ROS平台上使用C++实现A*算法
1.概述
在ROS平台上使用C++实现A*算法,并使用map_server在rviz上进行可视化,通过插件获取起点和终点进行路径规划,并通过path显示。
2.平台
ubuntu20.04 +ROS-Noetic
3.具体步骤
假设我们已经有了一张map,将其导入rviz中,我们可以获得地图数据,然后获取起点/initialpose和终点/move_base_simple/goal的坐标,这时我们就可以编写代码进行算法层面上的路径规划,最后将该路径上所有的坐标点通过path话题传入rviz中并在上面进行可视化显示。
3.1载入地图
roscore
rosrun map_server map_server map.yaml
//载入地图的路径,一定要在yaml路径
然后输入代码打开rviz可视化平台:
rviz
左下角Add添加Map,在下拉Topic中选择/map话题,就可以看到相应的地图了。
以上完成了载入地图到rviz中的操作,但是吧,这样只是在rviz上的显示,并没有实际获取地图的数据,所以我们需要订阅/map话题,然后读取其中的数据。/map话题中的数据类型为nav_msgs/OccupancyGrid,消息具体内容如下:
geometry_msgs/Point position
现在数据类型知道了,就需要通过订阅话题/map,然后读取话题内容的地图信息存储到代码中,这里比较重要的有两点,首先是,我们一般会通过如下步骤进行订阅topic:
//主函数中
ros::NodeHandle n;
ros::Subscriber map_sub = n.subscribe("/map", 10, &MapCallback);//重写MapCallback回调函数
void MapCallback(const nav_msgs::OccupancyGrid msg){
/*code*/
}
先定义一个类,然后再写订阅代码:
class MapClass{public:int origin_x = 0;int origin_y = 0;float resolution = 0;int width = 0;int height = 0;vector<vector<int>> map_data; //采用二维容器,可以自适应调整长度,存储栅栏属性public:void MapCallback(const nav_msgs::OccupancyGrid msg);
};void MapClass::MapCallback(const nav_msgs::OccupancyGrid msg){origin_x = msg.info.origin.position.x;origin_y = msg.info.origin.position.y;resolution = msg.info.resolution; //像素:米/像素width = msg.info.width;height = msg.info.height;// for(int i=0;i<height;i++) map_data.push_back(vector<int>());// for(int i=0;i<height;i++){// for(int j=0;j<width;j++){// map_data[j].push_back(msg.data[i*width + j]);// }// } //祭奠我傻逼的获取地图数据for(int i=0;i<width;i++) map_data.push_back(vector<int>());for(int i=0;i<width;i++){for(int j=0;j<height;j++){map_data[i].push_back(msg.data[j*width+i]);}}
}
//主函数中
MapClass mapclass;
ros::NodeHandle n;
ros::Subscriber map_sub = n.subscribe("/map", 10, &MapClass::MapCallback, &mapclass);//这是重点,使用MapClass的类里的回调函数就可以把地图数据获取出来
这样我们就把获得了一个含有地图全部数据的类mapclass。
3.2 通过/initialpose在rviz设置起点
这里比较好理解,在rviz上设置起点,实际上就是鼠标在地图上选一个点,然后这个点的信息会通过/initialpose话题发布,这时只需要订阅该话题就可以获取其中信息,选取起点的步骤如下:
这个话题中的信息类型是"geometry_msgs/PoseWithCovarianceStamped",查看消息数据格式为:
rosmsg show geometry_msgs/PoseWithCovarianceStamped
std_msgs/Header header uint32 seq //帧数time stamp //时间戳string frame_id //地图参考的坐标系id
geometry_msgs/PoseWithCovariance posegeometry_msgs/Pose posegeometry_msgs/Point position //位置float64 xfloat64 yfloat64 zgeometry_msgs/Quaternion orientation //四元数姿态float64 xfloat64 yfloat64 zfloat64 wfloat64[36] covariance
这里面暂时四元数姿态用不上,具体需要frame_id和pose就够用了,订阅方式我暂时用了最简单的方法:
float init_pose_point[3]={0};
void InitPoseCallback(const geometry_msgs::PoseWithCovarianceStamped msg){init_pose_point[0] = 1; //标志位,1表示这个起点没被用过,0表示被用完了init_pose_point[1] = msg.pose.pose.position.x; init_pose_point[2] = msg.pose.pose.position.y;
}
//main中
ros::Subscriber init_pose_sub = n.subscribe("/initialpose",10, &InitPoseCallback);
这样就获得了初始点的坐标。
3.3通过move_base_simple/goal获取终点坐标
消息类型为geometry_msgs::PoseStamped
消息格式为
rosmsg show geometry_msgs/PoseStamped
std_msgs/Header headeruint32 seq //帧数time stamp //时间戳string frame_id //地图参考的坐标系id
