路径规划（1） move_base总体概述

本文将对move_base进行一些说明。由于move_base的代码已经看过很长时间了，现在再看竟然细节大都记不清了。所以还是写一篇博客吧，为了以后的再次查看。

首先move_base可以实现机器人的实时的动态路径规划。它的路径规划分为了全局路径规划和局部路径规划。

全局路径规划主要依据：A*算法和Dijkstra算法。

局部路径规划主要依据TEB算法和dwa算法。

另外在路径规划过程中需要recovery机制来解决机器人卡主的问题，如果机器人卡住了就使用该机制消除障碍物，如果该机制运行完成，机器人仍然不能规划出合法的路径，那么机器人才会最终卡住。

地图使用的是gridmap形式的二维地图。

由于概述部分不涉及到算法所以直接上代码解析了。主要是怕自己忘记了而写的，所以解析的非常简略。

namespace move_base {

  MoveBase::MoveBase(tf::TransformListener& tf) :
    tf_(tf),
    as_(NULL),
    planner_costmap_ros_(NULL), controller_costmap_ros_(NULL),
    bgp_loader_("nav_core", "nav_core::BaseGlobalPlanner"),//初始化，全局规划
    blp_loader_("nav_core", "nav_core::BaseLocalPlanner"), //初始化，路径规划
    recovery_loader_("nav_core", "nav_core::RecoveryBehavior"),//初始化，recoverybehavior
    planner_plan_(NULL), latest_plan_(NULL), controller_plan_(NULL),//三次的路径规划，buffer
    runPlanner_(false), setup_(false), p_freq_change_(false), c_freq_change_(false), new_global_plan_(false) {

    as_ = new MoveBaseActionServer(ros::NodeHandle(), "move_base", boost::bind(&MoveBase::executeCb, this, _1), false);//actionserver，其相应的目标为：move_base goal

    ros::NodeHandle private_nh("~");
    ros::NodeHandle nh;

    recovery_trigger_ = PLANNING_R;

    //get some parameters that will be global to the move base node
    std::string global_planner, local_planner;
    private_nh.param("base_global_planner", global_planner, std::string("navfn/NavfnROS"));
    private_nh.param("base_local_planner", local_planner, std::string("base_local_planner/TrajectoryPlannerROS"));
    private_nh.param("global_costmap/robot_base_frame", robot_base_frame_, std::string("base_link"));
    private_nh.param("global_costmap/global_frame", global_frame_, std::string("/map"));
    private_nh.param("planner_frequency", planner_frequency_, 0.0);
    private_nh.param("controller_frequency", controller_frequency_, 20.0);
    private_nh.param("planner_patience", planner_patience_, 5.0);
    private_nh.param("controller_patience", controller_patience_, 15.0);

    private_nh.param("oscillation_timeout", oscillation_timeout_, 0.0);
    private_nh.param("oscillation_distance", oscillation_distance_, 0.5);

    //set up plan triple buffer
    planner_plan_ = new std::vector();
    latest_plan_ = new std::vector();
    controller_plan_ = new std::vector();

    //set up the planner's thread
    planner_thread_ = new boost::thread(boost::bind(&MoveBase::planThread, this));//开启路径规划线程，得到的是全局的路径规划。planThread等待executeCb通知他工作，规划一下路径，这个是global plan

    //for comanding the base
    vel_pub_ = nh.advertise("cmd_vel", 1);
    current_goal_pub_ = private_nh.advertise("current_goal", 0 );

    ros::NodeHandle action_nh("move_base");
    action_goal_pub_ = action_nh.advertise("goal", 1);

    //we'll provide a mechanism for some people to send goals as PoseStamped messages over a topic
    //they won't get any useful information back about its status, but this is useful for tools
    //like nav_view and rviz
    ros::NodeHandle simple_nh("move_base_simple");
    goal_sub_ = simple_nh.subscribe("goal", 1, boost::bind(&MoveBase::goalCB, this, _1));//从rviz得到目标

