algorithmn

算法的解释

Dijkstra

其实就是A star或者Dijkstra(基于priority queue实现的)的路径规划算法,关键是相邻点之间的cost怎么计算,怎么从可行点找到path

Navfn's optimal path is based on a path's "potential"(可能可以行走的目标). The potential is the relative cost of a

path based on the distance from the goal and from the existing path itself.(怎么计算两个点之间的距离cost) It must be noted that Navfn update's each cell's potential in the potential map, or potarr(定义的potential array变量) as it's called in navfn, as it checks that cell. This way,it can step back through the potential array to find the best possible path. The potential is determined by the cost of traversing a cell (traversability factor, hf)

and the distance away that the next cell is from the previous cell.

parameter

navfn 参数

global planner

上面两个链接一个是navfn的配置,一个是具体哪个global planner的配置,具体的global planner 是可以覆盖navfn中的配置(要是大家有相同的参数)

比如下面这个参数global planner中的可以覆盖navfn中的配置:

~<name>/allow_unknown (bool, default: true)

这个参数可以让你看见potential array的图像,看计算出的cost是怎么样子(颜色深浅代表距离起始点的远近)

~<name>/visualize_potential (bool, default: false)

code

void GlobalPlanner::initialize(std::string name, costmap_2d::Costmap2D* costmap, std::string frame_id) {
if(!old_navfn_behavior_)
convert_offset_ = 0.5;
else
convert_offset_ = 0.0; if (use_quadratic)
p_calc_ = new QuadraticCalculator(cx, cy);
else
p_calc_ = new PotentialCalculator(cx, cy); if (use_dijkstra)
{
DijkstraExpansion* de = new DijkstraExpansion(p_calc_, cx, cy);
if(!old_navfn_behavior_)
de->setPreciseStart(true);
planner_ = de;
}
else
planner_ = new AStarExpansion(p_calc_, cx, cy); if (use_grid_path)
path_maker_ = new GridPath(p_calc_);
else
path_maker_ = new GradientPath(p_calc_);
//发布一个make_plan的service
make_plan_srv_ = private_nh.advertiseService("make_plan", &GlobalPlanner::makePlanService, this); } bool GlobalPlanner::makePlanService(nav_msgs::GetPlan::Request& req, nav_msgs::GetPlan::Response& resp) {
makePlan(req.start, req.goal, resp.plan.poses);
} bool GlobalPlanner::makePlan(const geometry_msgs::PoseStamped& start, const geometry_msgs::PoseStamped& goal,
std::vector<geometry_msgs::PoseStamped>& plan) {
return makePlan(start, goal, default_tolerance_, plan);
} bool GlobalPlanner::makePlan(const geometry_msgs::PoseStamped& start, const geometry_msgs::PoseStamped& goal,
double tolerance, std::vector<geometry_msgs::PoseStamped>& plan) {
double wx = start.pose.position.x;
double wy = start.pose.position.y; if (!costmap_->worldToMap(wx, wy, start_x_i, start_y_i)) {
ROS_WARN("The robot's start position is off the global costmap. Planning will always fail, are you sure the robot has been properly localized?");
return false;
} if(old_navfn_behavior_){
start_x = start_x_i;
start_y = start_y_i;
}else{
worldToMap(wx, wy, start_x, start_y);
} wx = goal.pose.position.x;
wy = goal.pose.position.y; if (!costmap_->worldToMap(wx, wy, goal_x_i, goal_y_i)) {
ROS_WARN_THROTTLE(1.0,"The goal sent to the global planner is off the global costmap. Planning will always fail to this goal.");
return false;
}
if(old_navfn_behavior_){
goal_x = goal_x_i;
goal_y = goal_y_i;
}else{
worldToMap(wx, wy, goal_x, goal_y);
}
//clear the starting cell within the costmap because we know it can't be an obstacle
//设置起点为FREE_SPACE
clearRobotCell(start_pose, start_x_i, start_y_i); int nx = costmap_->getSizeInCellsX(), ny = costmap_->getSizeInCellsY(); //make sure to resize the underlying array that Navfn uses
p_calc_->setSize(nx, ny);
planner_->setSize(nx, ny);
path_maker_->setSize(nx, ny);
potential_array_ = new float[nx * ny]; //将costmap的四周(边界)变为LETHAL_OBSTACLE
outlineMap(costmap_->getCharMap(), nx, ny, costmap_2d::LETHAL_OBSTACLE); // 寻找potential array
bool found_legal = planner_->calculatePotentials(costmap_->getCharMap(), start_x, start_y, goal_x, goal_y,
nx * ny * 2, potential_array_);
//对终点的处理
if(!old_navfn_behavior_)
planner_->clearEndpoint(costmap_->getCharMap(), potential_array_, goal_x_i, goal_y_i, 2);
if(publish_potential_)
publishPotential(potential_array_); if (found_legal) {
//extract the plan,提取路径
if (getPlanFromPotential(start_x, start_y, goal_x, goal_y, goal, plan)) {
//make sure the goal we push on has the same timestamp as the rest of the plan
geometry_msgs::PoseStamped goal_copy = goal;
goal_copy.header.stamp = ros::Time::now();
plan.push_back(goal_copy);
} else {
ROS_ERROR("Failed to get a plan from potential when a legal potential was found. This shouldn't happen.");
}
}else{
ROS_ERROR("Failed to get a plan.");
} // add orientations if needed,对方向的处理
orientation_filter_->processPath(start, plan); //publish the plan for visualization purposes
publishPlan(plan);
delete potential_array_;
return !plan.empty(); nx * ny * 2, potential_array_);
}

