首页 > 编程语言 >A*算法实现最优路径规划

A*算法实现最优路径规划

时间:2024-06-21 16:00:44浏览次数:12  
标签:blocklist self 路径 father pos CELL WIDTH 算法 最优

  1.  用A*算法实现最优路径规划,绿色五角星为起点,红色四角星为终点,黑色方块为障碍物,如下图所示。简要介绍问题的估价函数、算法步骤、搜索路径、代码实现和代码运行结果。

  2. import math
    from random import randint
    import pygame
    from enum import Enum
    # 定义全局变量:地图中节点的像素大小
    CELL_WIDTH = 50  # 单元格宽度
    CELL_HEIGHT = 50  # 单元格长度
    BORDER_WIDTH = 1  # 边框宽度
    BLOCK_NUM = 5  # 地图中的障碍物数量
    class Color(Enum):
        ''' 颜色 '''
        RED = (255, 0, 0)
        GREEN = (0, 255, 0)
        BLUE = (0, 0, 255)
        WHITE = (255, 255, 255)
        BLACK = (0, 0, 0)
        GREY = (128, 128, 128)
        @staticmethod
        def random_color():
            '''设置随机颜色'''
            r = randint(0, 255)
            g = randint(0, 255)
            b = randint(0, 255)
            return (r, g, b)
    class Map(object):
        def __init__(self, mapsize):
            self.mapsize = mapsize
        def generate_cell(self, cell_width, cell_height):
            '''
            定义一个生成器,用来生成地图中的所有节点坐标
            :param cell_width: 节点宽度
            :param cell_height: 节点长度
            :return: 返回地图中的节点
            '''
            x_cell = -cell_width
            for num_x in range(self.mapsize[0] // cell_width):
                y_cell = -cell_height
                x_cell += cell_width
                for num_y in range(self.mapsize[1] // cell_height):
                    y_cell += cell_height
                    yield (x_cell, y_cell)
    class Node(object):
        def __init__(self, pos):
            self.pos = pos
            self.father = None
            self.gvalue = 0
            self.fvalue = 0
        def compute_fx(self, enode, father):
            if father == None:
                print('未设置当前节点的父节点!')
            gx_father = father.gvalue
    
            # 采用欧氏距离计算父节点到当前节点的距离
            # gx_f2n = math.sqrt((father.pos[0] - self.pos[0])**2 + (father.pos[1] - self.pos[1])**2)
            gx_f2n = abs(father.pos[0] - self.pos[0]) + abs(father.pos[1] - self.pos[1])
            gvalue = gx_f2n + gx_father
    
            # hx_n2enode = math.sqrt((self.pos[0] - enode.pos[0])**2 + (self.pos[1] - enode.pos[1])**2)
            hx_n2enode = abs(self.pos[0] - enode.pos[0]) + abs(self.pos[1] - enode.pos[1])
            fvalue = gvalue + hx_n2enode
            return gvalue, fvalue
    
        def set_fx(self, enode, father):
            self.gvalue, self.fvalue = self.compute_fx(enode, father)
            self.father = father
    
        def update_fx(self, enode, father):
            gvalue, fvalue = self.compute_fx(enode, father)
            if fvalue < self.fvalue:
                self.gvalue, self.fvalue = gvalue, fvalue
                self.father = father
    
    class AStar(object):
        def __init__(self, mapsize, pos_sn, pos_en):
            self.mapsize = mapsize  # 表示地图的投影大小,并非屏幕上的地图像素大小
            self.openlist, self.closelist, self.blocklist = [], [], []
            self.snode = Node(pos_sn)  # 用于存储路径规划的起始节点
            self.enode = Node(pos_en)  # 用于存储路径规划的目标节点
            self.cnode = self.snode  # 用于存储当前搜索到的节点
        def run(self):
            self.openlist.append(self.snode)
            while (len(self.openlist) > 0):
                # 查找openlist中fx最小的节点
                fxlist = list(map(lambda x: x.fvalue, self.openlist))
                index_min = fxlist.index(min(fxlist))
                self.cnode = self.openlist[index_min]
                del self.openlist[index_min]
                self.closelist.append(self.cnode)
                # 扩展当前fx最小的节点,并进入下一次循环搜索
                self.extend(self.cnode)
                # 如果openlist列表为空,或者当前搜索节点为目标节点,则跳出循环
                if len(self.openlist) == 0 or self.cnode.pos == self.enode.pos:
                    break
    
