org.cicirello.search.problems.scheduling.ATCS Maven / Gradle / Ivy

Go to download

Show more of this group Show more artifacts with this name
Show all versions of chips-n-salsa Show documentation

Chips-n-Salsa is a Java library of customizable, hybridizable, iterative, parallel, stochastic, and self-adaptive local search algorithms. The library includes implementations of several stochastic local search algorithms, including simulated annealing, hill climbers, as well as constructive search algorithms such as stochastic sampling. Chips-n-Salsa now also includes genetic algorithms as well as evolutionary algorithms more generally. The library very extensively supports simulated annealing. It includes several classes for representing solutions to a variety of optimization problems. For example, the library includes a BitVector class that implements vectors of bits, as well as classes for representing solutions to problems where we are searching for an optimal vector of integers or reals. For each of the built-in representations, the library provides the most common mutation operators for generating random neighbors of candidate solutions, as well as common crossover operators for use with evolutionary algorithms. Additionally, the library provides extensive support for permutation optimization problems, including implementations of many different mutation operators for permutations, and utilizing the efficiently implemented Permutation class of the JavaPermutationTools (JPT) library. Chips-n-Salsa is customizable, making extensive use of Java's generic types, enabling using the library to optimize other types of representations beyond what is provided in the library. It is hybridizable, providing support for integrating multiple forms of local search (e.g., using a hill climber on a solution generated by simulated annealing), creating hybrid mutation operators (e.g., local search using multiple mutation operators), as well as support for running more than one type of search for the same problem concurrently using multiple threads as a form of algorithm portfolio. Chips-n-Salsa is iterative, with support for multistart metaheuristics, including implementations of several restart schedules for varying the run lengths across the restarts. It also supports parallel execution of multiple instances of the same, or different, stochastic local search algorithms for an instance of a problem to accelerate the search process. The library supports self-adaptive search in a variety of ways, such as including implementations of adaptive annealing schedules for simulated annealing, such as the Modified Lam schedule, implementations of the simpler annealing schedules but which self-tune the initial temperature and other parameters, and restart schedules that adapt to run length.

There is a newer version: 7.0.1

Show newest version

/*
 * Chips-n-Salsa: A library of parallel self-adaptive local search algorithms.
 * Copyright (C) 2002-2020  Vincent A. Cicirello
 *
 * This file is part of Chips-n-Salsa (https://chips-n-salsa.cicirello.org/).
 *
 * Chips-n-Salsa is free software: you can redistribute it and/or modify
 * it under the terms of the GNU General Public License as published by
 * the Free Software Foundation, either version 3 of the License, or
 * (at your option) any later version.
 *
 * Chips-n-Salsa is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with this program.  If not, see .
 */

package org.cicirello.search.problems.scheduling;

import org.cicirello.permutations.Permutation;
import org.cicirello.search.ss.IncrementalEvaluation;
import org.cicirello.search.ss.Partial;

/**
 * This is an implementation of the ATCS (Apparent Tardiness Cost with Setups) heuristic. ATCS is
 * defined as: h(j) = (w[j]/p[j]) exp( -max(0,d[j] - T - p[j]) / (k₁ p̄) ) exp(
 * -s[i][j] / (k₂ s̄)), where w[j] is the weight of job j, p[j] is its processing
 * time, d[j] is the job's due date, T is the current time, and s[i][j] is any setup time of the job
 * if it follows job i. The k₁ and k₂ are parameters that can be tuned based
 * on problem instance characteristics, p̄ is the average processing time of all jobs, and
 * s̄ is the average setup time of all jobs. The p̄ and s̄ are computed once at the
 * start. For a version of the heuristic that dynamically updates these, see the {@link DynamicATCS}
 * class.
 *
 * The constant {@link #MIN_H} defines the minimum value the heuristic will return, preventing
 * h(j)=0 in support of stochastic sampling algorithms for which h(j)=0 is problematic. This
 * implementation returns max( {@link #MIN_H}, h(j)), where {@link #MIN_H} is a small non-zero
 * value.
 *
 * @author Vincent A. Cicirello, https://www.cicirello.org/
 * @version 9.4.2020
 */
public final class ATCS extends WeightedShortestProcessingTime {

  private final double k1;
  private final double k2;
  private final double pAve;
  private final double sAve;