    //we'll assume the radius of the robot to be consistent with what's specified for the costmaps
    private_nh.param("local_costmap/inscribed_radius", inscribed_radius_, 0.325);
    private_nh.param("local_costmap/circumscribed_radius", circumscribed_radius_, 0.46);
    private_nh.param("clearing_radius", clearing_radius_, circumscribed_radius_);
    private_nh.param("conservative_reset_dist", conservative_reset_dist_, 3.0);

    private_nh.param("shutdown_costmaps", shutdown_costmaps_, false);
    private_nh.param("clearing_rotation_allowed", clearing_rotation_allowed_, true);
    private_nh.param("recovery_behavior_enabled", recovery_behavior_enabled_, true);

    //create the ros wrapper for the planner's costmap... and initializer a pointer we'll use with the underlying map
    planner_costmap_ros_ = new costmap_2d::Costmap2DROS("global_costmap", tf_);//全局路径规划所需要的costmap的初始化。
    planner_costmap_ros_->pause();

    //initialize the global planner
    try {
      planner_ = bgp_loader_.createInstance(global_planner);
      planner_->initialize(bgp_loader_.getName(global_planner), planner_costmap_ros_);
    } catch (const pluginlib::PluginlibException& ex) {
      ROS_FATAL("Failed to create the %s planner, are you sure it is properly registered and that the containing library is built? Exception: %s", global_planner.c_str(), ex.what());
      exit(1);
    }

    //create the ros wrapper for the controller's costmap... and initializer a pointer we'll use with the underlying map
    controller_costmap_ros_ = new costmap_2d::Costmap2DROS("local_costmap", tf_);//局部路径规划costmap封装的建立。
    controller_costmap_ros_->pause();

    //create a local planner
    try {
      tc_ = blp_loader_.createInstance(local_planner);
      ROS_INFO("Created local_planner %s", local_planner.c_str());
      tc_->initialize(blp_loader_.getName(local_planner), &tf_, controller_costmap_ros_);
    } catch (const pluginlib::PluginlibException& ex) {
      ROS_FATAL("Failed to create the %s planner, are you sure it is properly registered and that the containing library is built? Exception: %s", local_planner.c_str(), ex.what());
      exit(1);
    }

    // Start actively updating costmaps based on sensor data//更新costmap,各种层的添加
    planner_costmap_ros_->start();
    controller_costmap_ros_->start();

    //advertise a service for getting a plan
    make_plan_srv_ = private_nh.advertiseService("make_plan", &MoveBase::planService, this);

    //advertise a service for clearing the costmaps
    clear_costmaps_srv_ = private_nh.advertiseService("clear_costmaps", &MoveBase::clearCostmapsService, this);

    //if we shutdown our costmaps when we're deactivated... we'll do that now
    if(shutdown_costmaps_){
      ROS_DEBUG_NAMED("move_base","Stopping costmaps initially");
      planner_costmap_ros_->stop();
      controller_costmap_ros_->stop();
    }

    //load any user specified recovery behaviors, and if that fails load the defaults
    if(!loadRecoveryBehaviors(private_nh)){
      loadDefaultRecoveryBehaviors();
    }

    //initially, we'll need to make a plan
    state_ = PLANNING;

    //we'll start executing recovery behaviors at the beginning of our list
    recovery_index_ = 0;

    //we're all set up now so we can start the action server
    as_->start();

    dsrv_ = new dynamic_reconfigure::Server(ros::NodeHandle("~"));
    dynamic_reconfigure::Server::CallbackType cb = boost::bind(&MoveBase::reconfigureCB, this, _1, _2);
    dsrv_->setCallback(cb);//将通过.cfg文件生成的config的头文件通过reconfigureCB函数进行加载处理。对局部路径规划和全局路径规划进行了相应的初始化处理，确定了使用何种算法和参数。
  }

  void MoveBase::reconfigureCB(move_base::MoveBaseConfig &config, uint32_t level){
    boost::recursive_mutex::scoped_lock l(configuration_mutex_);

