c# 多容器 二维矩形优化排版 遗传算法代码

以下是一个简单的C#遗传算法实现多容器二维矩形优化排版的示例代码：

using System;
using System.Collections.Generic;

namespace GeneticAlgorithm
{
    class Rectangle
    {
        public int Width { get; set; }
        public int Height { get; set; }

        public Rectangle(int width, int height)
        {
            Width = width;
            Height = height;
        }
    }

    class Container
    {
        public int Width { get; set; }
        public int Height { get; set; }
        public List<Rectangle> Rectangles { get; set; }

        public Container(int width, int height)
        {
            Width = width;
            Height = height;
            Rectangles = new List<Rectangle>();
        }
    }

    class Individual
    {
        public List<Container> Containers { get; set; }
        public int Fitness { get; set; }

        public Individual(List<Container> containers)
        {
            Containers = containers;
            Fitness = CalculateFitness();
        }

        private int CalculateFitness()
        {
            // 计算适应度函数的值
            int fitness = 0;
            foreach (var container in Containers)
            {
                int maxHeight = 0;
                foreach (var rectangle in container.Rectangles)
                {
                    maxHeight = Math.Max(maxHeight, rectangle.Height);
                }
                fitness += maxHeight;
            }
            return fitness;
        }
    }

    class GeneticAlgorithm
    {
        private Random random;
        private int populationSize;
        private double crossoverProbability;
        private double mutationProbability;
        private int maxGenerations;
        private List<Container> containers;
        private List<Individual> population;
        private int currentGeneration;

        public GeneticAlgorithm(int populationSize, double crossoverProbability, double mutationProbability, int maxGenerations, List<Container> containers)
        {
            random = new Random();
            this.populationSize = populationSize;
            this.crossoverProbability = crossoverProbability;
            this.mutationProbability = mutationProbability;
            this.maxGenerations = maxGenerations;
            this.containers = containers;
            population = new List<Individual>();
            currentGeneration = 0;
        }

        private void InitializePopulation()
        {
            // 初始化种群
            for (int i = 0; i < populationSize; i++)
            {
                List<Container> containersCopy = new List<Container>();
                foreach (var container in containers)
                {
                    Container containerCopy = new Container(container.Width, container.Height);
                    containersCopy.Add(containerCopy);
                }
                Individual individual = new Individual(containersCopy);
                population.Add(individual);
            }
        }

        private Individual SelectParent()
        {
            // 选择父代个体
            int index1 = random.Next(populationSize);
            int index2 = random.Next(populationSize);
            return population[index1].Fitness < population[index2].Fitness ? population[index1] : population[index2];
        }

        private void Crossover(Individual parent1, Individual parent2, out Individual child1, out Individual child2)
        {
            // 交叉操作
            child1 = new Individual(new List<Container>(parent1.Containers));
            child2 = new Individual(new List<Container>(parent2.Containers));
            if (random.NextDouble() < crossoverProbability)
            {
                int containerIndex = random.Next(containers.Count);
                Container container1 = child1.Containers[containerIndex];
                Container container2 = child2.Containers[containerIndex];
                for (int i = 0; i < container1.Rectangles.Count; i++)
                {
                    if (random.NextDouble() < 0.5)
                    {
                        Rectangle rectangle = container1.Rectangles[i];
                        container1.Rectangles.RemoveAt(i);
                        container2.Rectangles.Add(rectangle);
                    }
                }
            }
        }

        private void Mutate(Individual individual)
        {
            // 变异操作
            if (random.NextDouble() < mutationProbability)
            {
                int containerIndex = random.Next(containers.Count);
                Container container = individual.Containers[containerIndex];
                int rectangleIndex = random.Next(container.Rectangles.Count);
                Rectangle rectangle = container.Rectangles[rectangleIndex];
                int newX = random.Next(container.Width - rectangle.Width);
                int newY = random.Next(container.Height - rectangle.Height);
                rectangle.Width = newX;
                rectangle.Height = newY;
            }
        }

        private void Reproduce()
        {
            // 繁殖操作，生成新一代
            List<Individual> newPopulation = new List<Individual>();
            while (newPopulation.Count < populationSize)
            {
                Individual parent1 = SelectParent();
                Individual parent2 = SelectParent();
                Individual child1, child2;
                Crossover(parent1, parent2, out child1, out child2);
                Mutate(child1);
                Mutate(child2);
                newPopulation.Add(child1);
                newPopulation.Add(child2);
            }
            population = newPopulation;
        }

        private Individual GetBestIndividual()
        {
            // 获取适应度函数值最小的个体
            Individual bestIndividual = population[0];
            foreach (var individual in population)
            {
                if (individual.Fitness < bestIndividual.Fitness)
                {
                    bestIndividual = individual;
                }
            }
            return bestIndividual;
        }

        public Individual Run()
        {
            InitializePopulation();
            while (currentGeneration < maxGenerations)
            {
                Reproduce();
                currentGeneration++;
            }
            return GetBestIndividual();
        }
    }

    class Program
    {
        static void Main(string[] args)
        {
            // 创建容器
            Container container1 = new Container(100, 100);
            Container container2 = new Container(200, 200);
            Container container3 = new Container(300, 300);

            // 创建矩形
            Rectangle rectangle1 = new Rectangle(50, 50);
            Rectangle rectangle2 = new Rectangle(100, 100);
            Rectangle rectangle3 = new Rectangle(150, 150);

            // 将矩形添加到容器中
            container1.Rectangles.Add(rectangle1);
            container1.Rectangles.Add(rectangle2);
            container2.Rectangles.Add(rectangle1);
            container2.Rectangles.Add(rectangle2);
            container2.Rectangles.Add(rectangle3);
            container3.Rectangles.Add(rectangle1);
            container3.Rectangles.Add(rectangle2);
            container3.Rectangles.Add(rectangle3);

            // 创建容器列表
            List<Container> containers = new List<Container>();
            containers.Add(container1);
            containers.Add(container2);
            containers.Add(container3);

            // 运行遗传算法
            GeneticAlgorithm ga = new GeneticAlgorithm(100, 0.8, 0.1, 1000, containers);
            Individual bestIndividual = ga.Run();

            // 输出结果
            Console.WriteLine("最优解：");
            foreach (var container in bestIndividual.Containers)
            {
                Console.WriteLine($"容器({container.Width}, {container.Height}):");
                foreach (var rectangle in container.Rectangles)
                {
                    Console.WriteLine($"矩形({rectangle.Width}, {rectangle.Height})");
                }
            }
            Console.WriteLine($"适应度函数值：{bestIndividual.Fitness}");
            Console.ReadLine();
        }
    }
}

在上面的示例代码中，我们定义了三个类：`Rectangle`、`Container`和`Individual`。`Rectangle`表示矩形，包含宽度和高度两个属性。`Container`表示容器，包含宽度、高度和矩形列表三个属性。`Individual`表示个体，包含容器列表和适应度函数值两个属性。

在`GeneticAlgorithm`类中，我们定义了遗传算法的各种操作，包括初始化种群、选择父代个体、交叉操作、变异操作、繁殖操作和获取适应度函数值最小的个体等。在`Main`函数中，我们首先创建了三个容器和三个矩形，并将矩形添加到容器中。然后，我们将容器列表传入遗传算法中，并指定种群大小、交叉概率、变异概率和最大迭代次数等参数。最后，我们运行遗传算法，并输出结果。

请注意，这只是一个简单的示例代码，需要根据具体情况进行修改和优化。