此页面由 Cloud Translation API 翻译。

解决多个背包问题

本部分介绍如何解决多个背包的背包问题 MIP 求解器和 CP-SAT 求解器。在这种情况下，通常将容器称为“bin”，

下例展示了如何找到将商品打包到 5 个分箱中的最佳方式。

示例

与上一个示例一样，您首先构建了具有不同权重和值的项目的集合。问题在于，这些项目的子集会分成五个箱，每个箱的最大容量为 100；使打包总值达到最大值。

以下部分介绍了可解决此问题的程序部分。有关完整的程序，请参阅完整程序。

MIP 解决方案

以下部分介绍了如何使用 MPSolver 封装容器。

导入库

以下代码会导入所需的库。

Python

from ortools.linear_solver import pywraplp

C++

#include <iostream>
#include <memory>
#include <numeric>
#include <vector>

#include "absl/strings/str_format.h"
#include "ortools/linear_solver/linear_expr.h"
#include "ortools/linear_solver/linear_solver.h"

Java

import com.google.ortools.Loader;
import com.google.ortools.linearsolver.MPConstraint;
import com.google.ortools.linearsolver.MPObjective;
import com.google.ortools.linearsolver.MPSolver;
import com.google.ortools.linearsolver.MPVariable;
import java.util.stream.IntStream;

C#

using System;
using System.Collections.Generic;
using System.Linq;
using Google.OrTools.LinearSolver;

创建数据

以下代码会为问题创建数据。

Python

data = {}
data["weights"] = [48, 30, 42, 36, 36, 48, 42, 42, 36, 24, 30, 30, 42, 36, 36]
data["values"] = [10, 30, 25, 50, 35, 30, 15, 40, 30, 35, 45, 10, 20, 30, 25]
assert len(data["weights"]) == len(data["values"])
data["num_items"] = len(data["weights"])
data["all_items"] = range(data["num_items"])

data["bin_capacities"] = [100, 100, 100, 100, 100]
data["num_bins"] = len(data["bin_capacities"])
data["all_bins"] = range(data["num_bins"])

C++

const std::vector<int> weights = {
    {48, 30, 42, 36, 36, 48, 42, 42, 36, 24, 30, 30, 42, 36, 36}};
const std::vector<int> values = {
    {10, 30, 25, 50, 35, 30, 15, 40, 30, 35, 45, 10, 20, 30, 25}};
const int num_items = weights.size();
std::vector<int> all_items(num_items);
std::iota(all_items.begin(), all_items.end(), 0);

const std::vector<int> bin_capacities = {{100, 100, 100, 100, 100}};
const int num_bins = bin_capacities.size();
std::vector<int> all_bins(num_bins);
std::iota(all_bins.begin(), all_bins.end(), 0);

Java

final double[] weights = {48, 30, 42, 36, 36, 48, 42, 42, 36, 24, 30, 30, 42, 36, 36};
final double[] values = {10, 30, 25, 50, 35, 30, 15, 40, 30, 35, 45, 10, 20, 30, 25};
final int numItems = weights.length;
final int[] allItems = IntStream.range(0, numItems).toArray();

final double[] binCapacities = {100, 100, 100, 100, 100};
final int numBins = binCapacities.length;
final int[] allBins = IntStream.range(0, numBins).toArray();

C#

double[] Weights = { 48, 30, 42, 36, 36, 48, 42, 42, 36, 24, 30, 30, 42, 36, 36 };
double[] Values = { 10, 30, 25, 50, 35, 30, 15, 40, 30, 35, 45, 10, 20, 30, 25 };
int NumItems = Weights.Length;
int[] allItems = Enumerable.Range(0, NumItems).ToArray();

double[] BinCapacities = { 100, 100, 100, 100, 100 };
int NumBins = BinCapacities.Length;
int[] allBins = Enumerable.Range(0, NumBins).ToArray();

这些数据包括以下内容：

weights：包含项权重的向量。
values：包含项值的向量。
capacities：包含分箱容量的向量。

在此示例中，所有分箱的容量都相同，但无需为 true 。

声明 MIP 求解器

以下代码声明了 MIP 求解器。

Python

solver = pywraplp.Solver.CreateSolver("SCIP")
if solver is None:
    print("SCIP solver unavailable.")
    return

C++

  std::unique_ptr<MPSolver> solver(MPSolver::CreateSolver("SCIP"));
  if (!solver) {
    LOG(WARNING) << "SCIP solver unavailable.";
    return;
  }

Java

// Create the linear solver with the SCIP backend.
MPSolver solver = MPSolver.createSolver("SCIP");
if (solver == null) {
  System.out.println("Could not create solver SCIP");
  return;
}

C#

// Create the linear solver with the SCIP backend.
Solver solver = Solver.CreateSolver("SCIP");
if (solver is null)
{
    return;
}

创建变量

以下代码会为问题创建变量。

Python

# x[i, b] = 1 if item i is packed in bin b.
x = {}
for i in data["all_items"]:
    for b in data["all_bins"]:
        x[i, b] = solver.BoolVar(f"x_{i}_{b}")

C++

// x[i][b] = 1 if item i is packed in bin b.
std::vector<std::vector<const MPVariable*>> x(
    num_items, std::vector<const MPVariable*>(num_bins));
for (int i : all_items) {
  for (int b : all_bins) {
    x[i][b] = solver->MakeBoolVar(absl::StrFormat("x_%d_%d", i, b));
  }
}

Java

MPVariable[][] x = new MPVariable[numItems][numBins];
for (int i : allItems) {
  for (int b : allBins) {
    x[i][b] = solver.makeBoolVar("x_" + i + "_" + b);
  }
}

C#

Variable[,] x = new Variable[NumItems, NumBins];
foreach (int i in allItems)
{
    foreach (int b in allBins)
    {
        x[i, b] = solver.MakeBoolVar($"x_{i}_{b}");
    }
}

每个 x[(i, j)] 都是 0-1 的变量，其中 i 为一项，j 为分箱。在解：如果将 i 项放在 bin j 中，x[(i, j)] 将为 1，0 将为 0 否则。

定义限制条件

以下代码定义了该问题的约束条件：

Python

# Each item is assigned to at most one bin.
for i in data["all_items"]:
    solver.Add(sum(x[i, b] for b in data["all_bins"]) <= 1)

# The amount packed in each bin cannot exceed its capacity.
for b in data["all_bins"]:
    solver.Add(
        sum(x[i, b] * data["weights"][i] for i in data["all_items"])
        <= data["bin_capacities"][b]
    )

C++

// Each item is assigned to at most one bin.
for (int i : all_items) {
  LinearExpr sum;
  for (int b : all_bins) {
    sum += x[i][b];
  }
  solver->MakeRowConstraint(sum <= 1.0);
}
// The amount packed in each bin cannot exceed its capacity.
for (int b : all_bins) {
  LinearExpr bin_weight;
  for (int i : all_items) {
    bin_weight += LinearExpr(x[i][b]) * weights[i];
  }
  solver->MakeRowConstraint(bin_weight <= bin_capacities[b]);
}