    //The first time we're called, we just want to make sure we have the
    //original configuration
    if(!setup_)
    {
      last_config_ = config;
      default_config_ = config;
      setup_ = true;
      return;
    }

    if(config.restore_defaults) {
      config = default_config_;
      //if someone sets restore defaults on the parameter server, prevent looping
      config.restore_defaults = false;
    }

    if(planner_frequency_ != config.planner_frequency)
    {
      planner_frequency_ = config.planner_frequency;
      p_freq_change_ = true;
    }

    if(controller_frequency_ != config.controller_frequency)
    {
      controller_frequency_ = config.controller_frequency;
      c_freq_change_ = true;
    }

    planner_patience_ = config.planner_patience;
    controller_patience_ = config.controller_patience;
    conservative_reset_dist_ = config.conservative_reset_dist;

    recovery_behavior_enabled_ = config.recovery_behavior_enabled;
    clearing_rotation_allowed_ = config.clearing_rotation_allowed;
    shutdown_costmaps_ = config.shutdown_costmaps;

    oscillation_timeout_ = config.oscillation_timeout;
    oscillation_distance_ = config.oscillation_distance;
    if(config.base_global_planner != last_config_.base_global_planner) {
      boost::shared_ptr old_planner = planner_;
      //initialize the global planner
      ROS_INFO("Loading global planner %s", config.base_global_planner.c_str());
      try {
        planner_ = bgp_loader_.createInstance(config.base_global_planner);

        // wait for the current planner to finish planning
        boost::unique_lock lock(planner_mutex_);

        // Clean up before initializing the new planner
        planner_plan_->clear();
        latest_plan_->clear();
        controller_plan_->clear();
        resetState();
        planner_->initialize(bgp_loader_.getName(config.base_global_planner), planner_costmap_ros_);

        lock.unlock();
      } catch (const pluginlib::PluginlibException& ex) {
        ROS_FATAL("Failed to create the %s planner, are you sure it is properly registered and that the \
                   containing library is built? Exception: %s", config.base_global_planner.c_str(), ex.what());
        planner_ = old_planner;
        config.base_global_planner = last_config_.base_global_planner;
      }
    }

    if(config.base_local_planner != last_config_.base_local_planner){
      boost::shared_ptr old_planner = tc_;
      //create a local planner
      try {
        tc_ = blp_loader_.createInstance(config.base_local_planner);
        // Clean up before initializing the new planner
        planner_plan_->clear();
        latest_plan_->clear();
        controller_plan_->clear();
        resetState();
        tc_->initialize(blp_loader_.getName(config.base_local_planner), &tf_, controller_costmap_ros_);
      } catch (const pluginlib::PluginlibException& ex) {
        ROS_FATAL("Failed to create the %s planner, are you sure it is properly registered and that the \
                   containing library is built? Exception: %s", config.base_local_planner.c_str(), ex.what());
        tc_ = old_planner;
        config.base_local_planner = last_config_.base_local_planner;
      }
    }

    last_config_ = config;
  }

  void MoveBase::goalCB(const geometry_msgs::PoseStamped::ConstPtr& goal){//从rviz等得到goal
    ROS_DEBUG_NAMED("move_base","In ROS goal callback, wrapping the PoseStamped in the action message and re-sending to the server.");
    move_base_msgs::MoveBaseActionGoal action_goal;
    action_goal.header.stamp = ros::Time::now();
    action_goal.goal.target_pose = *goal;

    action_goal_pub_.publish(action_goal);
  }

  void MoveBase::clearCostmapWindows(double size_x, double size_y){
    tf::Stamped global_pose;