            if self.cnode.pos == self.enode.pos:
                self.enode.father = self.cnode.father
                return 1
            else:
                return -1
        def get_minroute(self):
            minroute = []
            current_node = self.enode
            while (True):
                minroute.append(current_node.pos)
                current_node = current_node.father
                if current_node.pos == self.snode.pos:
                    break
            minroute.append(self.snode.pos)
            minroute.reverse()
            return minroute
        def extend(self, cnode):
            nodes_neighbor = self.get_neighbor(cnode)
            for node in nodes_neighbor:
                # 判断节点node是否在closelist和blocklist中,因为closelist和blocklist中元素均为Node类,所以要用map函数转换为坐标集合
                if node.pos in list(map(lambda x: x.pos, self.closelist)) or node.pos in self.blocklist:
                    continue
                else:
                    if node.pos in list(map(lambda x: x.pos, self.openlist)):
                        node.update_fx(self.enode, cnode)
                    else:
                        node.set_fx(self.enode, cnode)
                        self.openlist.append(node)
        def setBlock(self, blocklist):
            '''
            获取地图中的障碍物节点,并存入self.blocklist列表中
            注意:self.blocklist列表中存储的是障碍物坐标,不是Node类
            :param blocklist:
            :return:
            '''
            self.blocklist.extend(blocklist)
        def get_neighbor(self, cnode):
            # offsets = [(-1,1),(0,1),(1,1),(-1,0),(1,0),(-1,-1),(0,-1),(1,-1)]
            offsets = [(-1, 0), (1, 0), (0, 1), (0, -1)]
            nodes_neighbor = []
            x, y = cnode.pos[0], cnode.pos[1]
            for os in offsets:
                x_new, y_new = x + os[0], y + os[1]
                pos_new = (x_new, y_new)
                # 判断是否在地图范围内,超出范围跳过
                if x_new < 0 or x_new > self.mapsize[0] - 1 or y_new < 0 or y_new > self.mapsize[1]:
                    continue
                nodes_neighbor.append(Node(pos_new))
            return nodes_neighbor
    def main():
        mapsize = tuple(map(int, input('请输入地图大小,以逗号隔开:').split(',')))
        pos_snode = tuple(map(int, input('请输入起点坐标,以逗号隔开:').split(',')))
        pos_enode = tuple(map(int, input('请输入终点坐标,以逗号隔开:').split(',')))
        myAstar = AStar(mapsize, pos_snode, pos_enode)
        blocklist = gen_blocks(mapsize[0], mapsize[1])
        myAstar.setBlock(blocklist)
        routelist = []  # 记录搜索到的最优路径
        if myAstar.run() == 1:
            routelist = myAstar.get_minroute()
            print(routelist)
            showresult(mapsize, pos_snode, pos_enode, blocklist, routelist)
        else:
            print('路径规划失败!')
    def gen_blocks(width, height):
        '''
        随机生成障碍物
        :param width: 地图宽度
        :param height: 地图高度
        :return:返回障碍物坐标集合
        '''
        i, blocklist = 0, []
    