Java

// Each item is assigned to at most one bin.
for (int i : allItems) {
  MPConstraint constraint = solver.makeConstraint(0, 1, "");
  for (int b : allBins) {
    constraint.setCoefficient(x[i][b], 1);
  }
}

// The amount packed in each bin cannot exceed its capacity.
for (int b : allBins) {
  MPConstraint constraint = solver.makeConstraint(0, binCapacities[b], "");
  for (int i : allItems) {
    constraint.setCoefficient(x[i][b], weights[i]);
  }
}

C#

// Each item is assigned to at most one bin.
foreach (int i in allItems)
{
    Constraint constraint = solver.MakeConstraint(0, 1, "");
    foreach (int b in allBins)
    {
        constraint.SetCoefficient(x[i, b], 1);
    }
}

// The amount packed in each bin cannot exceed its capacity.
foreach (int b in allBins)
{
    Constraint constraint = solver.MakeConstraint(0, BinCapacities[b], "");
    foreach (int i in allItems)
    {
        constraint.SetCoefficient(x[i, b], Weights[i]);
    }
}

限制条件如下：

每件商品最多只能放置在 1 个箱中。此限制条件由要求所有分箱的x[i, j]总和 j 小于或等于更改为 1。
每个垃圾桶中装卸的总重量不得超过其容量。这个约束条件是通过要求放入分箱中的物品的权重总和来设置的 j 应小于或等于垃圾桶的容量。

定义目标

以下代码定义了问题的目标函数，包装商品的总价值。

Python

# Maximize total value of packed items.
objective = solver.Objective()
for i in data["all_items"]:
    for b in data["all_bins"]:
        objective.SetCoefficient(x[i, b], data["values"][i])
objective.SetMaximization()

C++

// Maximize total value of packed items.
MPObjective* const objective = solver->MutableObjective();
LinearExpr objective_value;
for (int i : all_items) {
  for (int b : all_bins) {
    objective_value += LinearExpr(x[i][b]) * values[i];
  }
}
objective->MaximizeLinearExpr(objective_value);

Java

// Maximize total value of packed items.
MPObjective objective = solver.objective();
for (int i : allItems) {
  for (int b : allBins) {
    objective.setCoefficient(x[i][b], values[i]);
  }
}
objective.setMaximization();

C#

Objective objective = solver.Objective();
foreach (int i in allItems)
{
    foreach (int b in allBins)
    {
        objective.SetCoefficient(x[i, b], Values[i]);
    }
}
objective.SetMaximization();

请注意，x[i, j] * data['values'][i] 会将项 i 的值添加到目标 (j)。如果 i 未放置在任何分箱中，则其对目标没有帮助。

调用求解器

以下代码会调用求解器。

Python

print(f"Solving with {solver.SolverVersion()}")
status = solver.Solve()

C++

const MPSolver::ResultStatus result_status = solver->Solve();

Java

final MPSolver.ResultStatus status = solver.solve();

C#

Solver.ResultStatus resultStatus = solver.Solve();

输出解决方案

以下代码会输出该问题的解决方案。

Python

if status == pywraplp.Solver.OPTIMAL:
    print(f"Total packed value: {objective.Value()}")
    total_weight = 0
    for b in data["all_bins"]:
        print(f"Bin {b}")
        bin_weight = 0
        bin_value = 0
        for i in data["all_items"]:
            if x[i, b].solution_value() > 0:
                print(
                    f"Item {i} weight: {data['weights'][i]} value:"
                    f" {data['values'][i]}"
                )
                bin_weight += data["weights"][i]
                bin_value += data["values"][i]
        print(f"Packed bin weight: {bin_weight}")
        print(f"Packed bin value: {bin_value}\n")
        total_weight += bin_weight
    print(f"Total packed weight: {total_weight}")
else:
    print("The problem does not have an optimal solution.")

C++

if (result_status == MPSolver::OPTIMAL) {
  LOG(INFO) << "Total packed value: " << objective->Value();
  double total_weight = 0.0;
  for (int b : all_bins) {
    LOG(INFO) << "Bin " << b;
    double bin_weight = 0.0;
    double bin_value = 0.0;
    for (int i : all_items) {
      if (x[i][b]->solution_value() > 0) {
        LOG(INFO) << "Item " << i << " weight: " << weights[i]
                  << " value: " << values[i];
        bin_weight += weights[i];
        bin_value += values[i];
      }
    }
    LOG(INFO) << "Packed bin weight: " << bin_weight;
    LOG(INFO) << "Packed bin value: " << bin_value;
    total_weight += bin_weight;
  }
  LOG(INFO) << "Total packed weight: " << total_weight;
} else {
  LOG(INFO) << "The problem does not have an optimal solution.";
}

Java

// Check that the problem has an optimal solution.
if (status == MPSolver.ResultStatus.OPTIMAL) {
  System.out.println("Total packed value: " + objective.value());
  double totalWeight = 0;
  for (int b : allBins) {
    double binWeight = 0;
    double binValue = 0;
    System.out.println("Bin " + b);
    for (int i : allItems) {
      if (x[i][b].solutionValue() == 1) {
        System.out.println("Item " + i + " weight: " + weights[i] + " value: " + values[i]);
        binWeight += weights[i];
        binValue += values[i];
      }
    }
    System.out.println("Packed bin weight: " + binWeight);
    System.out.println("Packed bin value: " + binValue);
    totalWeight += binWeight;
  }
  System.out.println("Total packed weight: " + totalWeight);
} else {
  System.err.println("The problem does not have an optimal solution.");
}

C#

// Check that the problem has an optimal solution.
if (resultStatus == Solver.ResultStatus.OPTIMAL)
{
    Console.WriteLine($"Total packed value: {solver.Objective().Value()}");
    double TotalWeight = 0.0;
    foreach (int b in allBins)
    {
        double BinWeight = 0.0;
        double BinValue = 0.0;
        Console.WriteLine("Bin " + b);
        foreach (int i in allItems)
        {
            if (x[i, b].SolutionValue() == 1)
            {
                Console.WriteLine($"Item {i} weight: {Weights[i]} values: {Values[i]}");
                BinWeight += Weights[i];
                BinValue += Values[i];
            }
        }
        Console.WriteLine("Packed bin weight: " + BinWeight);
        Console.WriteLine("Packed bin value: " + BinValue);
        TotalWeight += BinWeight;
    }
    Console.WriteLine("Total packed weight: " + TotalWeight);
}
else
{
    Console.WriteLine("The problem does not have an optimal solution!");
}

对于每个分箱，该代码会显示放入该分箱的物品，以及分箱的总值和重量。该代码还会显示总价值和包装物品的重量。

运行该程序时，它会显示以下输出。

Total packed value: 395.0
Bin  0

Item 3 - weight: 36  value: 50
Item 13 - weight: 36  value: 30
Packed bin weight: 72
Packed bin value: 80

Bin  1

Item 5 - weight: 48  value: 30
Item 7 - weight: 42  value: 40
Packed bin weight: 90
Packed bin value: 70

Bin  2

Item 1 - weight: 30  value: 30
Item 10 - weight: 30  value: 45
Item 14 - weight: 36  value: 25
Packed bin weight: 96
Packed bin value: 100