    //clear the planner's costmap
    planner_costmap_ros_->getRobotPose(global_pose);

    std::vector clear_poly;
    double x = global_pose.getOrigin().x();
    double y = global_pose.getOrigin().y();
    geometry_msgs::Point pt;

    pt.x = x - size_x / 2;
    pt.y = y - size_y / 2;
    clear_poly.push_back(pt);

    pt.x = x + size_x / 2;
    pt.y = y - size_y / 2;
    clear_poly.push_back(pt);

    pt.x = x + size_x / 2;
    pt.y = y + size_y / 2;
    clear_poly.push_back(pt);

    pt.x = x - size_x / 2;
    pt.y = y + size_y / 2;
    clear_poly.push_back(pt);

    planner_costmap_ros_->getCostmap()->setConvexPolygonCost(clear_poly, costmap_2d::FREE_SPACE);

    //clear the controller's costmap
    controller_costmap_ros_->getRobotPose(global_pose);

    clear_poly.clear();
    x = global_pose.getOrigin().x();
    y = global_pose.getOrigin().y();

    pt.x = x - size_x / 2;
    pt.y = y - size_y / 2;
    clear_poly.push_back(pt);

    pt.x = x + size_x / 2;
    pt.y = y - size_y / 2;
    clear_poly.push_back(pt);

    pt.x = x + size_x / 2;
    pt.y = y + size_y / 2;
    clear_poly.push_back(pt);

    pt.x = x - size_x / 2;
    pt.y = y + size_y / 2;
    clear_poly.push_back(pt);

    controller_costmap_ros_->getCostmap()->setConvexPolygonCost(clear_poly, costmap_2d::FREE_SPACE);
  }

  bool MoveBase::clearCostmapsService(std_srvs::Empty::Request &req, std_srvs::Empty::Response &resp){//其实就是将costmap的各个层进行重置
    //clear the costmaps
    planner_costmap_ros_->resetLayers();
    controller_costmap_ros_->resetLayers();
    return true;
  }


  bool MoveBase::planService(nav_msgs::GetPlan::Request &req, nav_msgs::GetPlan::Response &resp){
    if(as_->isActive()){
      ROS_ERROR("move_base must be in an inactive state to make a plan for an external user");
      return false;
    }
    //make sure we have a costmap for our planner
    if(planner_costmap_ros_ == NULL){
      ROS_ERROR("move_base cannot make a plan for you because it doesn't have a costmap");
      return false;
    }
    tf::Stamped global_pose;
    if(!planner_costmap_ros_->getRobotPose(global_pose)){
      ROS_ERROR("move_base cannot make a plan for you because it could not get the start pose of the robot");
      return false;
    }
    geometry_msgs::PoseStamped start;
    //if the user does not specify a start pose, identified by an empty frame id, then use the robot's pose
    if(req.start.header.frame_id == "")
      tf::poseStampedTFToMsg(global_pose, start);
    else
      start = req.start;

    //update the copy of the costmap the planner uses
    clearCostmapWindows(2 * clearing_radius_, 2 * clearing_radius_);

    //first try to make a plan to the exact desired goal
    std::vector global_plan;
    if(!planner_->makePlan(start, req.goal, global_plan) || global_plan.empty()){
      ROS_DEBUG_NAMED("move_base","Failed to find a plan to exact goal of (%.2f, %.2f), searching for a feasible goal within tolerance", 
          req.goal.pose.position.x, req.goal.pose.position.y);

      //search outwards for a feasible goal within the specified tolerance
      geometry_msgs::PoseStamped p;
      p = req.goal;
      bool found_legal = false;
      float resolution = planner_costmap_ros_->getCostmap()->getResolution();
      float search_increment = resolution*3.0;
      if(req.tolerance > 0.0 && req.tolerance < search_increment) search_increment = req.tolerance;
      for(float max_offset = search_increment; max_offset <= req.tolerance && !found_legal; max_offset += search_increment) {
        for(float y_offset = 0; y_offset <= max_offset && !found_legal; y_offset += search_increment) {
          for(float x_offset = 0; x_offset <= max_offset && !found_legal; x_offset += search_increment) {
            //don't search again inside the current outer layer
            if(x_offset < max_offset-1e-9 && y_offset < max_offset-1e-9) continue;