        block = (1, 0)
        if block not in blocklist:
                    blocklist.append(block)
        block = (3, 1)
        if block not in blocklist:
            blocklist.append(block)
        block = (2, 3)
        if block not in blocklist:
            blocklist.append(block)
        block = (4, 3)
        if block not in blocklist:
            blocklist.append(block)
        block = (1, 5)
        if block not in blocklist:
            blocklist.append(block)
        return blocklist
    def showresult(mapsize, pos_sn, pos_en, blocklist, routelist):
        # 初始化导入的Pygame模块
        pygame.init()
        # 此处要将地图投影大小转换为像素大小,此处设地图中每个单元格的大小为CELL_WIDTH*CELL_HEIGHT像素
        mymap = Map((mapsize[0] * CELL_WIDTH, mapsize[1] * CELL_HEIGHT))
        pix_sn = (pos_sn[0] * CELL_WIDTH, pos_sn[1] * CELL_HEIGHT)
        pix_en = (pos_en[0] * CELL_WIDTH, pos_en[1] * CELL_HEIGHT)
        # 对blocklist和routelist中的坐标同样要转换为像素值
        bl_pix = list(map(transform, blocklist))
        rl_pix = list(map(transform, routelist))
        # 初始化显示的窗口并设置尺寸
        screen = pygame.display.set_mode(mymap.mapsize)
        # 设置窗口标题
        pygame.display.set_caption('A*算法路径搜索演示:')
        # 用白色填充屏幕3
        screen.fill(Color.WHITE.value)
        # 绘制屏幕中的所有单元格
        for (x, y) in mymap.generate_cell(CELL_WIDTH, CELL_HEIGHT):
            if (x, y) in bl_pix:
                # 绘制黑色的障碍物单元格,并留出2个像素的边框
                pygame.draw.rect(screen, Color.BLACK.value, (
                (x + BORDER_WIDTH, y + BORDER_WIDTH), (CELL_WIDTH - 2 * BORDER_WIDTH, CELL_HEIGHT - 2 * BORDER_WIDTH)))
            else:
                # 绘制绿色的可通行单元格,并留出2个像素的边框
                pygame.draw.rect(screen, Color.GREEN.value, (
                (x + BORDER_WIDTH, y + BORDER_WIDTH), (CELL_WIDTH - 2 * BORDER_WIDTH, CELL_HEIGHT - 2 * BORDER_WIDTH)))
        # 绘制起点和终点
        pygame.draw.circle(screen, Color.BLUE.value, (pix_sn[0] + CELL_WIDTH // 2, pix_sn[1] + CELL_HEIGHT // 2),
                           CELL_WIDTH // 2 - 1)
        pygame.draw.circle(screen, Color.RED.value, (pix_en[0] + CELL_WIDTH // 2, pix_en[1] + CELL_HEIGHT // 2),
                           CELL_WIDTH // 2 - 1)
    
        # 绘制搜索得到的最优路径
        for (x, y) in mymap.generate_cell(CELL_WIDTH, CELL_HEIGHT):
            if (x, y) in rl_pix and (x, y) != (pix_sn[0], pix_sn[1]) and (x, y) != (pix_en[0], pix_en[1]):
                pygame.draw.rect(screen, Color.GREY.value, (
                (x + BORDER_WIDTH, y + BORDER_WIDTH), (CELL_WIDTH - 2 * BORDER_WIDTH, CELL_HEIGHT - 2 * BORDER_WIDTH)))
        # pygame.draw.aalines(screen, Color.RED.value, False, rl_pix)
        keepGoing = True
        while keepGoing:
            pygame.time.delay(100)
            for event in pygame.event.get():
                if event.type == pygame.QUIT:
                    keepGoing = False
            pygame.display.flip()
    def transform(pos):
        xnew, ynew = pos[0] * CELL_WIDTH, pos[1] * CELL_HEIGHT
        return (xnew, ynew)
    if __name__ == '__main__':
        main()

标签:blocklist,self,路径,father,pos,CELL,WIDTH,算法,最优
From: https://blog.csdn.net/2201_75648765/article/details/139862972

