.Pow(c1Y - c2Y, 2));
|
---|
| 100 | }
|
---|
| 101 |
|
---|
| 102 | private double CalculateMeanCentroidDistance(Tour t1, IList<Tour> tours, DoubleMatrix coordinates) {
|
---|
| 103 | double sum = 0;
|
---|
| 104 |
|
---|
| 105 | for (int i = 0; i < tours.Count; i++) {
|
---|
| 106 | sum += CalculateCentroidDistance(t1, tours[i], coordinates);
|
---|
| 107 | }
|
---|
| 108 |
|
---|
| 109 | return sum / tours.Count;
|
---|
| 110 | }
|
---|
| 111 |
|
---|
| 112 | private int SelectCityBiasedByNeighborDistance(IRandom random, Tour tour, DistanceMatrix distMatrix) {
|
---|
| 113 | int cityIndex = -1;
|
---|
| 114 |
|
---|
| 115 | double sum = 0.0;
|
---|
| 116 | double[] probabilities = new double[tour.Cities.Count];
|
---|
| 117 | for (int i = 0; i < tour.Cities.Count; i++) {
|
---|
| 118 | int next = i + 1;
|
---|
| 119 | if (next >= tour.Cities.Count)
|
---|
| 120 | next = 0;
|
---|
| 121 | else
|
---|
| 122 | next = tour.Cities[next];
|
---|
| 123 | double distance = VRPUtilities.GetDistance(
|
---|
| 124 | tour.Cities[i], next, distMatrix);
|
---|
| 125 |
|
---|
| 126 | int prev = i - 1;
|
---|
| 127 | if (prev < 0)
|
---|
| 128 | prev = 0;
|
---|
| 129 | else
|
---|
| 130 | prev = tour.Cities[prev];
|
---|
| 131 | distance += VRPUtilities.GetDistance(
|
---|
| 132 | tour.Cities[i], prev, distMatrix);
|
---|
| 133 |
|
---|
| 134 | probabilities[i] = distance;
|
---|
| 135 | sum += probabilities[i];
|
---|
| 136 | }
|
---|
| 137 |
|
---|
| 138 | for (int i = 0; i < probabilities.Length; i++)
|
---|
| 139 | probabilities[i] = probabilities[i] / sum;
|
---|
| 140 |
|
---|
| 141 | double rand = random.NextDouble();
|
---|
| 142 | double cumulatedProbabilities = 0.0;
|
---|
| 143 | int index = 0;
|
---|
| 144 | while (cityIndex == -1 && index < probabilities.Length) {
|
---|
| 145 | if (cumulatedProbabilities <= rand && rand <= cumulatedProbabilities + probabilities[index])
|
---|
| 146 | cityIndex = index;
|
---|
| 147 |
|
---|
| 148 | cumulatedProbabilities += probabilities[index];
|
---|
| 149 | index++;
|
---|
| 150 | }
|
---|
| 151 |
|
---|
| 152 | return cityIndex;
|
---|
| 153 | }
|
---|
| 154 |
|
---|
| 155 | private bool FindRouteInsertionPlace(
|
---|
| 156 | Tour tour,
|
---|
| 157 | DoubleArray dueTimeArray,
|
---|
| 158 | DoubleArray serviceTimeArray, DoubleArray readyTimeArray, DoubleArray demandArray, DoubleValue capacity,
|
---|
[6459] | 159 | DistanceMatrix distMatrix, int city, bool allowInfeasible, out int place) {
|
---|
[6448] | 160 | place = -1;
|
---|
| 161 | bool bestFeasible = false;
|
---|
| 162 | double minDetour = 0;
|
---|
| 163 |
|
---|
| 164 | for (int i = 0; i <= tour.Cities.Count; i++) {
|
---|
| 165 | double length = tour.GetLength(distMatrix);
|
---|
| 166 |
|
---|
| 167 | tour.Cities.Insert(i, city);
|
---|
| 168 |
|
---|
| 169 | bool feasible = tour.Feasible(dueTimeArray, serviceTimeArray, readyTimeArray, demandArray,
|
---|
| 170 | capacity, distMatrix);
|
---|
| 171 |
|
---|