Bin  3

Item 2 - weight: 42  value: 25
Item 12 - weight: 42  value: 20
Packed bin weight: 84
Packed bin value: 45

Bin  4

Item 4 - weight: 36  value: 35
Item 8 - weight: 36  value: 30
Item 9 - weight: 24  value: 35
Packed bin weight: 96
Packed bin value: 100

Total packed weight: 438

完成计划

下面显示了适用于多件背包的完整计划。

Python

"""Solve a multiple knapsack problem using a MIP solver."""
from ortools.linear_solver import pywraplp


def main():
    data = {}
    data["weights"] = [48, 30, 42, 36, 36, 48, 42, 42, 36, 24, 30, 30, 42, 36, 36]
    data["values"] = [10, 30, 25, 50, 35, 30, 15, 40, 30, 35, 45, 10, 20, 30, 25]
    assert len(data["weights"]) == len(data["values"])
    data["num_items"] = len(data["weights"])
    data["all_items"] = range(data["num_items"])

    data["bin_capacities"] = [100, 100, 100, 100, 100]
    data["num_bins"] = len(data["bin_capacities"])
    data["all_bins"] = range(data["num_bins"])

    # Create the mip solver with the SCIP backend.
    solver = pywraplp.Solver.CreateSolver("SCIP")
    if solver is None:
        print("SCIP solver unavailable.")
        return

    # Variables.
    # x[i, b] = 1 if item i is packed in bin b.
    x = {}
    for i in data["all_items"]:
        for b in data["all_bins"]:
            x[i, b] = solver.BoolVar(f"x_{i}_{b}")

    # Constraints.
    # Each item is assigned to at most one bin.
    for i in data["all_items"]:
        solver.Add(sum(x[i, b] for b in data["all_bins"]) <= 1)

    # The amount packed in each bin cannot exceed its capacity.
    for b in data["all_bins"]:
        solver.Add(
            sum(x[i, b] * data["weights"][i] for i in data["all_items"])
            <= data["bin_capacities"][b]
        )

    # Objective.
    # Maximize total value of packed items.
    objective = solver.Objective()
    for i in data["all_items"]:
        for b in data["all_bins"]:
            objective.SetCoefficient(x[i, b], data["values"][i])
    objective.SetMaximization()

    print(f"Solving with {solver.SolverVersion()}")
    status = solver.Solve()

    if status == pywraplp.Solver.OPTIMAL:
        print(f"Total packed value: {objective.Value()}")
        total_weight = 0
        for b in data["all_bins"]:
            print(f"Bin {b}")
            bin_weight = 0
            bin_value = 0
            for i in data["all_items"]:
                if x[i, b].solution_value() > 0:
                    print(
                        f"Item {i} weight: {data['weights'][i]} value:"
                        f" {data['values'][i]}"
                    )
                    bin_weight += data["weights"][i]
                    bin_value += data["values"][i]
            print(f"Packed bin weight: {bin_weight}")
            print(f"Packed bin value: {bin_value}\n")
            total_weight += bin_weight
        print(f"Total packed weight: {total_weight}")
    else:
        print("The problem does not have an optimal solution.")


if __name__ == "__main__":
    main()

C++

// Solve a multiple knapsack problem using a MIP solver.
#include <iostream>
#include <memory>
#include <numeric>
#include <vector>

#include "absl/strings/str_format.h"
#include "ortools/linear_solver/linear_expr.h"
#include "ortools/linear_solver/linear_solver.h"

namespace operations_research {

void MultipleKnapsackMip() {
  const std::vector<int> weights = {
      {48, 30, 42, 36, 36, 48, 42, 42, 36, 24, 30, 30, 42, 36, 36}};
  const std::vector<int> values = {
      {10, 30, 25, 50, 35, 30, 15, 40, 30, 35, 45, 10, 20, 30, 25}};
  const int num_items = weights.size();
  std::vector<int> all_items(num_items);
  std::iota(all_items.begin(), all_items.end(), 0);

  const std::vector<int> bin_capacities = {{100, 100, 100, 100, 100}};
  const int num_bins = bin_capacities.size();
  std::vector<int> all_bins(num_bins);
  std::iota(all_bins.begin(), all_bins.end(), 0);

  // Create the mip solver with the SCIP backend.
  std::unique_ptr<MPSolver> solver(MPSolver::CreateSolver("SCIP"));
  if (!solver) {
    LOG(WARNING) << "SCIP solver unavailable.";
    return;
  }

  // Variables.
  // x[i][b] = 1 if item i is packed in bin b.
  std::vector<std::vector<const MPVariable*>> x(
      num_items, std::vector<const MPVariable*>(num_bins));
  for (int i : all_items) {
    for (int b : all_bins) {
      x[i][b] = solver->MakeBoolVar(absl::StrFormat("x_%d_%d", i, b));
    }
  }

  // Constraints.
  // Each item is assigned to at most one bin.
  for (int i : all_items) {
    LinearExpr sum;
    for (int b : all_bins) {
      sum += x[i][b];
    }
    solver->MakeRowConstraint(sum <= 1.0);
  }
  // The amount packed in each bin cannot exceed its capacity.
  for (int b : all_bins) {
    LinearExpr bin_weight;
    for (int i : all_items) {
      bin_weight += LinearExpr(x[i][b]) * weights[i];
    }
    solver->MakeRowConstraint(bin_weight <= bin_capacities[b]);
  }

  // Objective.
  // Maximize total value of packed items.
  MPObjective* const objective = solver->MutableObjective();
  LinearExpr objective_value;
  for (int i : all_items) {
    for (int b : all_bins) {
      objective_value += LinearExpr(x[i][b]) * values[i];
    }
  }
  objective->MaximizeLinearExpr(objective_value);

  const MPSolver::ResultStatus result_status = solver->Solve();

  if (result_status == MPSolver::OPTIMAL) {
    LOG(INFO) << "Total packed value: " << objective->Value();
    double total_weight = 0.0;
    for (int b : all_bins) {
      LOG(INFO) << "Bin " << b;
      double bin_weight = 0.0;
      double bin_value = 0.0;
      for (int i : all_items) {
        if (x[i][b]->solution_value() > 0) {
          LOG(INFO) << "Item " << i << " weight: " << weights[i]
                    << " value: " << values[i];
          bin_weight += weights[i];
          bin_value += values[i];
        }
      }
      LOG(INFO) << "Packed bin weight: " << bin_weight;
      LOG(INFO) << "Packed bin value: " << bin_value;
      total_weight += bin_weight;
    }
    LOG(INFO) << "Total packed weight: " << total_weight;
  } else {
    LOG(INFO) << "The problem does not have an optimal solution.";
  }
}
}  // namespace operations_research

int main(int argc, char** argv) {
  operations_research::MultipleKnapsackMip();
  return EXIT_SUCCESS;
}

Java

// Solve a multiple knapsack problem using a MIP solver.
package com.google.ortools.linearsolver.samples;
import com.google.ortools.Loader;
import com.google.ortools.linearsolver.MPConstraint;
import com.google.ortools.linearsolver.MPObjective;
import com.google.ortools.linearsolver.MPSolver;
import com.google.ortools.linearsolver.MPVariable;
import java.util.stream.IntStream;