            //search to both sides of the desired goal
            for(float y_mult = -1.0; y_mult <= 1.0 + 1e-9 && !found_legal; y_mult += 2.0) {

              //if one of the offsets is 0, -1*0 is still 0 (so get rid of one of the two)
              if(y_offset < 1e-9 && y_mult < -1.0 + 1e-9) continue;

              for(float x_mult = -1.0; x_mult <= 1.0 + 1e-9 && !found_legal; x_mult += 2.0) {
                if(x_offset < 1e-9 && x_mult < -1.0 + 1e-9) continue;

                p.pose.position.y = req.goal.pose.position.y + y_offset * y_mult;
                p.pose.position.x = req.goal.pose.position.x + x_offset * x_mult;

                if(planner_->makePlan(start, p, global_plan)){
                  if(!global_plan.empty()){

                    //adding the (unreachable) original goal to the end of the global plan, in case the local planner can get you there
                    //(the reachable goal should have been added by the global planner)
                    global_plan.push_back(req.goal);

                    found_legal = true;
                    ROS_DEBUG_NAMED("move_base", "Found a plan to point (%.2f, %.2f)", p.pose.position.x, p.pose.position.y);
                    break;
                  }
                }
                else{
                  ROS_DEBUG_NAMED("move_base","Failed to find a plan to point (%.2f, %.2f)", p.pose.position.x, p.pose.position.y);
                }
              }
            }
          }
        }
      }
    }

    //copy the plan into a message to send out
    resp.plan.poses.resize(global_plan.size());
    for(unsigned int i = 0; i < global_plan.size(); ++i){
      resp.plan.poses[i] = global_plan[i];
    }

    return true;
  }

  MoveBase::~MoveBase(){
    recovery_behaviors_.clear();

    delete dsrv_;

    if(as_ != NULL)
      delete as_;

    if(planner_costmap_ros_ != NULL)
      delete planner_costmap_ros_;

    if(controller_costmap_ros_ != NULL)
      delete controller_costmap_ros_;

    planner_thread_->interrupt();
    planner_thread_->join();

    delete planner_thread_;

    delete planner_plan_;
    delete latest_plan_;
    delete controller_plan_;

    planner_.reset();
    tc_.reset();
  }

  bool MoveBase::makePlan(const geometry_msgs::PoseStamped& goal, std::vector& plan){//依据global costmap进行全局路径规划
    boost::unique_lock lock(*(planner_costmap_ros_->getCostmap()->getMutex()));

    //make sure to set the plan to be empty initially
    plan.clear();

    //since this gets called on handle activate
    if(planner_costmap_ros_ == NULL) {//根据是否有costmap判断，是否可以进行全局路径规划
      ROS_ERROR("Planner costmap ROS is NULL, unable to create global plan");
      return false;
    }

    //get the starting pose of the robot
    tf::Stamped global_pose;
    if(!planner_costmap_ros_->getRobotPose(global_pose)) {//获取机器人的初始位姿
      ROS_WARN("Unable to get starting pose of robot, unable to create global plan");
      return false;
    }

    geometry_msgs::PoseStamped start;
    tf::poseStampedTFToMsg(global_pose, start);