[6459] | 172 | if ((!allowInfeasible && feasible) || (allowInfeasible && (!bestFeasible || feasible))) {
|
---|
[6448] | 173 | double newLength = tour.GetLength(distMatrix);
|
---|
| 174 | double detour = newLength - length;
|
---|
| 175 |
|
---|
[6459] | 176 | if (place <= 0 || detour < minDetour ||
|
---|
| 177 | (allowInfeasible && ((!(bestFeasible && !feasible)) && detour < minDetour || (feasible && !bestFeasible)))) {
|
---|
[6448] | 178 | place = i;
|
---|
| 179 | minDetour = detour;
|
---|
| 180 |
|
---|
| 181 | if (feasible)
|
---|
| 182 | bestFeasible = true;
|
---|
| 183 | }
|
---|
| 184 | }
|
---|
| 185 |
|
---|
| 186 | tour.Cities.RemoveAt(i);
|
---|
| 187 | }
|
---|
| 188 |
|
---|
| 189 | return place >= 0;
|
---|
| 190 | }
|
---|
| 191 |
|
---|
| 192 | protected override PotvinEncoding Crossover(IRandom random, PotvinEncoding parent1, PotvinEncoding parent2) {
|
---|
| 193 | PotvinEncoding child = new PotvinEncoding();
|
---|
| 194 | bool success = true;
|
---|
| 195 |
|
---|
| 196 | BoolValue useDistanceMatrix = UseDistanceMatrixParameter.ActualValue;
|
---|
| 197 | DoubleMatrix coordinates = CoordinatesParameter.ActualValue;
|
---|
| 198 | DistanceMatrix distMatrix = VRPUtilities.GetDistanceMatrix(coordinates, DistanceMatrixParameter, useDistanceMatrix);
|
---|
| 199 | DoubleArray dueTime = DueTimeParameter.ActualValue;
|
---|
| 200 | DoubleArray readyTime = ReadyTimeParameter.ActualValue;
|
---|
| 201 | DoubleArray serviceTime = ServiceTimeParameter.ActualValue;
|
---|
| 202 | DoubleArray demand = DemandParameter.ActualValue;
|
---|
| 203 | DoubleValue capacity = CapacityParameter.ActualValue;
|
---|
| 204 |
|
---|
[6459] | 205 | bool allowInfeasible = AllowInfeasibleSolutions.Value.Value;
|
---|
| 206 |
|
---|
[6448] | 207 | List<Tour> R1 = new List<Tour>();
|
---|
| 208 | PotvinEncoding p1Clone = parent1.Clone() as PotvinEncoding;
|
---|
| 209 |
|
---|
[6455] | 210 | int length = Math.Min(Length.Value.Value, parent1.Tours.Count) + 1;
|
---|
| 211 | int k = random.Next(1, length);
|
---|
[6448] | 212 | for (int i = 0; i < k; i++) {
|
---|
| 213 | int index = SelectRandomTourBiasedByLength(random, p1Clone);
|
---|
| 214 | R1.Add(p1Clone.Tours[index]);
|
---|
| 215 | p1Clone.Tours.RemoveAt(index);
|
---|
| 216 | }
|
---|
| 217 |
|
---|
| 218 | foreach (Tour r1 in R1) {
|
---|
| 219 | List<int> R2 = new List<int>();
|
---|
| 220 |
|
---|
| 221 | double r = CalculateMeanCentroidDistance(r1, parent2.Tours, coordinates);
|
---|
| 222 | foreach (Tour tour in parent2.Tours) {
|
---|
| 223 | if (CalculateCentroidDistance(r1, tour, coordinates) <= r) {
|
---|
| 224 | R2.AddRange(tour.Cities);
|
---|
| 225 | }
|
---|
| 226 | }
|
---|
| 227 |
|
---|
| 228 | Tour childTour = new Tour();
|
---|
| 229 | childTour.Cities.AddRange(r1.Cities);
|
---|
| 230 |
|
---|
| 231 | //DESTROY - remove cities from r1
|
---|
| 232 | int removed = random.Next(1, r1.Cities.Count + 1);
|
---|
| 233 | for (int i = 0; i < removed; i++) {