/** Multiple knapsack problem. */
public class MultipleKnapsackMip {
  public static void main(String[] args) {
    Loader.loadNativeLibraries();
    // Instantiate the data problem.
    final double[] weights = {48, 30, 42, 36, 36, 48, 42, 42, 36, 24, 30, 30, 42, 36, 36};
    final double[] values = {10, 30, 25, 50, 35, 30, 15, 40, 30, 35, 45, 10, 20, 30, 25};
    final int numItems = weights.length;
    final int[] allItems = IntStream.range(0, numItems).toArray();

    final double[] binCapacities = {100, 100, 100, 100, 100};
    final int numBins = binCapacities.length;
    final int[] allBins = IntStream.range(0, numBins).toArray();

    // Create the linear solver with the SCIP backend.
    MPSolver solver = MPSolver.createSolver("SCIP");
    if (solver == null) {
      System.out.println("Could not create solver SCIP");
      return;
    }

    // Variables.
    MPVariable[][] x = new MPVariable[numItems][numBins];
    for (int i : allItems) {
      for (int b : allBins) {
        x[i][b] = solver.makeBoolVar("x_" + i + "_" + b);
      }
    }

    // Constraints.
    // Each item is assigned to at most one bin.
    for (int i : allItems) {
      MPConstraint constraint = solver.makeConstraint(0, 1, "");
      for (int b : allBins) {
        constraint.setCoefficient(x[i][b], 1);
      }
    }

    // The amount packed in each bin cannot exceed its capacity.
    for (int b : allBins) {
      MPConstraint constraint = solver.makeConstraint(0, binCapacities[b], "");
      for (int i : allItems) {
        constraint.setCoefficient(x[i][b], weights[i]);
      }
    }

    // Objective.
    // Maximize total value of packed items.
    MPObjective objective = solver.objective();
    for (int i : allItems) {
      for (int b : allBins) {
        objective.setCoefficient(x[i][b], values[i]);
      }
    }
    objective.setMaximization();

    final MPSolver.ResultStatus status = solver.solve();

    // Check that the problem has an optimal solution.
    if (status == MPSolver.ResultStatus.OPTIMAL) {
      System.out.println("Total packed value: " + objective.value());
      double totalWeight = 0;
      for (int b : allBins) {
        double binWeight = 0;
        double binValue = 0;
        System.out.println("Bin " + b);
        for (int i : allItems) {
          if (x[i][b].solutionValue() == 1) {
            System.out.println("Item " + i + " weight: " + weights[i] + " value: " + values[i]);
            binWeight += weights[i];
            binValue += values[i];
          }
        }
        System.out.println("Packed bin weight: " + binWeight);
        System.out.println("Packed bin value: " + binValue);
        totalWeight += binWeight;
      }
      System.out.println("Total packed weight: " + totalWeight);
    } else {
      System.err.println("The problem does not have an optimal solution.");
    }
  }

  private MultipleKnapsackMip() {}
}

C#

// Solve a multiple knapsack problem using a MIP solver.
using System;
using System.Collections.Generic;
using System.Linq;
using Google.OrTools.LinearSolver;

public class MultipleKnapsackMip
{
    public static void Main()
    {
        // Instantiate the data problem.
        double[] Weights = { 48, 30, 42, 36, 36, 48, 42, 42, 36, 24, 30, 30, 42, 36, 36 };
        double[] Values = { 10, 30, 25, 50, 35, 30, 15, 40, 30, 35, 45, 10, 20, 30, 25 };
        int NumItems = Weights.Length;
        int[] allItems = Enumerable.Range(0, NumItems).ToArray();

        double[] BinCapacities = { 100, 100, 100, 100, 100 };
        int NumBins = BinCapacities.Length;
        int[] allBins = Enumerable.Range(0, NumBins).ToArray();

        // Create the linear solver with the SCIP backend.
        Solver solver = Solver.CreateSolver("SCIP");
        if (solver is null)
        {
            return;
        }

        // Variables.
        Variable[,] x = new Variable[NumItems, NumBins];
        foreach (int i in allItems)
        {
            foreach (int b in allBins)
            {
                x[i, b] = solver.MakeBoolVar($"x_{i}_{b}");
            }
        }

        // Constraints.
        // Each item is assigned to at most one bin.
        foreach (int i in allItems)
        {
            Constraint constraint = solver.MakeConstraint(0, 1, "");
            foreach (int b in allBins)
            {
                constraint.SetCoefficient(x[i, b], 1);
            }
        }

        // The amount packed in each bin cannot exceed its capacity.
        foreach (int b in allBins)
        {
            Constraint constraint = solver.MakeConstraint(0, BinCapacities[b], "");
            foreach (int i in allItems)
            {
                constraint.SetCoefficient(x[i, b], Weights[i]);
            }
        }

        // Objective.
        Objective objective = solver.Objective();
        foreach (int i in allItems)
        {
            foreach (int b in allBins)
            {
                objective.SetCoefficient(x[i, b], Values[i]);
            }
        }
        objective.SetMaximization();

        Solver.ResultStatus resultStatus = solver.Solve();

        // Check that the problem has an optimal solution.
        if (resultStatus == Solver.ResultStatus.OPTIMAL)
        {
            Console.WriteLine($"Total packed value: {solver.Objective().Value()}");
            double TotalWeight = 0.0;
            foreach (int b in allBins)
            {
                double BinWeight = 0.0;
                double BinValue = 0.0;
                Console.WriteLine("Bin " + b);
                foreach (int i in allItems)
                {
                    if (x[i, b].SolutionValue() == 1)
                    {
                        Console.WriteLine($"Item {i} weight: {Weights[i]} values: {Values[i]}");
                        BinWeight += Weights[i];
                        BinValue += Values[i];
                    }
                }
                Console.WriteLine("Packed bin weight: " + BinWeight);
                Console.WriteLine("Packed bin value: " + BinValue);
                TotalWeight += BinWeight;
            }
            Console.WriteLine("Total packed weight: " + TotalWeight);
        }
        else
        {
            Console.WriteLine("The problem does not have an optimal solution!");
        }
    }
}

CP SAT 解决方案

下文介绍如何使用 CP-SAT 求解器求解题目。

导入库

以下代码会导入所需的库。

Python

from ortools.sat.python import cp_model

C++

#include <stdlib.h>

#include <map>
#include <numeric>
#include <tuple>
#include <vector>

#include "absl/strings/str_format.h"
#include "ortools/base/logging.h"
#include "ortools/sat/cp_model.h"
#include "ortools/sat/cp_model.pb.h"
#include "ortools/sat/cp_model_solver.h"

Java

import com.google.ortools.Loader;
import com.google.ortools.sat.CpModel;
import com.google.ortools.sat.CpSolver;
import com.google.ortools.sat.CpSolverStatus;
import com.google.ortools.sat.LinearExpr;
import com.google.ortools.sat.LinearExprBuilder;
import com.google.ortools.sat.Literal;
import java.util.ArrayList;
import java.util.List;
import java.util.stream.IntStream;