  /**
   * Constructs an ATCS heuristic.
   *
   * @param problem The instance of a scheduling problem that is the target of the heuristic.
   * @param k1 A parameter to the heuristic, which must be positive.
   * @param k2 A parameter to the heuristic, which must be positive.
   * @throws IllegalArgumentException if problem.hasDueDates() returns false.
   * @throws IllegalArgumentException if k ≤ 0.0.
   */
  public ATCS(SingleMachineSchedulingProblem problem, double k1, double k2) {
    super(problem);
    if (!data.hasDueDates()) {
      throw new IllegalArgumentException("This heuristic requires due dates.");
    }
    if (k1 <= 0.0 || k2 <= 0.0) {
      throw new IllegalArgumentException("k1 and k2 must be positive");
    }
    int n = data.numberOfJobs();
    pAve = ((double) sumOfProcessingTimes()) / n;
    sAve = ((double) sumOfSetupTimes()) / (n * n);
    this.k1 = k1;
    this.k2 = k2;
  }

  /**
   * Constructs an ATCS heuristic. Sets the values of k1 and k2 according to the procedure described
   * by the authors of the ATCS heuristic.
   *
   * @param problem The instance of a scheduling problem that is the target of the heuristic.
   * @throws IllegalArgumentException if problem.hasDueDates() returns false.
   */
  public ATCS(SingleMachineSchedulingProblem problem) {
    super(problem);
    if (!data.hasDueDates()) {
      throw new IllegalArgumentException("This heuristic requires due dates.");
    }

    int n = data.numberOfJobs();
    double cv = 0;
    double eta = 0;
    int pSum = sumOfProcessingTimes();
    double cmax = pSum;
    pAve = ((double) pSum) / n;
    sAve = ((double) sumOfSetupTimes()) / (n * n);
    if (sAve > 0) {
      eta = n * sAve / pSum;

      cv = setupVariance(sAve) / (sAve * sAve);
      if (cv < 1E-10) cv = 0;
      else if (cv > 0.3333333333) cv = 0.3333333333;
      final double BETA_MIN = n < 153 ? -0.097 * Math.log(n) + 0.6876 : 0.2;
      final double BETA = cv == 0 ? 1.0 : 1.0 - (1.0 - BETA_MIN) * cv / 0.3333333333;
      cmax += n * sAve * BETA;
    }
    double[] d_stats = computeDueDateStats();
    double R = (d_stats[1] - d_stats[0]) / cmax;
    double tau = 1.0 - d_stats[2] / cmax;
    if (R <= 0.5) k1 = 4.5 + R;
    else if (R <= 2.5) {
      // R should never be > 1 for any realistic instance
      // wider range in if condition is to handle degenerate scheduling
      // problem instance ensuring k1 >= 1.
      k1 = 6 - R - R;
    } else {
      k1 = 1;
    }
    double temp = eta > 0 && tau > 0 ? 0.5 * tau / Math.sqrt(eta) : 1;
    k2 = temp >= 1 ? temp : 1;
  }

  @Override
  public double h(Partial p, int element, IncrementalEvaluation incEval) {
    double value = super.h(p, element, incEval);
    if (value > MIN_H) {
      double num =
          data.getDueDate(element)
              - data.getProcessingTime(element)
              - ((IncrementalTimeCalculator) incEval).currentTime();
      if (num > 0) {
        double denom = k1 * pAve;
        value *= Math.exp(-num / denom);
        if (value <= MIN_H) return MIN_H;
      }
      if (HAS_SETUPS && sAve > 0) {
        num = p.size() == 0 ? data.getSetupTime(element) : data.getSetupTime(p.getLast(), element);
        if (num > 0) {
          double denom = k2 * sAve;
          value *= Math.exp(-num / denom);
          if (value <= MIN_H) return MIN_H;
        }
      }
    }
    return value;
  }

  @Override
  public IncrementalEvaluation createIncrementalEvaluation() {
    return new IncrementalTimeCalculator();
  }

  private double[] computeDueDateStats() {
    int min = data.getDueDate(0);
    int max = min;
    int sum = min;
    int n = data.numberOfJobs();
    for (int i = 1; i < n; i++) {
      int d = data.getDueDate(i);
      if (d < min) min = d;
      else if (d > max) max = d;
      sum += d;
    }
    return new double[] {min, max, ((double) sum) / n};
  }

  private double setupVariance(double mean) {
    int n = data.numberOfJobs();
    double total = 0;
    for (int i = 0; i < n; i++) {
      for (int j = 0; j < n; j++) {
        int s = data.getSetupTime(i, j);
        total += (s * s);
      }
    }
    return total / (n * n) - mean * mean;
  }

  /* package-private: here to support testing only */
  final double getSetupAverage() {
    return sAve;
  }
}