|
---|
| 234 | childTour.Cities.RemoveAt(SelectCityBiasedByNeighborDistance(random, childTour, distMatrix));
|
---|
| 235 | }
|
---|
| 236 |
|
---|
| 237 | //REPAIR - add cities from R2
|
---|
| 238 | bool insertSuccess = true;
|
---|
| 239 | int maxCount = random.Next(1, Math.Min(5, R2.Count));
|
---|
| 240 | int count = 0;
|
---|
| 241 |
|
---|
| 242 | while (count < maxCount && R2.Count != 0) {
|
---|
| 243 | PotvinEncoding newChild = child.Clone() as PotvinEncoding;
|
---|
| 244 | newChild.Tours.Add(childTour);
|
---|
| 245 |
|
---|
| 246 | int index = random.Next(R2.Count);
|
---|
| 247 | int city = R2[index];
|
---|
| 248 | R2.RemoveAt(index);
|
---|
| 249 |
|
---|
| 250 | int place = -1;
|
---|
[6459] | 251 | if(FindRouteInsertionPlace(childTour, dueTime, serviceTime, readyTime,
|
---|
| 252 | demand, capacity, distMatrix, city, allowInfeasible, out place)) {
|
---|
[6448] | 253 | childTour.Cities.Insert(place, city);
|
---|
| 254 |
|
---|
[6459] | 255 | if (!Repair(random, child, childTour, distMatrix, dueTime, readyTime, serviceTime, demand, capacity, allowInfeasible)) {
|
---|
[6448] | 256 | childTour.Cities.RemoveAt(place);
|
---|
| 257 | insertSuccess = false;
|
---|
| 258 | } else {
|
---|
| 259 | count++;
|
---|
| 260 | }
|
---|
| 261 | }
|
---|
| 262 | }
|
---|
| 263 |
|
---|
| 264 | child.Tours.Add(childTour);
|
---|
[6459] | 265 | if (!Repair(random, child, childTour, distMatrix, dueTime, readyTime, serviceTime, demand, capacity, allowInfeasible)) {
|
---|
[6455] | 266 | /*success = false;
|
---|
| 267 | break;*/
|
---|
[6448] | 268 | }
|
---|
| 269 | }
|
---|
| 270 |
|
---|
| 271 | if (success) {
|
---|
| 272 | for (int i = 0; i < p1Clone.Tours.Count; i++) {
|
---|
| 273 | Tour childTour = p1Clone.Tours[i].Clone() as Tour;
|
---|
| 274 | child.Tours.Add(childTour);
|
---|
[6459] | 275 | if (!Repair(random, child, childTour, distMatrix, dueTime, readyTime, serviceTime, demand, capacity, allowInfeasible)) {
|
---|
[6455] | 276 | /*success = false;
|
---|
| 277 | break;*/
|
---|
[6448] | 278 | }
|
---|
| 279 | }
|
---|
| 280 | }
|
---|
| 281 |
|
---|
| 282 | if (success) {
|
---|
| 283 | //route unrouted customers
|
---|
| 284 | for (int i = 1; i <= parent1.Cities; i++) {
|
---|
| 285 | if (FindRoute(child, i) == null)
|
---|
| 286 | child.Unrouted.Add(i);
|
---|
| 287 | }
|
---|
| 288 |
|
---|
[6459] | 289 | if (!RouteUnrouted(child, distMatrix, dueTime, readyTime, serviceTime, demand, capacity, allowInfeasible)) {
|
---|
[6448] | 290 | success = false;
|
---|
| 291 | }
|
---|
| 292 | }
|
---|
| 293 |
|
---|
[6459] | 294 | if (success || allowInfeasible)
|
---|
[6448] | 295 | return child;
|
---|
| 296 | else {
|
---|
[6455] | 297 | if (random.NextDouble() < 0.5)
|
---|
[6448] | 298 | return parent1.Clone() as PotvinEncoding;
|
---|
| 299 | else
|
---|
[6455] | 300 | return parent2.Clone() as PotvinEncoding;
|
---|
[6448] | 301 | }
|
---|
| 302 | }
|
---|
| 303 | }
|
---|
| 304 | }
|
---|