组合优化类粒子群算法代码

以下是一个基于粒子群算法的组合优化问题的 Python 代码示例：

import random

class Particle:
    def __init__(self, num_variables):
        self.position = [random.randint(0, 1) for i in range(num_variables)]
        self.velocity = [0 for i in range(num_variables)]
        self.fitness = 0
        self.best_position = self.position[:]
        self.best_fitness = 0

class PSO:
    def __init__(self, num_particles, num_iterations, num_variables, fitness_function):
        self.num_particles = num_particles
        self.num_iterations = num_iterations
        self.num_variables = num_variables
        self.fitness_function = fitness_function
        self.swarm = [Particle(num_variables) for i in range(num_particles)]
        self.best_particle = self.swarm[0]

    def optimize(self):
        for i in range(self.num_iterations):
            for particle in self.swarm:
                particle.fitness = self.fitness_function(particle.position)
                if particle.fitness > particle.best_fitness:
                    particle.best_fitness = particle.fitness
                    particle.best_position = particle.position[:]
                if particle.fitness > self.best_particle.best_fitness:
                    self.best_particle = particle

                for j in range(self.num_variables):
                    r1 = random.random()
                    r2 = random.random()
                    particle.velocity[j] = 0.5 * particle.velocity[j] + \
                                           1.0 * r1 * (particle.best_position[j] - particle.position[j]) + \
                                           1.0 * r2 * (self.best_particle.position[j] - particle.position[j])
                    particle.position[j] = 1 if random.random() < self.sigmoid(particle.velocity[j]) else 0

    def sigmoid(self, x):
        return 1.0 / (1.0 + pow(2.718, -x))

# Example usage
def fitness_function(position):
    # Evaluate the fitness of the binary string
    # For example, count the number of 1s in the string
    return sum(position)

pso = PSO(num_particles=10, num_iterations=100, num_variables=5, fitness_function=fitness_function)
pso.optimize()
print(pso.best_particle.best_position)

这个代码示例中，我们定义了一个 `Particle` 类，它代表了一个粒子，包括当前位置、速度、适应度、最佳位置和最佳适应度。我们还定义了一个 `PSO` 类，它代表了整个粒子群算法，包括粒子群、迭代次数、变量数、适应度函数等。`optimize` 方法是整个算法的核心，它包括了粒子的更新、适应度的评估、全局最佳位置的更新等操作。

在这个示例中，我们使用了一个简单的适应度函数，即计算二进制字符串中 1 的个数。你可以将适应度函数替换成你自己的问题的适应度函数。