pengjiawei's profile - activity

Hi,

for (unsigned int i = 0; i < size_y; ++i)  
            {    
               for (unsigned int j = 0; j < size_x; ++j)

                {

                unsigned char value = new_map->data[index];
                costmap_[index] = interpretValue(value);
                  ++index;

                }
            }

the above is from ros navigation package , /navigation/costmap_2d/plugins/static_layer.cpp 198~206 it update the map according to laserScan.But I am confused why it update costmap from 0 to xy not from x to xy

costmap use the Index(x*nx+y) to represent the two dimensional coordinate,doesn't it?

I have this confusion becuase I see that in globalPlanner,the dijkstra.cpp use index() to getCost,and then it calculatePotentials

dijkstra.cpp 95~116

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);
}

when it push_cur,it calls the function getCost() in dijkstra.h.But I find that float c = costs[n] in Function getCost() n represents index(x*nx+y)

So,why we use x and y when we put value into char* costmap,But we use index() get value from costmap. where is the value from costmap[0] to costmap [nx],nx means the x size of map.

help me , Did I make a mistake

I wanna know costmap use x and y or index to represent the values in map

thx~ PS:why MarkDown Grammar ``` can't be used in this

/**

• @brief Transforms the global plan of the robot from the planner frame to the frame of the costmap,

• selects only the (first) part of the plan that is within the costmap area.

• @param tf A reference to a transform listener
• @param global_plan The plan to be transformed
• @param robot_pose The pose of the robot in the global frame (same as costmap)
• @param costmap A reference to the costmap being used so the window size for transforming can be computed
• @param global_frame The frame to transform the plan to
• @param transformed_plan Populated with the transformed plan */

bool transformGlobalPlan(const tf::TransformListener& tf,

const std::vector<geometry_msgs::PoseStamped>& global_plan,
  const tf::Stamped<tf::Pose>& global_robot_pose,
  const costmap_2d::Costmap2D& costmap,
  const std::string& global_frame,
  std::vector<geometry_msgs::PoseStamped>& transformed_plan);

what is the meaning of this function? I have got the global_plan from the globalPlanner in the global_frame,have I? Why and when should I call this function? and what is "the planner frame",Does it equal to the global_frame?

maybe you can debug tf first,such as viewing tf tree

oh ,Ok.and then what does the "essai1.launch" looks like in amcl package?

what does the odom publisher looks like?

There might be a problem.

what is the package "tf_hector_odom"

try catkin_make clean && catkin_make

does "a same topic M" means that the name and type of topic is same?

try this while(n.ok()){ subscriber; ros::spinOnce(); }

what does the global plan looks like?Is there only local plan in the video?

what does the global plan looks like?

map_server load a .yaml file

maybe you can publish the goals one by one?

did odom have no position?

it's a list,operator[]

what can you see if you do rostopic echo /base_scan and rostopic echo /map

i think i almost figure it out, there is a paper talk about this

what does your configuration of base_local_planner look like

how to complete coverage path? hello,as you see in the picture. I just wanna a optimal coverage algorithm.the robot st

how to complete coverage path? hello,as you see in the picture. I just wanna a optimal coverage algorithm.the robot st

Did you add the inflation layer ?

yes,try this dynamic_reconfigure tutorial and you can figure it out

have you ever configured the footprint in global_costmap

try this hector_navigation,you can focus on hector_exploration_planner

do you means rqt_reconfigure

amcl is in package navigation github,here is navigation wiki ,the amcl called adaptive monte calro localization will wor

I want to update static layer with the data from lidar.Is that OK?

My thought is that call this function StaticLayer::incomingMap(new_map),the parameter "new_map"is from the lidar?

Why it can be done? or why not? thanks

i have never try it in PI3 but I did it between my ubuntu16.04 and my android phone. There is a OccupancyGridLayer at li

android_apps/map_nav that might help

what if you set the goal near your turtlebot?

Did you set the global frame of rviz

by the way . If you don't wanna build all projects in directory <your workspace path>/src , you can add -DCATKIN_W

maybe you can check your topic first with command rostopic list and check whether your controller has received velocity

try this source /opt/ros/kinetic/setup.bash