C#

using System;
using System.Collections.Generic;
using System.Linq;
using Google.OrTools.Sat;

public class MultipleKnapsackSat
{
    public static void Main(String[] args)
    {
        // Instantiate the data problem.
        int[] Weights = { 48, 30, 42, 36, 36, 48, 42, 42, 36, 24, 30, 30, 42, 36, 36 };
        int[] Values = { 10, 30, 25, 50, 35, 30, 15, 40, 30, 35, 45, 10, 20, 30, 25 };
        int NumItems = Weights.Length;
        int[] allItems = Enumerable.Range(0, NumItems).ToArray();

        int[] BinCapacities = { 100, 100, 100, 100, 100 };
        int NumBins = BinCapacities.Length;
        int[] allBins = Enumerable.Range(0, NumBins).ToArray();

        // Model.
        CpModel model = new CpModel();

        // Variables.
        ILiteral[,] x = new ILiteral[NumItems, NumBins];
        foreach (int i in allItems)
        {
            foreach (int b in allBins)
            {
                x[i, b] = model.NewBoolVar($"x_{i}_{b}");
            }
        }

        // Constraints.
        // Each item is assigned to at most one bin.
        foreach (int i in allItems)
        {
            List<ILiteral> literals = new List<ILiteral>();
            foreach (int b in allBins)
            {
                literals.Add(x[i, b]);
            }
            model.AddAtMostOne(literals);
        }

        // The amount packed in each bin cannot exceed its capacity.
        foreach (int b in allBins)
        {
            List<ILiteral> items = new List<ILiteral>();
            foreach (int i in allItems)
            {
                items.Add(x[i, b]);
            }
            model.Add(LinearExpr.WeightedSum(items, Weights) <= BinCapacities[b]);
        }

        // Objective.
        LinearExprBuilder obj = LinearExpr.NewBuilder();
        foreach (int i in allItems)
        {
            foreach (int b in allBins)
            {
                obj.AddTerm(x[i, b], Values[i]);
            }
        }
        model.Maximize(obj);

        // Solve
        CpSolver solver = new CpSolver();
        CpSolverStatus status = solver.Solve(model);

        // Print solution.
        // Check that the problem has a feasible solution.
        if (status == CpSolverStatus.Optimal || status == CpSolverStatus.Feasible)
        {
            Console.WriteLine($"Total packed value: {solver.ObjectiveValue}");
            double TotalWeight = 0.0;
            foreach (int b in allBins)
            {
                double BinWeight = 0.0;
                double BinValue = 0.0;
                Console.WriteLine($"Bin {b}");
                foreach (int i in allItems)
                {
                    if (solver.BooleanValue(x[i, b]))
                    {
                        Console.WriteLine($"Item {i} weight: {Weights[i]} values: {Values[i]}");
                        BinWeight += Weights[i];
                        BinValue += Values[i];
                    }
                }
                Console.WriteLine("Packed bin weight: " + BinWeight);
                Console.WriteLine("Packed bin value: " + BinValue);
                TotalWeight += BinWeight;
            }
            Console.WriteLine("Total packed weight: " + TotalWeight);
        }
        else
        {
            Console.WriteLine("No solution found.");
        }

        Console.WriteLine("Statistics");
        Console.WriteLine($"  conflicts: {solver.NumConflicts()}");
        Console.WriteLine($"  branches : {solver.NumBranches()}");
        Console.WriteLine($"  wall time: {solver.WallTime()}s");
    }
}

声明模型

以下代码声明了 CP-SAT 模型。

Python

model = cp_model.CpModel()

C++

CpModelBuilder cp_model;

Java

CpModel model = new CpModel();

C#

CpModel model = new CpModel();

创建数据

以下代码用于设置问题的数据。

Python

data = {}
data["weights"] = [48, 30, 42, 36, 36, 48, 42, 42, 36, 24, 30, 30, 42, 36, 36]
data["values"] = [10, 30, 25, 50, 35, 30, 15, 40, 30, 35, 45, 10, 20, 30, 25]
assert len(data["weights"]) == len(data["values"])
num_items = len(data["weights"])
all_items = range(num_items)

data["bin_capacities"] = [100, 100, 100, 100, 100]
num_bins = len(data["bin_capacities"])
all_bins = range(num_bins)

C++

const std::vector<int> weights = {
    {48, 30, 42, 36, 36, 48, 42, 42, 36, 24, 30, 30, 42, 36, 36}};
const std::vector<int> values = {
    {10, 30, 25, 50, 35, 30, 15, 40, 30, 35, 45, 10, 20, 30, 25}};
const int num_items = static_cast<int>(weights.size());
std::vector<int> all_items(num_items);
std::iota(all_items.begin(), all_items.end(), 0);

const std::vector<int> bin_capacities = {{100, 100, 100, 100, 100}};
const int num_bins = static_cast<int>(bin_capacities.size());
std::vector<int> all_bins(num_bins);
std::iota(all_bins.begin(), all_bins.end(), 0);

Java

final int[] weights = {48, 30, 42, 36, 36, 48, 42, 42, 36, 24, 30, 30, 42, 36, 36};
final int[] values = {10, 30, 25, 50, 35, 30, 15, 40, 30, 35, 45, 10, 20, 30, 25};
final int numItems = weights.length;
final int[] allItems = IntStream.range(0, numItems).toArray();

final int[] binCapacities = {100, 100, 100, 100, 100};
final int numBins = binCapacities.length;
final int[] allBins = IntStream.range(0, numBins).toArray();

C#

int[] Weights = { 48, 30, 42, 36, 36, 48, 42, 42, 36, 24, 30, 30, 42, 36, 36 };
int[] Values = { 10, 30, 25, 50, 35, 30, 15, 40, 30, 35, 45, 10, 20, 30, 25 };
int NumItems = Weights.Length;
int[] allItems = Enumerable.Range(0, NumItems).ToArray();

int[] BinCapacities = { 100, 100, 100, 100, 100 };
int NumBins = BinCapacities.Length;
int[] allBins = Enumerable.Range(0, NumBins).ToArray();

costs 数组与费用表相对应用于向任务分配工作器，如上所示。

创建变量

以下代码会为问题创建二进制整数变量。

Python

# x[i, b] = 1 if item i is packed in bin b.
x = {}
for i in all_items:
    for b in all_bins:
        x[i, b] = model.new_bool_var(f"x_{i}_{b}")

C++

// x[i, b] = 1 if item i is packed in bin b.
std::map<std::tuple<int, int>, BoolVar> x;
for (int i : all_items) {
  for (int b : all_bins) {
    auto key = std::make_tuple(i, b);
    x[key] = cp_model.NewBoolVar().WithName(absl::StrFormat("x_%d_%d", i, b));
  }
}

Java

Literal[][] x = new Literal[numItems][numBins];
for (int i : allItems) {
  for (int b : allBins) {
    x[i][b] = model.newBoolVar("x_" + i + "_" + b);
  }
}

C#

ILiteral[,] x = new ILiteral[NumItems, NumBins];
foreach (int i in allItems)
{
    foreach (int b in allBins)
    {
        x[i, b] = model.NewBoolVar($"x_{i}_{b}");
    }
}