geometry_msgs/Pose posegeometry_msgs/Point position //位置float64 xfloat64 yfloat64 zgeometry_msgs/Quaternion orientation //四元数姿态float64 xfloat64 yfloat64 zfloat64 w
3.4、A*算法实现
通过上述三个步骤,已经获取了地图数据、起点和终点的坐标,那么就可以编写A*算法进行路径规划了。
(1)A*算法理论:A*算法理论视频
(2)算法实现:A*算法理论文章
这篇文章里面的A*算法简单易懂,并且代码对于初学C++的我来说显得很厉害,很简单,包括对类、vector、list的指针的使用。略微修改后如下:
astar.h
#pragma once
/*
//A*算法对象类
*/
#include <vector>
#include <list>
const int kCost1 = 10; //直移一格消耗
const int kCost2 = 17; //斜移一格消耗struct Point
{int x, y; //点坐标,这里为了方便按照C++的数组来计算,x代表横排,y代表竖列int F, G, H; //F=G+HPoint *parent; //parent的坐标,这里没有用指针,从而简化代码Point(int _x, int _y) :x(_x), y(_y), F(0), G(0), H(0), parent(NULL) //变量初始化{}
};class Astar
{
public://构造函数,输入为_maze的地址void InitAstar(std::vector<std::vector<int>> &_maze);//路径点liststd::list<Point *> GetPath(Point &startPoint, Point &endPoint, bool isIgnoreCorner);
private://寻找路径的函数,返回的是点的指针Point *findPath(Point &startPoint, Point &endPoint, bool isIgnoreCorner);//获取周围点std::vector<Point *> getSurroundPoints(const Point *point, bool isIgnoreCorner) const;//判断某点是否可以用于下一步判断bool isCanreach(const Point *point, const Point *target, bool isIgnoreCorner) const; //判断开启/关闭列表中是否包含某点Point *isInList(const std::list<Point *> &list, const Point *point) const;//从开启列表中返回F值最小的节点Point *getLeastFpoint(); //计算FGH值int calcG(Point *temp_start, Point *point);int calcH(Point *point, Point *end);int calcF(Point *point);
private:std::vector<std::vector<int>> maze;std::list<Point *> openList; //开启列表std::list<Point *> closeList; //关闭列表
public:bool maze_flag = true; //加载地图的标志位,省得多次加载浪费资源
};
astar.cpp
#include <math.h>
#include "astar.h"
#include <ros/ros.h>
void Astar::InitAstar(std::vector<std::vector<int>> &_maze){maze = _maze;
}Point *Astar::getLeastFpoint() //返回最小F的点
{if (!openList.empty()){auto resPoint = openList.front(); //返回第一个元素for (auto &point : openList){if (point->F<resPoint->F)resPoint = point;}return resPoint;}return NULL;
}
Point *Astar::findPath(Point &startPoint, Point &endPoint, bool isIgnoreCorner)
{openList.push_back(new Point(startPoint.x, startPoint.y)); //置入起点,拷贝开辟一个节点,内外隔离while (!openList.empty()){auto curPoint = getLeastFpoint(); //找到F值最小的点//ROS_INFO("%d,%d",curPoint->x,curPoint->y);openList.remove(curPoint); //从开启列表中删除closeList.push_back(curPoint); //放到关闭列表//1,找到当前周围八个格中可以通过的格子//ROS_INFO("改找周围点了");auto surroundPoints = getSurroundPoints(curPoint, isIgnoreCorner);for (auto &target : surroundPoints){//ROS_INFO("开始判断周围点了");//2,对某一个格子,如果它不在开启列表中,加入到开启列表,设置当前格为其父节点,计算F G Hif (!isInList(openList, target)){//BROS_INFO("判断每个目标点");target->parent = curPoint; //给子节点放父节点target->G = calcG(curPoint, target);target->H = calcH(target, &endPoint);target->F = calcF(target);openList.push_back(target);}//3,对某一个格子,它在开启列表中,计算G值, 如果比原来的大, 就什么都不做, 否则设置它的父节点为当前点,并更新G和Felse{int tempG = calcG(curPoint, target);if (tempG<target->G){target->parent = curPoint;target->G = tempG;target->F = calcF(target);}}{ //如果到这一步,就说明已经结束了Point *resPoint = isInList(openList, &endPoint);if (resPoint)return resPoint; //返回列表里的节点指针,不要用原来传入的endpoint指针,因为发生了深拷贝}}}return NULL;