主要是以下三个函数

可能不同的配置有不同的实现,但是每个函数的实现功能是一样的。

计算所有的可行点

namespace global_planner {
bool DijkstraExpansion::calculatePotentials(unsigned char* costs, double start_x, double start_y, double end_x, double end_y,
int cycles, float* potential) {
cells_visited_ = 0;
// priority buffers
threshold_ = lethal_cost_;
currentBuffer_ = buffer1_;
currentEnd_ = 0;
nextBuffer_ = buffer2_;
nextEnd_ = 0;
overBuffer_ = buffer3_;
overEnd_ = 0;
memset(pending_, 0, ns_ * sizeof(bool));
std::fill(potential, potential + ns_, POT_HIGH); // set goal
int k = toIndex(start_x, start_y); if(precise_)
{
double dx = start_x - (int)start_x, dy = start_y - (int)start_y;
dx = floorf(dx * 100 + 0.5) / 100;
dy = floorf(dy * 100 + 0.5) / 100;
potential[k] = neutral_cost_ * 2 * dx * dy;
potential[k+1] = neutral_cost_ * 2 * (1-dx)*dy;
potential[k+nx_] = neutral_cost_*2*dx*(1-dy);
potential[k+nx_+1] = neutral_cost_*2*(1-dx)*(1-dy);//*/ push_cur(k+2);
push_cur(k-1);
push_cur(k+nx_-1);
push_cur(k+nx_+2); push_cur(k-nx_);
push_cur(k-nx_+1);
push_cur(k+nx_*2);
push_cur(k+nx_*2+1);
}else{
potential[k] = 0;
push_cur(k+1);
push_cur(k-1);
push_cur(k-nx_);
push_cur(k+nx_);
} int nwv = 0; // max priority block size
int nc = 0; // number of cells put into priority blocks
int cycle = 0; // which cycle we're on // set up start cell
int startCell = toIndex(end_x, end_y); for (; cycle < cycles; cycle++) // go for this many cycles, unless interrupted
{
//
if (currentEnd_ == 0 && nextEnd_ == 0) // priority blocks empty
return false; // stats
nc += currentEnd_;
if (currentEnd_ > nwv)
nwv = currentEnd_; // reset pending_ flags on current priority buffer
int *pb = currentBuffer_;
int i = currentEnd_;
while (i-- > 0)
pending_[*(pb++)] = false; // process current priority buffer
pb = currentBuffer_;
i = currentEnd_;
while (i-- > 0)
updateCell(costs, potential, *pb++); // swap priority blocks currentBuffer_ <=> nextBuffer_
currentEnd_ = nextEnd_;
nextEnd_ = 0;
pb = currentBuffer_; // swap buffers
currentBuffer_ = nextBuffer_;
nextBuffer_ = pb; // see if we're done with this priority level
if (currentEnd_ == 0) {
threshold_ += priorityIncrement_; // increment priority threshold
currentEnd_ = overEnd_; // set current to overflow block
overEnd_ = 0;
pb = currentBuffer_; // swap buffers
currentBuffer_ = overBuffer_;
overBuffer_ = pb;
} // check if we've hit the Start cell
if (potential[startCell] < POT_HIGH)
break;
}
//ROS_INFO("CYCLES %d/%d ", cycle, cycles);
if (cycle < cycles)
return true; // finished up here
else
return false;计算路径path
}
}