创建限制条件

以下代码为问题创建了约束条件。

Python

# Each item is assigned to at most one bin.
for i in all_items:
    model.add_at_most_one(x[i, b] for b in all_bins)

# The amount packed in each bin cannot exceed its capacity.
for b in all_bins:
    model.add(
        sum(x[i, b] * data["weights"][i] for i in all_items)
        <= data["bin_capacities"][b]
    )

C++

// Each item is assigned to at most one bin.
for (int i : all_items) {
  std::vector<BoolVar> copies;
  for (int b : all_bins) {
    copies.push_back(x[std::make_tuple(i, b)]);
  }
  cp_model.AddAtMostOne(copies);
}

// The amount packed in each bin cannot exceed its capacity.
for (int b : all_bins) {
  LinearExpr bin_weight;
  for (int i : all_items) {
    bin_weight += x[std::make_tuple(i, b)] * weights[i];
  }
  cp_model.AddLessOrEqual(bin_weight, bin_capacities[b]);
}

Java

// Each item is assigned to at most one bin.
for (int i : allItems) {
  List<Literal> bins = new ArrayList<>();
  for (int b : allBins) {
    bins.add(x[i][b]);
  }
  model.addAtMostOne(bins);
}

// The amount packed in each bin cannot exceed its capacity.
for (int b : allBins) {
  LinearExprBuilder load = LinearExpr.newBuilder();
  for (int i : allItems) {
    load.addTerm(x[i][b], weights[i]);
  }
  model.addLessOrEqual(load, binCapacities[b]);
}

C#

// Each item is assigned to at most one bin.
foreach (int i in allItems)
{
    List<ILiteral> literals = new List<ILiteral>();
    foreach (int b in allBins)
    {
        literals.Add(x[i, b]);
    }
    model.AddAtMostOne(literals);
}

// The amount packed in each bin cannot exceed its capacity.
foreach (int b in allBins)
{
    List<ILiteral> items = new List<ILiteral>();
    foreach (int i in allItems)
    {
        items.Add(x[i, b]);
    }
    model.Add(LinearExpr.WeightedSum(items, Weights) <= BinCapacities[b]);
}

创建目标函数

以下代码为问题创建目标函数。

Python

# maximize total value of packed items.
objective = []
for i in all_items:
    for b in all_bins:
        objective.append(cp_model.LinearExpr.term(x[i, b], data["values"][i]))
model.maximize(cp_model.LinearExpr.sum(objective))

C++

// Maximize total value of packed items.
LinearExpr objective;
for (int i : all_items) {
  for (int b : all_bins) {
    objective += x[std::make_tuple(i, b)] * values[i];
  }
}
cp_model.Maximize(objective);

Java

// Maximize total value of packed items.
LinearExprBuilder obj = LinearExpr.newBuilder();
for (int i : allItems) {
  for (int b : allBins) {
    obj.addTerm(x[i][b], values[i]);
  }
}
model.maximize(obj);

C#

LinearExprBuilder obj = LinearExpr.NewBuilder();
foreach (int i in allItems)
{
    foreach (int b in allBins)
    {
        obj.AddTerm(x[i, b], Values[i]);
    }
}
model.Maximize(obj);

目标函数的值是分配给所有变量的求解器求出的值为 1。

调用求解器

以下代码会调用求解器。

Python

solver = cp_model.CpSolver()
status = solver.solve(model)

C++

const CpSolverResponse response = Solve(cp_model.Build());

Java

CpSolver solver = new CpSolver();
final CpSolverStatus status = solver.solve(model);

C#

CpSolver solver = new CpSolver();
CpSolverStatus status = solver.Solve(model);

输出解决方案

以下代码会输出该问题的解决方案。

Python

if status == cp_model.OPTIMAL:
    print(f"Total packed value: {solver.objective_value}")
    total_weight = 0
    for b in all_bins:
        print(f"Bin {b}")
        bin_weight = 0
        bin_value = 0
        for i in all_items:
            if solver.value(x[i, b]) > 0:
                print(
                    f'Item:{i} weight:{data["weights"][i]} value:{data["values"][i]}'
                )
                bin_weight += data["weights"][i]
                bin_value += data["values"][i]
        print(f"Packed bin weight: {bin_weight}")
        print(f"Packed bin value: {bin_value}\n")
        total_weight += bin_weight
    print(f"Total packed weight: {total_weight}")
else:
    print("The problem does not have an optimal solution.")

C++

if (response.status() == CpSolverStatus::OPTIMAL ||
    response.status() == CpSolverStatus::FEASIBLE) {
  LOG(INFO) << "Total packed value: " << response.objective_value();
  double total_weight = 0.0;
  for (int b : all_bins) {
    LOG(INFO) << "Bin " << b;
    double bin_weight = 0.0;
    double bin_value = 0.0;
    for (int i : all_items) {
      auto key = std::make_tuple(i, b);
      if (SolutionIntegerValue(response, x[key]) > 0) {
        LOG(INFO) << "Item " << i << " weight: " << weights[i]
                  << " value: " << values[i];
        bin_weight += weights[i];
        bin_value += values[i];
      }
    }
    LOG(INFO) << "Packed bin weight: " << bin_weight;
    LOG(INFO) << "Packed bin value: " << bin_value;
    total_weight += bin_weight;
  }
  LOG(INFO) << "Total packed weight: " << total_weight;
} else {
  LOG(INFO) << "The problem does not have an optimal solution.";
}

Java

// Check that the problem has an optimal solution.
if (status == CpSolverStatus.OPTIMAL) {
  System.out.println("Total packed value: " + solver.objectiveValue());
  long totalWeight = 0;
  for (int b : allBins) {
    long binWeight = 0;
    long binValue = 0;
    System.out.println("Bin " + b);
    for (int i : allItems) {
      if (solver.booleanValue(x[i][b])) {
        System.out.println("Item " + i + " weight: " + weights[i] + " value: " + values[i]);
        binWeight += weights[i];
        binValue += values[i];
      }
    }
    System.out.println("Packed bin weight: " + binWeight);
    System.out.println("Packed bin value: " + binValue);
    totalWeight += binWeight;
  }
  System.out.println("Total packed weight: " + totalWeight);
} else {
  System.err.println("The problem does not have an optimal solution.");
}

C#

// Check that the problem has a feasible solution.
if (status == CpSolverStatus.Optimal || status == CpSolverStatus.Feasible)
{
    Console.WriteLine($"Total packed value: {solver.ObjectiveValue}");
    double TotalWeight = 0.0;
    foreach (int b in allBins)
    {
        double BinWeight = 0.0;
        double BinValue = 0.0;
        Console.WriteLine($"Bin {b}");
        foreach (int i in allItems)
        {
            if (solver.BooleanValue(x[i, b]))
            {
                Console.WriteLine($"Item {i} weight: {Weights[i]} values: {Values[i]}");
                BinWeight += Weights[i];
                BinValue += Values[i];
            }
        }
        Console.WriteLine("Packed bin weight: " + BinWeight);
        Console.WriteLine("Packed bin value: " + BinValue);
        TotalWeight += BinWeight;
    }
    Console.WriteLine("Total packed weight: " + TotalWeight);
}
else
{
    Console.WriteLine("No solution found.");
}