}std::list<Point *> Astar::GetPath(Point &startPoint, Point &endPoint, bool isIgnoreCorner)
{//ROS_INFO("开始坐标:(%d,%d),结束坐标:(%d,%d)",startPoint.x,startPoint.y,endPoint.x,endPoint.y);Point *result = findPath(startPoint, endPoint, isIgnoreCorner);//ROS_INFO("找到路径了");std::list<Point *> path;//返回路径,如果没找到路径,返回空链表while (result){path.push_front(result);result = result->parent;}// 清空临时开闭列表,防止重复执行GetPath导致结果异常openList.clear();closeList.clear();ROS_INFO("获取路径成功");return path;
}Point *Astar::isInList(const std::list<Point *> &list, const Point *point) const
{//判断某个节点是否在列表中,这里不能比较指针,因为每次加入列表是新开辟的节点,只能比较坐标for (auto p : list){if (p->x == point->x&&p->y == point->y)return p;}return NULL;
}bool Astar::isCanreach(const Point *point, const Point *target, bool isIgnoreCorner) const
{//ROS_INFO("判断是否可以到达");//判断是否可以到达if (target->x<0 || target->x>maze.size()-1 //x不在地图边界内|| target->y<0 || target->y>maze[0].size() - 1 //y不在地图边界内|| maze[target->x][target->y] == 100 //障碍物|| (target->x == point->x&&target->y == point->y) //与当前节点重合|| isInList(closeList, target)) //如果点与当前节点重合、超出地图、是障碍物、或者在关闭列表中,返回false{return false;}else{//ROS_INFO("可以到达");if (abs(point->x - target->x) + abs(point->y - target->y) == 1) //非斜角可以{//ROS_INFO("非斜角");return true;}else{//斜对角要判断是否绊住,意思就是我能不能从直线走到斜对角去if (maze[point->x][target->y] == 0 && maze[target->x][point->y] == 0){//ROS_INFO("拌不住");return true;}else{//ROS_INFO("绊住了");return isIgnoreCorner; //到这就说明是真的被绊住了,那就看用不用考虑会被绊住的条件}}}
}std::vector<Point *> Astar::getSurroundPoints(const Point *point, bool isIgnoreCorner) const
{std::vector<Point *> surroundPoints;for (int x = point->x - 1; x <= point->x + 1; x++){for (int y = point->y - 1; y <= point->y + 1; y++){if (isCanreach(point, new Point(x, y), isIgnoreCorner)){surroundPoints.push_back(new Point(x, y));//ROS_INFO("在放点");}}}return surroundPoints;
}int Astar::calcG(Point *temp_start, Point *point){//如果x,y绝对值加起来是0那么就是一个点,如果是差1,那么就是差一格,如果不是1那就是斜着走的int extraG = (abs(point->x - temp_start->x) + abs(point->y - temp_start->y)) == 1 ? kCost1 : kCost2; //如果是初始节点,则其父节点是空,父节点的G就为0int parentG = point->parent == NULL ? 0 : point->parent->G; return parentG + extraG;
}int Astar::calcH(Point *point, Point *end){//return sqrt((double)(end->x - point->x)*(double)(end->x - point->x) + (double)(end->y - point->y)*(double)(end->y - point->y))*kCost1;return (abs(end->x - point->x)+abs(end->y - point->y)) * kCost1;
}int Astar::calcF(Point *point){return point->G + point->H;
}
这样就就完成了A*算法的修改,主函数具体调用为:
main.cpp
#include <ros/ros.h>
#include <geometry_msgs/PoseStamped.h>
#include <geometry_msgs/PoseWithCovarianceStamped.h>
#include <nav_msgs/OccupancyGrid.h>
#include <nav_msgs/Path.h>
#include <astar.h>
using namespace std;
int main(int argc, char *argv[])
