pso.cpp

微粒群算法与其它进化类算法相类似

//附录1、标准微粒群算法源程序
/*该程序用于计算Goldstein-Price函数的函数最小值 */
//库文件
#include"stdio.h"
#include"stdlib.h"
#include"time.h"
#include"math.h"
//随机数定义
#define rdint(i) (rand()%(int)(i))
#define rdft() (float)((double)rdint(16384)/(16383.0))
#define rnd(a,b) (rdint((int)(b)-(int)(a)+1)+(int)(a))
//宏定义
#define POPSIZE 20
#define DIMENSION 2
//全局变量定义
float W=1.0;     //W=0.0;//*
float C1=1.8;
float C2=1.8;
float VMAX=2;
float XMIN=-2.0;
float XMAX=2.0;
float P[DIMENSION];
float PBEST;

struct indi
{
	float number[DIMENSION];
	float best[DIMENSION];
	float bestfitness;
	float fitness;
	float speed[DIMENSION];
}individual[POPSIZE];

void initiate(void);
void calculation(int number);
void globalbest(int number);
void localbest(int number);
//void generate(int number);
/*
void generate(int num)
{
	int j;
	for(j=0;j<DIMENSION;j++)
		individual[num].number[j]=rdft()*(XMAX-XMIN)+XMIN;
	calculation(num);
	individual[num].bestfitness=individual[num].fitness;
	for(j=0;j<DIMENSION;j++)
		individual[num].best[j]=individual[num].number[j];
}
*/
//程序初始化定义
void initiate()
{
	int i,j;
	for(i=0;i<POPSIZE;i++)
		for(j=0;j<DIMENSION;j++)
			individual[i].number[j]=rdft()*(XMAX-XMIN)+XMIN;

	for(i=0;i<POPSIZE;i++)
		for(j=0;j<DIMENSION;j++)
			individual[i].speed[j]=VMAX*rdft();

	for(i=0;i<POPSIZE;i++)
		for(j=0;j<DIMENSION;j++)
			individual[i].best[j]=individual[i].number[j];

	for(i=0;i<POPSIZE;i++)
		calculation(i);

	for(i=0;i<POPSIZE;i++)
		individual[i].bestfitness=individual[i].fitness;

	globalbest(0);
}
//微粒历史最优位置修改程序
void localbest(int number)
{
	int i;
	if(individual[number].bestfitness>individual[number].fitness)
		for(i=0;i<DIMENSION;i++)
			individual[number].best[i]=individual[number].number[i];

	individual[number].bestfitness=individual[number].fitness;
}
//种群历史最优位置修改程序
void globalbest(int number)
{
	int i,j;
	float s=0;
	int flag=0;
	if(number==0)
	{
		s=individual[0].fitness;
		flag=0;
		for(i=1;i<POPSIZE;i++)
			if(individual[i].fitness<s)
			{
				s=individual[i].fitness;
				flag=i;
			}
		for(i=0;i<DIMENSION;i++)
			P[i]=individual[flag].number[i];
		PBEST=individual[flag].fitness;
	}
	else
	{
		for(i=0;i<POPSIZE;i++)
		if(individual[i].bestfitness<PBEST)
		{
			for(j=0;j<DIMENSION;j++)
				P[j]=individual[i].best[j];
			PBEST=individual[i].bestfitness;
		}
	}
}
//适应值函数计算程序
void calculation(int num)
{
	int i;
	float s=0.0,h=0.0;
	s=pow(individual[num].number[0]+individual[num].number[1]+1.0,2.0)*(19.0-14.0*individual[num].number[0]+3.0*pow(individual[num].number[0],2.0)-14.0*individual[num].number[1]+6.0*individual[num].number[0]*individual[num].number[1]+3.0*pow(individual[num].number[1],2.0))+1.0;
	h=30.0+pow(2.0*individual[num].number[0]-3.0*individual[num].number[1],2.0)*(18.0-32.0*individual[num].number[0]+12.0*pow(individual[num].number[0],2.0)+48.0*individual[num].number[1]-36.0*individual[num].number[0]*individual[num].number[1]+27.0*pow(individual[num].number[1],2.0));
	individual[num].fitness=s*h;
}
//主程序
main()
{
	int i,j,k,t,total=0;
	float sum=0;
	srand(time(0));
	for(j=0;j<50;j++)
	{
		initiate();
		for(i=0;i<500;i++)
		{
			W=1.0-i*0.6/499;
			if((PBEST-3.0)<0.0001)
			{
				total++;
				sum=sum+i;
				break;
			}
			else
			{
				for(k=0;k<POPSIZE;k++)
				{
					for(t=0;t<DIMENSION;t++)
					{
						individual[k].speed[t]=W*individual[k].speed[t]+C1*rdft()*(individual[k].best[t]-individual[k].number[t])+C2*rdft()*(P[t]-individual[k].number[t]);
						if(individual[k].speed[t]>VMAX)
							individual[k].speed[t]=VMAX;
						individual[k].number[t]=individual[k].number[t]+individual[k].speed[t];
						if(individual[k].number[t]<XMIN)
							individual[k].number[t]=2*XMIN-individual[k].number[t];
						if(individual[k].number[t]>XMAX)
							individual[k].number[t]=2*XMAX-individual[k].number[t];
					}
					calculation(k);
					localbest(k);
				}
				globalbest(1);
			}
		}
		printf("%d,%f\t",i,PBEST);
	}
	printf("Total number is%d\n",total);
	printf("Sum is %f\n",sum);
	printf("Average is %f\n",sum/total);
}
/*
main()
{
	int i,j,k,t,total=0,flag=-1,fflag;
	float sum=0,temp[DIMENSION],s;
	srand(time(NULL));
	for(j=0;j<50;j++)
	{
		initiate();
		for(i=0;i<500;i++)
		{
			flag=-1;
		//寻找是否存在该最优微粒
			for(k=0;k<POPSIZE;k++)
			{
				if((individual[k].fitness==individual[k].bestfitness)&&(individual[k].fitness==PBEST))
				{
					flag=k;
					break;
				}
			}
			if(flag!=-1)
			generate(flag);
			for(k=0;k<POPSIZE;k++)
				if(k!=flag)
					{
						for(t=0;t<DIMENSION;t++)
						{
							individual[k].speed[t]=W*individual[k].speed[t]+C1*rdft()*(individual[k].best[t]-individual[k].number[t])+C2*rdft()*(P[t]-individual[k].number[t]);
							if(individual[k].speed[t]>VMAX)
								individual[k].speed[t]=VMAX;
							individual[k].number[t]=individual[k].number[t]+individual[k].speed[t];
							if(individual[k].number[t]<XMIN)
								individual[k].number[t]=2*XMIN-individual[k].number[t];
							if(individual[k].number[t]>XMAX)
								individual[k].number[t]=2*XMAX-individual[k].number[t];
						}
						calculation(k);
						localbest(k);
					}
					globalbest(1);
					if((PBEST-3.0)<0.00001)
					{
						total++;
						sum=sum+i;
						break;
					}
			}
				printf("%d,%f\t",i,PBEST);
	}
	printf("Total number is%d\n",total);
	printf("Sum is %f\n",sum);
	printf("Average is %f\n",sum/total);
}
  */