遗传算法是一种基于自然选择和遗传学原理的优化算法，也很适合解决任务分配问题, 比如达到任务总耗时最短, 比如再兼顾每个工人工作量相对均衡.
下面代码中 TaskAssignmentProblem(单目标优化) 和 BalancedTaskAssignmentProblem(多目标优化) .
package com.example;import java.util.ArrayList;
import java.util.List;import org.uma.jmetal.algorithm.Algorithm;
import org.uma.jmetal.algorithm.examples.AlgorithmRunner;
import org.uma.jmetal.algorithm.multiobjective.nsgaii.NSGAIIBuilder;
import org.uma.jmetal.operator.crossover.impl.IntegerSBXCrossover;
import org.uma.jmetal.operator.mutation.impl.IntegerPolynomialMutation;
import org.uma.jmetal.operator.selection.SelectionOperator;
import org.uma.jmetal.operator.selection.impl.BinaryTournamentSelection;
import org.uma.jmetal.problem.Problem;
import org.uma.jmetal.problem.integerproblem.impl.AbstractIntegerProblem;
import org.uma.jmetal.solution.integersolution.IntegerSolution;
import org.uma.jmetal.util.AbstractAlgorithmRunner;
import org.uma.jmetal.util.JMetalLogger;
import org.uma.jmetal.util.fileoutput.SolutionListOutput;
import org.uma.jmetal.util.fileoutput.impl.DefaultFileOutputContext;public class App {public static void main(String[] args) {// 1. 定义问题Problem<IntegerSolution> problem = null;//problem = new TaskAssignmentProblem();problem = new BalancedTaskAssignmentProblem();// 2. 设置交叉和变异算子 和 设置选择算子// 定义交叉操作: SBX交叉double crossoverProbability = 0.9;double crossoverDistributionIndex = 20.0;var crossover = new IntegerSBXCrossover(crossoverProbability, crossoverDistributionIndex);// 定义变异操作: 多项式变异double mutationProbability = 1.0 / problem.numberOfVariables();double mutationDistributionIndex = 20.0;var mutation = new IntegerPolynomialMutation(mutationProbability, mutationDistributionIndex);// 定义选择操作: 二元竞标赛选择SelectionOperator<List<IntegerSolution>, IntegerSolution> selection = new BinaryTournamentSelection<IntegerSolution>();// 3. 迭代次数和种群大小int populationSize = 100;// 4. 定义算法（NSGA-II） Algorithm<List<IntegerSolution>> algorithm = new NSGAIIBuilder<IntegerSolution>(problem, crossover, mutation,populationSize).setSelectionOperator(selection).setMaxEvaluations(25000).build();// 5. 运行算法AlgorithmRunner algorithmRunner = new AlgorithmRunner.Executor(algorithm).execute();List<IntegerSolution> solutionSet = algorithm.result();long computingTime = algorithmRunner.getComputingTime();JMetalLogger.logger.info("Total execution time: " + computingTime + "ms");// 6. 打印非支配排序结果,每个solution包含决策变量取值和目标函数取值.for (IntegerSolution solution : solutionSet) {JMetalLogger.logger.info("Solution: " + solution);}JMetalLogger.logger.info("Solution Count: " + solutionSet.size());// 7. save to tsv filesnew SolutionListOutput(solutionSet).setVarFileOutputContext(new DefaultFileOutputContext("VAR.csv", ",")).setFunFileOutputContext(new DefaultFileOutputContext("FUN.csv", ",")).print();AbstractAlgorithmRunner.printFinalSolutionSet(solutionSet);}}/** 有4个任务需要3个工人完成, 需要找出最节省时间的任务分配方式.* 目标函数一个, 即所有任务总计耗时* 每个任务由哪个worker完成, 即决策变量, 所以共有4个变量, 变量的取值为 workerId, 范围从0~3 * * 任务和工人的时间Cost矩阵如下: * * 任务 工人A 工人B 工人C* 任务A 2 3 1* 任务B 4 2 3* 任务C 3 4 2* 任务D 1 2 4*/
class TaskAssignmentProblem extends AbstractIntegerProblem {private int evaluationCount;private static final int NUMBER_OF_TASK = 4;private static final int[][] COMPLETION_TIMES = {{ 2, 3, 1 },{ 4, 2, 3 },{ 3, 4, 2 },{ 1, 2, 4 },};public TaskAssignmentProblem() {numberOfObjectives(1);name("TaskAssignmentProblem");// 4 varabiles for 4 tasksvar lowerList = new ArrayList<Integer>();lowerList.add(0); // workerId lower valuelowerList.add(0); // workerId lower valuelowerList.add(0); // workerId lower valuelowerList.add(0); // workerId lower valuevar upperList = new ArrayList<Integer>();upperList.add(2); // workerId upper valueupperList.add(2); // workerId upper valueupperList.add(2); // workerId upper valueupperList.add(2); // workerId upper valuevariableBounds(lowerList, upperList);}@Overridepublic IntegerSolution evaluate(IntegerSolution solution) {int totalTime = 0;for (int i = 0; i < NUMBER_OF_TASK; i++) {var workerId = solution.variables().get(i);totalTime += COMPLETION_TIMES[i][workerId];}solution.objectives()[0] = totalTime;return solution;}
}/** 有4个任务需要3个工人完成, 需要找出最节省时间的任务分配方式, 同时要求每个人的工作量尽量平衡* 目标函数2个, (1)所有任务总计耗时最小, (2)每个人总耗时最小, 即最忙的那个人耗时最小* 每个任务由哪个worker完成, 即决策变量, 所以共有4个变量, 变量的取值为 workerId, 范围从0~3 * * 任务和工人的时间Cost矩阵如下: * * 任务 工人A 工人B 工人C* 任务A 3 2 4* 任务B 5 4 3* 任务C 2 3 2* 任务D 1 4 2*/
class BalancedTaskAssignmentProblem extends AbstractIntegerProblem {private int evaluationCount;private static final int NUMBER_OF_TASK = 4;private static final int NUMBER_OF_WORKER = 3;private static final int[][] COMPLETION_TIMES = {{ 3, 2, 4 },{ 5, 4, 3 },{ 2, 3, 2 },{ 1, 4, 2 },};public BalancedTaskAssignmentProblem() {numberOfObjectives(2);name("TaskAssignmentProblem");// 4 varabiles for 4 tasksvar lowerList = new ArrayList<Integer>();lowerList.add(0); // workerId lower valuelowerList.add(0); // workerId lower valuelowerList.add(0); // workerId lower valuelowerList.add(0); // workerId lower valuevar upperList = new ArrayList<Integer>();upperList.add(2); // workerId upper valueupperList.add(2); // workerId upper valueupperList.add(2); // workerId upper valueupperList.add(2); // workerId upper valuevariableBounds(lowerList, upperList);}@Overridepublic IntegerSolution evaluate(IntegerSolution solution) {int totalTime = 0;int[] workerTime= new int[NUMBER_OF_WORKER] ;//目标函数: 总耗时最小for (int i = 0; i < NUMBER_OF_TASK; i++) {var workerId = solution.variables().get(i);totalTime += COMPLETION_TIMES[i][workerId];workerTime[workerId]+=COMPLETION_TIMES[i][workerId];}solution.objectives()[0] = totalTime;//目标函数:每个人总耗时最小int maxTime=0 ;for (int time : workerTime) {if (maxTime<time){maxTime=time;}}solution.objectives()[1]=maxTime;return solution;}
}