以下是程序的输出。

Total packed value: 395.0
Bin  0

Item 3 - weight: 36  value: 50
Item 13 - weight: 36  value: 30
Packed bin weight: 72
Packed bin value: 80

Bin  1

Item 5 - weight: 48  value: 30
Item 7 - weight: 42  value: 40
Packed bin weight: 90
Packed bin value: 70

Bin  2

Item 1 - weight: 30  value: 30
Item 10 - weight: 30  value: 45
Item 14 - weight: 36  value: 25
Packed bin weight: 96
Packed bin value: 100

Bin  3

Item 2 - weight: 42  value: 25
Item 12 - weight: 42  value: 20
Packed bin weight: 84
Packed bin value: 45

Bin  4

Item 4 - weight: 36  value: 35
Item 8 - weight: 36  value: 30
Item 9 - weight: 24  value: 35
Packed bin weight: 96
Packed bin value: 100

Total packed weight: 438

完成计划

以下是 CP-SAT 解决方案的完整计划。

Python

"""Solves a multiple knapsack problem using the CP-SAT solver."""
from ortools.sat.python import cp_model


def main() -> None:
    data = {}
    data["weights"] = [48, 30, 42, 36, 36, 48, 42, 42, 36, 24, 30, 30, 42, 36, 36]
    data["values"] = [10, 30, 25, 50, 35, 30, 15, 40, 30, 35, 45, 10, 20, 30, 25]
    assert len(data["weights"]) == len(data["values"])
    num_items = len(data["weights"])
    all_items = range(num_items)

    data["bin_capacities"] = [100, 100, 100, 100, 100]
    num_bins = len(data["bin_capacities"])
    all_bins = range(num_bins)

    model = cp_model.CpModel()

    # Variables.
    # x[i, b] = 1 if item i is packed in bin b.
    x = {}
    for i in all_items:
        for b in all_bins:
            x[i, b] = model.new_bool_var(f"x_{i}_{b}")

    # Constraints.
    # Each item is assigned to at most one bin.
    for i in all_items:
        model.add_at_most_one(x[i, b] for b in all_bins)

    # The amount packed in each bin cannot exceed its capacity.
    for b in all_bins:
        model.add(
            sum(x[i, b] * data["weights"][i] for i in all_items)
            <= data["bin_capacities"][b]
        )

    # Objective.
    # maximize total value of packed items.
    objective = []
    for i in all_items:
        for b in all_bins:
            objective.append(cp_model.LinearExpr.term(x[i, b], data["values"][i]))
    model.maximize(cp_model.LinearExpr.sum(objective))

    solver = cp_model.CpSolver()
    status = solver.solve(model)

    if status == cp_model.OPTIMAL:
        print(f"Total packed value: {solver.objective_value}")
        total_weight = 0
        for b in all_bins:
            print(f"Bin {b}")
            bin_weight = 0
            bin_value = 0
            for i in all_items:
                if solver.value(x[i, b]) > 0:
                    print(
                        f'Item:{i} weight:{data["weights"][i]} value:{data["values"][i]}'
                    )
                    bin_weight += data["weights"][i]
                    bin_value += data["values"][i]
            print(f"Packed bin weight: {bin_weight}")
            print(f"Packed bin value: {bin_value}\n")
            total_weight += bin_weight
        print(f"Total packed weight: {total_weight}")
    else:
        print("The problem does not have an optimal solution.")


if __name__ == "__main__":
    main()

C++

// Solves a multiple knapsack problem using the CP-SAT solver.
#include <stdlib.h>

#include <map>
#include <numeric>
#include <tuple>
#include <vector>

#include "absl/strings/str_format.h"
#include "ortools/base/logging.h"
#include "ortools/sat/cp_model.h"
#include "ortools/sat/cp_model.pb.h"
#include "ortools/sat/cp_model_solver.h"

namespace operations_research {
namespace sat {

void MultipleKnapsackSat() {
  const std::vector<int> weights = {
      {48, 30, 42, 36, 36, 48, 42, 42, 36, 24, 30, 30, 42, 36, 36}};
  const std::vector<int> values = {
      {10, 30, 25, 50, 35, 30, 15, 40, 30, 35, 45, 10, 20, 30, 25}};
  const int num_items = static_cast<int>(weights.size());
  std::vector<int> all_items(num_items);
  std::iota(all_items.begin(), all_items.end(), 0);

  const std::vector<int> bin_capacities = {{100, 100, 100, 100, 100}};
  const int num_bins = static_cast<int>(bin_capacities.size());
  std::vector<int> all_bins(num_bins);
  std::iota(all_bins.begin(), all_bins.end(), 0);

  CpModelBuilder cp_model;

  // Variables.
  // x[i, b] = 1 if item i is packed in bin b.
  std::map<std::tuple<int, int>, BoolVar> x;
  for (int i : all_items) {
    for (int b : all_bins) {
      auto key = std::make_tuple(i, b);
      x[key] = cp_model.NewBoolVar().WithName(absl::StrFormat("x_%d_%d", i, b));
    }
  }

  // Constraints.
  // Each item is assigned to at most one bin.
  for (int i : all_items) {
    std::vector<BoolVar> copies;
    for (int b : all_bins) {
      copies.push_back(x[std::make_tuple(i, b)]);
    }
    cp_model.AddAtMostOne(copies);
  }

  // The amount packed in each bin cannot exceed its capacity.
  for (int b : all_bins) {
    LinearExpr bin_weight;
    for (int i : all_items) {
      bin_weight += x[std::make_tuple(i, b)] * weights[i];
    }
    cp_model.AddLessOrEqual(bin_weight, bin_capacities[b]);
  }

  // Objective.
  // Maximize total value of packed items.
  LinearExpr objective;
  for (int i : all_items) {
    for (int b : all_bins) {
      objective += x[std::make_tuple(i, b)] * values[i];
    }
  }
  cp_model.Maximize(objective);

  const CpSolverResponse response = Solve(cp_model.Build());

  if (response.status() == CpSolverStatus::OPTIMAL ||
      response.status() == CpSolverStatus::FEASIBLE) {
    LOG(INFO) << "Total packed value: " << response.objective_value();
    double total_weight = 0.0;
    for (int b : all_bins) {
      LOG(INFO) << "Bin " << b;
      double bin_weight = 0.0;
      double bin_value = 0.0;
      for (int i : all_items) {
        auto key = std::make_tuple(i, b);
        if (SolutionIntegerValue(response, x[key]) > 0) {
          LOG(INFO) << "Item " << i << " weight: " << weights[i]
                    << " value: " << values[i];
          bin_weight += weights[i];
          bin_value += values[i];
        }
      }
      LOG(INFO) << "Packed bin weight: " << bin_weight;
      LOG(INFO) << "Packed bin value: " << bin_value;
      total_weight += bin_weight;
    }
    LOG(INFO) << "Total packed weight: " << total_weight;
  } else {
    LOG(INFO) << "The problem does not have an optimal solution.";
  }