怎么计算一个点的可行点

namespace global_planner {

float QuadraticCalculator::calculatePotential(float* potential, unsigned char cost, int n, float prev_potential) {
// get neighbors
float u, d, l, r;namespace
l = potential[n - 1];
r = potential[n + 1];
u = potential[n - nx_];
d = potential[n + nx_];
// ROS_INFO("[Update] c: %f l: %f r: %f u: %f d: %f\n",
// potential[n], l, r, u, d);
// ROS_INFO("[Update] cost: %d\n", costs[n]); // find lowest, and its lowest neighbor
float ta, tc;
if (l < r)
tc = l;
else
tc = r;
if (u < d)
ta = u;
else
ta = d; float hf = cost; // traversability factor
float dc = tc - ta; // relative cost between ta,tc
if (dc < 0) // tc is lowest
{
dc = -dc;
ta = tc;
} // calculate new potential
if (dc >= hf) // if too large, use ta-only update
return ta + hf;
else // two-neighbor interpolation update
{
// use quadratic approximation
// might speed this up through table lookup, but still have to
// do the divide
float d = dc / hf;
float v = -0.2301 * d * d + 0.5307 * d + 0.7040;
return ta + hf * v;
}
}
};

从可行点中计算路径path

bool GradientPath::getPath(float* potential, double start_x, double start_y, double goal_x, double goal_y, std::vector<std::pair<float, float> >& path) {
std::pair<float, float> current;
int stc = getIndex(goal_x, goal_y); // set up offset
float dx = goal_x - (int)goal_x;
float dy = goal_y - (int)goal_y;
int ns = xs_ * ys_;
memset(gradx_, 0, ns * sizeof(float));
memset(grady_, 0, ns * sizeof(float)); int c = 0;
while (c++<ns*4) {
// check if near goal
double nx = stc % xs_ + dx, ny = stc / xs_ + dy; if (fabs(nx - start_x) < .5 && fabs(ny - start_y) < .5) {
current.first = start_x;
current.second = start_y;
path.push_back(current);
return true;
} if (stc < xs_ || stc > xs_ * ys_ - xs_) // would be out of bounds
{
printf("[PathCalc] Out of bounds\n");
return false;
} current.first = nx;
current.second = ny; //ROS_INFO("%d %d | %f %f ", stc%xs_, stc/xs_, dx, dy); path.push_back(current); bool oscillation_detected = false;
int npath = path.size();
if (npath > 2 && path[npath - 1].first == path[npath - 3].first
&& path[npath - 1].second == path[npath - 3].second) {
ROS_DEBUG("[PathCalc] oscillation detected, attempting fix.");
oscillation_detected = true;
} int stcnx = stc + xs_;
int stcpx = stc - xs_; // check for potentials at eight positions near cell
if (potential[stc] >= POT_HIGH || potential[stc + 1] >= POT_HIGH || potential[stc - 1] >= POT_HIGH
|| potential[stcnx] >= POT_HIGH || potential[stcnx + 1] >= POT_HIGH || potential[stcnx - 1] >= POT_HIGH
|| potential[stcpx] >= POT_HIGH || potential[stcpx + 1] >= POT_HIGH || potential[stcpx - 1] >= POT_HIGH
|| oscillation_detected) {
ROS_DEBUG("[Path] Pot fn boundary, following grid (%0.1f/%d)", potential[stc], (int) path.size());
// check eight neighbors to find the lowest