相关文章

  • 神经网络与模式识别课程报告-卷积神经网络(CNN)算法的应用
     =======================================================================================完整的神经网络与模式识别课程报告文档下载:https://wenku.baidu.com/view/393fbc7853e2524de518964bcf84b9d528ea2c92?aggId=393fbc7853e2524de518964bcf84b9d528ea2c92&fr=catalogM......
  • 头歌机器学习实训答案 第1关:集成学习常用算法详解
    任务描述本关任务:学习集成学习的基本概念以及常用算法并编程熟悉sklearn。相关知识为了完成本关任务,你需要掌握:1.个体与集成的概念,2.常用的集成学习算法。个体和集成集成学习(ensemblelearning)通过构建并结合多个学习器来完成学习任务,有时也被称为多分类器系统(multi-class......
  • 虚拟环境 反向解析,有名分组和无名分组的反向解析,路由分发,名称空间,虚拟环境,路径
    Ⅰ反向解析【一】基础的URL配置在实际的Django项目中,经常需要获取某个具体对象的URL,为生成的内容配置URL链接。比如,我要在页面上展示一列文章列表,每个条目都是个超级链接,点击就进入该文章的详细页面。现在我们的urlconf是这么配置的:path('post/<int:pk>/',views.some_view)......
  • 深度学习各算法的优缺点和适用场景!!纯干货,建议收藏。(下篇)
    ............纯   干  货........上篇地址:深度学习各算法的优缺点和适用场景!!纯干货,建议收藏。(上篇)-CSDN博客目录废话不说,直接上干货自编码器1、标准自编码器(VanillaAutoencoder)2、稀疏自编码器(SparseAutoencoder)3、去噪自编码器(Denoisin......
  • 深度学习各算法的优缺点和适用场景!!纯干货,建议收藏。(上篇)
     ........纯  干  货..........下篇地址:深度学习各算法的优缺点和适用场景!!纯干货,建议收藏。(下篇)-CSDN博客​目录前馈神经网络1、梯度下降(GradientDescent)2、随机梯度下降(StochasticGradientDescent,SGD)3、小批量梯度下降(Mini-batchGradi......
  • 算法合集
    算法合集这里是我的算法合集:博弈论Gametheory图论Graphtheory数论Numbertheory三角函数Trigonometricfunction字符串Strings计算几何Computationgeometry数据结构Structs动态规划DynamicprogrammingLIS问题LongestIncreasingSubsequenceproble......
  • 算法神尊
    作者:是轨迹呐标记说明:code,地点,算法武技、算法功法上篇算法大陆,诺伊宗,核心弟子居住区。一个面容清秀的少年正站在元始碑之下,望着碑上的文字默默祷告。那碑上,用黄金铸就的一行代码闪闪发亮:cout<<"Hello,world!";。元始碑是算法大陆的一种古老的传承,修炼每一种语言的码农都......
  • scau程序设计与算法(个人偷懒版(前两章
    目录1.四个常见问题18104 练习使用多case解题2.c++STL的运用19116 丑数18440 走迷宫18276 走迷宫218105 银行的叫号顺序18216 银行服务18118 勇者斗恶龙18107 校赛排名18104 练习使用多case解题时间限制:1000MS 代码长度限制:10KB提交次数:0通过次数......
  • 算法合集
    算法合集这里是我的算法合集:博弈论Gametheory图论Graphtheory数论Numbertheory三角函数Trigonometricfunction字符串Strings计算几何Computationgeometry数据结构Structs动态规划DynamicprogrammingLIS问题LongestIncreasingSubsequenceproble......
  • 代码随想录算法训练营第17天 | 二叉树04
    代码随想录算法训练营第17天找树左下角的值https://leetcode.cn/problems/find-bottom-left-tree-value/找树左下角的值代码随想录https://leetcode.cn/problems/find-bottom-left-tree-value/路径总和https://leetcode.cn/problems/path-sum/description/路径总和2https......