  // Statistics.
  LOG(INFO) << "Statistics";
  LOG(INFO) << CpSolverResponseStats(response);
}
}  // namespace sat
}  // namespace operations_research

int main() {
  operations_research::sat::MultipleKnapsackSat();
  return EXIT_SUCCESS;
}

Java

// Solves a multiple knapsack problem using the CP-SAT solver.
package com.google.ortools.sat.samples;
import com.google.ortools.Loader;
import com.google.ortools.sat.CpModel;
import com.google.ortools.sat.CpSolver;
import com.google.ortools.sat.CpSolverStatus;
import com.google.ortools.sat.LinearExpr;
import com.google.ortools.sat.LinearExprBuilder;
import com.google.ortools.sat.Literal;
import java.util.ArrayList;
import java.util.List;
import java.util.stream.IntStream;

/** Sample showing how to solve a multiple knapsack problem. */
public class MultipleKnapsackSat {
  public static void main(String[] args) {
    Loader.loadNativeLibraries();
    // Instantiate the data problem.
    final int[] weights = {48, 30, 42, 36, 36, 48, 42, 42, 36, 24, 30, 30, 42, 36, 36};
    final int[] values = {10, 30, 25, 50, 35, 30, 15, 40, 30, 35, 45, 10, 20, 30, 25};
    final int numItems = weights.length;
    final int[] allItems = IntStream.range(0, numItems).toArray();

    final int[] binCapacities = {100, 100, 100, 100, 100};
    final int numBins = binCapacities.length;
    final int[] allBins = IntStream.range(0, numBins).toArray();

    CpModel model = new CpModel();

    // Variables.
    Literal[][] x = new Literal[numItems][numBins];
    for (int i : allItems) {
      for (int b : allBins) {
        x[i][b] = model.newBoolVar("x_" + i + "_" + b);
      }
    }

    // Constraints.
    // Each item is assigned to at most one bin.
    for (int i : allItems) {
      List<Literal> bins = new ArrayList<>();
      for (int b : allBins) {
        bins.add(x[i][b]);
      }
      model.addAtMostOne(bins);
    }

    // The amount packed in each bin cannot exceed its capacity.
    for (int b : allBins) {
      LinearExprBuilder load = LinearExpr.newBuilder();
      for (int i : allItems) {
        load.addTerm(x[i][b], weights[i]);
      }
      model.addLessOrEqual(load, binCapacities[b]);
    }

    // Objective.
    // Maximize total value of packed items.
    LinearExprBuilder obj = LinearExpr.newBuilder();
    for (int i : allItems) {
      for (int b : allBins) {
        obj.addTerm(x[i][b], values[i]);
      }
    }
    model.maximize(obj);

    CpSolver solver = new CpSolver();
    final CpSolverStatus status = solver.solve(model);

    // Check that the problem has an optimal solution.
    if (status == CpSolverStatus.OPTIMAL) {
      System.out.println("Total packed value: " + solver.objectiveValue());
      long totalWeight = 0;
      for (int b : allBins) {
        long binWeight = 0;
        long binValue = 0;
        System.out.println("Bin " + b);
        for (int i : allItems) {
          if (solver.booleanValue(x[i][b])) {
            System.out.println("Item " + i + " weight: " + weights[i] + " value: " + values[i]);
            binWeight += weights[i];
            binValue += values[i];
          }
        }
        System.out.println("Packed bin weight: " + binWeight);
        System.out.println("Packed bin value: " + binValue);
        totalWeight += binWeight;
      }
      System.out.println("Total packed weight: " + totalWeight);
    } else {
      System.err.println("The problem does not have an optimal solution.");
    }
  }

  private MultipleKnapsackSat() {}
}

C#

// Solves a multiple knapsack problem using the CP-SAT solver.
using System;
using System.Collections.Generic;
using System.Linq;
using Google.OrTools.Sat;

public class MultipleKnapsackSat
{
    public static void Main(String[] args)
    {
        // Instantiate the data problem.
        int[] Weights = { 48, 30, 42, 36, 36, 48, 42, 42, 36, 24, 30, 30, 42, 36, 36 };
        int[] Values = { 10, 30, 25, 50, 35, 30, 15, 40, 30, 35, 45, 10, 20, 30, 25 };
        int NumItems = Weights.Length;
        int[] allItems = Enumerable.Range(0, NumItems).ToArray();

        int[] BinCapacities = { 100, 100, 100, 100, 100 };
        int NumBins = BinCapacities.Length;
        int[] allBins = Enumerable.Range(0, NumBins).ToArray();

        // Model.
        CpModel model = new CpModel();

        // Variables.
        ILiteral[,] x = new ILiteral[NumItems, NumBins];
        foreach (int i in allItems)
        {
            foreach (int b in allBins)
            {
                x[i, b] = model.NewBoolVar($"x_{i}_{b}");
            }
        }

        // Constraints.
        // Each item is assigned to at most one bin.
        foreach (int i in allItems)
        {
            List<ILiteral> literals = new List<ILiteral>();
            foreach (int b in allBins)
            {
                literals.Add(x[i, b]);
            }
            model.AddAtMostOne(literals);
        }

        // The amount packed in each bin cannot exceed its capacity.
        foreach (int b in allBins)
        {
            List<ILiteral> items = new List<ILiteral>();
            foreach (int i in allItems)
            {
                items.Add(x[i, b]);
            }
            model.Add(LinearExpr.WeightedSum(items, Weights) <= BinCapacities[b]);
        }

        // Objective.
        LinearExprBuilder obj = LinearExpr.NewBuilder();
        foreach (int i in allItems)
        {
            foreach (int b in allBins)
            {
                obj.AddTerm(x[i, b], Values[i]);
            }
        }
        model.Maximize(obj);

        // Solve
        CpSolver solver = new CpSolver();
        CpSolverStatus status = solver.Solve(model);

        // Print solution.
        // Check that the problem has a feasible solution.
        if (status == CpSolverStatus.Optimal || status == CpSolverStatus.Feasible)
        {
            Console.WriteLine($"Total packed value: {solver.ObjectiveValue}");
            double TotalWeight = 0.0;
            foreach (int b in allBins)
            {
                double BinWeight = 0.0;
                double BinValue = 0.0;
                Console.WriteLine($"Bin {b}");
                foreach (int i in allItems)
                {
                    if (solver.BooleanValue(x[i, b]))
                    {
                        Console.WriteLine($"Item {i} weight: {Weights[i]} values: {Values[i]}");
                        BinWeight += Weights[i];
                        BinValue += Values[i];
                    }
                }
                Console.WriteLine("Packed bin weight: " + BinWeight);
                Console.WriteLine("Packed bin value: " + BinValue);
                TotalWeight += BinWeight;
            }
            Console.WriteLine("Total packed weight: " + TotalWeight);
        }
        else
        {
            Console.WriteLine("No solution found.");
        }

        Console.WriteLine("Statistics");
        Console.WriteLine($"  conflicts: {solver.NumConflicts()}");
        Console.WriteLine($"  branches : {solver.NumBranches()}");
        Console.WriteLine($"  wall time: {solver.WallTime()}s");
    }
}