int minc = stc;
int minp = potential[stc];
int st = stcpx - 1;
if (potential[st] < minp) {
minp = potential[st];
minc = st;
}
st++;
if (potential[st] < minp) {
minp = potential[st];
minc = st;
}
st++;
if (potential[st] < minp) {
minp = potential[st];
minc = st;
}
st = stc - 1;
if (potential[st] < minp) {
minp = potential[st];
minc = st;
}
st = stc + 1;
if (potential[st] < minp) {
minp = potential[st];
minc = st;
}
st = stcnx - 1;
if (potential[st] < minp) {
minp = potential[st];
minc = st;
}
st++;
if (potential[st] < minp) {
minp = potential[st];
minc = st;
}
st++;
if (potential[st] < minp) {
minp = potential[st];
minc = st;
}
stc = minc;
dx = 0;
dy = 0; //ROS_DEBUG("[Path] Pot: %0.1f pos: %0.1f,%0.1f",
// potential[stc], path[npath-1].first, path[npath-1].second); if (potential[stc] >= POT_HIGH) {
ROS_DEBUG("[PathCalc] No path found, high potential");
//savemap("navfn_highpot");
return 0;
}
} // have a good gradient here
else { // get grad at four positions near cell
gradCell(potential, stc);
gradCell(potential, stc + 1);
gradCell(potential, stcnx);
gradCell(potential, stcnx + 1); // get interpolated gradient
float x1 = (1.0 - dx) * gradx_[stc] + dx * gradx_[stc + 1];
float x2 = (1.0 - dx) * gradx_[stcnx] + dx * gradx_[stcnx + 1];
float x = (1.0 - dy) * x1 + dy * x2; // interpolated x
float y1 = (1.0 - dx) * grady_[stc] + dx * grady_[stc + 1];
float y2 = (1.0 - dx) * grady_[stcnx] + dx * grady_[stcnx + 1];
float y = (1.0 - dy) * y1 + dy * y2; // interpolated y // show gradients
ROS_DEBUG(
"[Path] %0.2f,%0.2f %0.2f,%0.2f %0.2f,%0.2f %0.2f,%0.2f; final x=%.3f, y=%.3f\n", gradx_[stc], grady_[stc], gradx_[stc+1], grady_[stc+1], gradx_[stcnx], grady_[stcnx], gradx_[stcnx+1], grady_[stcnx+1], x, y); // check for zero gradient, failed
if (x == 0.0 && y == 0.0) {
ROS_DEBUG("[PathCalc] Zero gradient");
return 0;
} // move in the right direction
float ss = pathStep_ / hypot(x, y);
dx += x * ss;
dy += y * ss; // check for overflow
if (dx > 1.0) {
stc++;
dx -= 1.0;
}
if (dx < -1.0) {
stc--;
dx += 1.0;
}
if (dy > 1.0) {
stc += xs_;
dy -= 1.0;
}
if (dy < -1.0) {
stc -= xs_;
dy += 1.0;
} } //printf("[Path] Pot: %0.1f grad: %0.1f,%0.1f pos: %0.1f,%0.1f\n",
// potential[stc], dx, dy, path[npath-1].first, path[npath-1].second);
} return false;
}

todo

sbpl,这也是个global planner,没有试过

move_base的全局路径规划代码研究的更多相关文章

  1. move_base的 局部路径规划代码研究

    base_local_planner teb_local_planner parameter code g2o base_local_planner ROS wiki Given a plan to ...

  2. ROS源码解读(二)--全局路径规划

    博客转载自:https://blog.csdn.net/xmy306538517/article/details/79032324 ROS中,机器人全局路径规划默认使用的是navfn包 ,move_b ...

  3. 全局路径规划算法Dijkstra(迪杰斯特拉算法)- matlab

    参考博客链接:https://www.cnblogs.com/kex1n/p/4178782.html Dijkstra是常用的全局路径规划算法,其本质上是一个最短路径寻优算法.算法的详细介绍参考上述 ...