{setlocale(LC_ALL,"");ros::init(argc, argv, "astar_planning_node");MapClass mapclass;//创建地图类Astar astar;bool IsGetPath =false;list<Point *> path_copy; ros::NodeHandle n;ros::Subscriber map_sub = n.subscribe("/map", 10, &MapClass::MapCallback, &mapclass);ros::Subscriber init_pose_sub = n.subscribe("/initialpose",10, &InitPoseCallback);ros::Subscriber goal_pose_sub = n.subscribe("/move_base_simple/goal",10, &GoalPoseCallback);ros::Publisher path_pub = n.advertise<nav_msgs::Path>("path_Astar",10); nav_msgs::Path pathforpub;pathforpub.header.frame_id = "map";pathforpub.header.stamp = ros::Time::now();ros::Rate r(1);while (ros::ok()){if(mapclass.height>0 && mapclass.width> 0 && astar.maze_flag){astar.maze_flag = false;astar.InitAstar(mapclass.map_data);}if(init_pose_point[0]&&goal_pose_point[0]){init_pose_point[0] = 0;init_pose_point[1] = init_pose_point[1]/mapclass.resolution;init_pose_point[2] = init_pose_point[2]/mapclass.resolution;goal_pose_point[0] = 0;goal_pose_point[1] = goal_pose_point[1]/mapclass.resolution;goal_pose_point[2] = goal_pose_point[2]/mapclass.resolution;// ROS_INFO("%f,%f",init_pose_point[1],init_pose_point[2]);// ROS_INFO("%f,%f",goal_pose_point[1],goal_pose_point[2]);Point start((int)init_pose_point[1],(int)init_pose_point[2]);Point end((int)goal_pose_point[1],(int)goal_pose_point[2]);// ROS_INFO("%d,%d",start.x,start.y);// ROS_INFO("%d,%d",end.x,end.y);//A星算法寻找路径list<Point *> path = astar.GetPath(start, end, false); //这个其实是反过来的pathforpub.poses.clear();for (auto p: path) {geometry_msgs::PoseStamped this_pose_stamped;this_pose_stamped.header.frame_id="map";this_pose_stamped.header.stamp = ros::Time::now();this_pose_stamped.pose.position.x = p->x*mapclass.resolution;this_pose_stamped.pose.position.y = p->y*mapclass.resolution;this_pose_stamped.pose.orientation.x = 0;this_pose_stamped.pose.orientation.y = 0;this_pose_stamped.pose.orientation.z = 0;this_pose_stamped.pose.orientation.w = 0; //直接让四元数全为0了,省得pathforpub.poses.push_back(this_pose_stamped);path_pub.publish(pathforpub);}//path_pub.publish(pathforpub);}path_pub.publish(pathforpub);ros::spinOnce();r.sleep();}return 0;
}
3.5、发布规划的path到rviz上并显示
主函数我其实已经放了publisher一个path_pub,话题名称为/path,消息类型为nav_msgs::Path,具体格式为:
std_msgs/Header headeruint32 seqtime stampstring frame_id
geometry_msgs/PoseStamped[] poses
//这里是关键,这里其实是PoseStamped消息的一维数组,所以需要定义路径上每个点的PoseStamped消息,然后用push_back放在末尾std_msgs/Header headeruint32 seqtime stampstring frame_idgeometry_msgs/Pose posegeometry_msgs/Point positionfloat64 xfloat64 yfloat64 zgeometry_msgs/Quaternion orientationfloat64 xfloat64 yfloat64 zfloat64 w
3.6 编写launch文件
<launch><node name="map_server" pkg="map_server" type="map_server" args="$(find astar_planning)/map/map.yaml"/><node pkg="rviz" type="rviz" name="rviz" args="-d $(find astar_planning)/rviz/astar_planning.rviz"/><node pkg="astar_planning" type="astar_planning_node" name="astar_planning_node"/>
</launch>
4、效果图
5.代码地址
https://github.com/zhuofsky/ros/tree/master/catkin_ws/src/astar_planning
6.参考
https://zhuanlan.zhihu.com/p/666365243