  4. [python] A*算法基于栅格地图的全局路径规划

    # 所有节点的g值并没有初始化为无穷大 # 当两个子节点的f值一样时,程序选择最先搜索到的一个作为父节点加入closed # 对相同数值的不同对待,导致不同版本的A*算法找到等长的不同路径 # 最后c ...

  5. ROS探索总结(十四)——move_base(路径规划)

    在上一篇的博客中,我们一起学习了ROS定位于导航的总体框架,这一篇我们主要研究其中最重要的move_base包. 在总体框架图中可以看到,move_base提供了ROS导航的配置.运行.交互接口,它主 ...

  6. ROS机器人路径规划介绍--全局规划

    ROS机器人路径规划算法主要包括2个部分:1)全局路径规划算法:2)局部路径规划算法: 一.全局路径规划 global planner ROS 的navigation官方功能包提供了三种全局路径规划器 ...

  7. 【路径规划】 Optimal Trajectory Generation for Dynamic Street Scenarios in a Frenet Frame (附python代码实例)

    参考与前言 2010年,论文 Optimal Trajectory Generation for Dynamic Street Scenarios in a Frenet Frame 地址:https ...

  8. DWA局部路径规划算法论文阅读:The Dynamic Window Approach to Collision Avoidance。

    DWA(动态窗口)算法是用于局部路径规划的算法,已经在ROS中实现,在move_base堆栈中:http://wiki.ros.org/dwa_local_planner DWA算法第一次提出应该是1 ...

  9. ROS源码解读(一)--局部路径规划

    博客转载自:https://blog.csdn.net/xmy306538517/article/details/78772066 ROS局部路径导航包括Trajectory Rollout 和 Dy ...

随机推荐

  1. 【jpeg_Class 类】使用说明

    jpeg_Class类是针对图片操作类,可以获取图片属性.等比例缩略图片.裁切图片.图片上打印文字及打印水印等功能. 目录 原型 参数 返回 说明 Sub load(byVal path) path ...

  2. mac安装pkg 一直“正在验证” 卡着

    今天换了新mac, 但是之前wireshark(抓包工具) 不能用了 ,要安装Xquartz. 下载之后一直卡着, 网上找了半天没有解决方法. 最后我重启一下就好了... 重启一下. 2. 15款ma ...

  3. C/C++语言基础

    1. 一个子类中含有其他类对象,则构造函数的顺序是? 先执行基类的(如果基类当中有虚基类,要先执行虚基类的,其他基类则按照声明派生类是的顺序依次执行),在执行成员对象的,最后执行自己的. 2.spri ...

  4. 宿主机ssh免密登录docker容器

    一.检查系统内核 二.安装docker 1.yum install docker  -y 2.docker version                    #查看docker版本 3.syste ...

  5. Redis5.0:在这些场景使用,高效率还低成本!

    很多大型电商网站.视频直播和游戏应用等,存在大规模数据访问,对数据查询效率要求高,且数据结构简单,不涉及太多关联查询. 这种场景使用Redis,在速度上对传统磁盘数据库有很大优势,能够有效减少数据库磁 ...

  6. 【RL系列】马尔可夫决策过程——Jack‘s Car Rental

    本篇请结合课本Reinforcement Learning: An Introduction学习 Jack's Car Rental是一个经典的应用马尔可夫决策过程的问题,翻译过来,我们就直接叫它“租 ...

  7. 7. I/O复用

    一.I/O复用的特点 能同时监听多个文件描述符 自身是阻塞的 当多个文件描述符同时就绪时,如果不采取额外的措施,程序就只能按顺序依次处理其中的每一个文件描述符 由于其第三个特点,所以服务器程序看起来仍 ...

  8. 第五周PSP作业

    PSP表格: 累积进度条: 折线图: 饼状图:

  9. python apply()函数

    python apply函数的具体的含义: apply(func [, args [, kwargs ]]) 函数用于当函数参数已经存在于一个元组或字典中时,间接地调用函数.args是一个包含将要提供 ...

  10. bat获取当前日期的前一天

    批处理做这样的事情很麻烦,你可以用cscript来实现,比如把下面的内容保存为a.js文件:var d=new Date();d.setTime(d.getTime()-24*3600*1000);v ...