    //if the planner fails or returns a zero length plan, planning failed
    if(!planner_->makePlan(start, goal, plan) || plan.empty()){//这里面的makeplan，是调用的全局路径规划的makeplan获取全局的路径。目前使用的是GlobalPlanner::makePlan
      ROS_DEBUG_NAMED("move_base","Failed to find a  plan to point (%.2f, %.2f)", goal.pose.position.x, goal.pose.position.y);
      return false;
    }

    return true;
  }

  void MoveBase::publishZeroVelocity(){
    geometry_msgs::Twist cmd_vel;
    cmd_vel.linear.x = 0.0;
    cmd_vel.linear.y = 0.0;
    cmd_vel.angular.z = 0.0;
    vel_pub_.publish(cmd_vel);
  }

  bool MoveBase::isQuaternionValid(const geometry_msgs::Quaternion& q){
    //first we need to check if the quaternion has nan's or infs
    if(!std::isfinite(q.x) || !std::isfinite(q.y) || !std::isfinite(q.z) || !std::isfinite(q.w)){
      ROS_ERROR("Quaternion has nans or infs... discarding as a navigation goal");
      return false;
    }

    tf::Quaternion tf_q(q.x, q.y, q.z, q.w);

    //next, we need to check if the length of the quaternion is close to zero
    if(tf_q.length2() < 1e-6){
      ROS_ERROR("Quaternion has length close to zero... discarding as navigation goal");
      return false;
    }

    //next, we'll normalize the quaternion and check that it transforms the vertical vector correctly
    tf_q.normalize();

    tf::Vector3 up(0, 0, 1);

    double dot = up.dot(up.rotate(tf_q.getAxis(), tf_q.getAngle()));

    if(fabs(dot - 1) > 1e-3){
      ROS_ERROR("Quaternion is invalid... for navigation the z-axis of the quaternion must be close to vertical.");
      return false;
    }

    return true;
  }

  geometry_msgs::PoseStamped MoveBase::goalToGlobalFrame(const geometry_msgs::PoseStamped& goal_pose_msg){
    std::string global_frame = planner_costmap_ros_->getGlobalFrameID();
    tf::Stamped goal_pose, global_pose;
    poseStampedMsgToTF(goal_pose_msg, goal_pose);

    //just get the latest available transform... for accuracy they should send
    //goals in the frame of the planner
    goal_pose.stamp_ = ros::Time();

    try{
      tf_.transformPose(global_frame, goal_pose, global_pose);
    }
    catch(tf::TransformException& ex){
      ROS_WARN("Failed to transform the goal pose from %s into the %s frame: %s",
          goal_pose.frame_id_.c_str(), global_frame.c_str(), ex.what());
      return goal_pose_msg;
    }

    geometry_msgs::PoseStamped global_pose_msg;
    tf::poseStampedTFToMsg(global_pose, global_pose_msg);
    return global_pose_msg;
  }

  void MoveBase::wakePlanner(const ros::TimerEvent& event)
  {
    // we have slept long enough for rate
    planner_cond_.notify_one();
  }

  void MoveBase::planThread(){//此处注意该段程序与executecb的两个线程之间的关系，防止出现程序问题，出出现全局路径与最终目标的同时操作等问题。
    ROS_DEBUG_NAMED("move_base_plan_thread","Starting planner thread...");
    ros::NodeHandle n;
    ros::Timer timer;
    bool wait_for_wake = false;
    boost::unique_lock lock(planner_mutex_);
    while(n.ok()){
      //check if we should run the planner (the mutex is locked)
      while(wait_for_wake || !runPlanner_){//如果是需要等待唤醒状态，或者目前已经被锁住，则进行等待，直到可以，等待上面的executeCb函数使得runPlanner_ = true
        //if we should not be running the planner then suspend this thread
        ROS_DEBUG_NAMED("move_base_plan_thread","Planner thread is suspending");
        planner_cond_.wait(lock);
        wait_for_wake = false;
      }
      ros::Time start_time = ros::Time::now();

      //time to plan! get a copy of the goal and unlock the mutex
      geometry_msgs::PoseStamped temp_goal = planner_goal_;
      lock.unlock();
      ROS_DEBUG_NAMED("move_base_plan_thread","Planning...");

      //run planner
      planner_plan_->clear();//旧的路径规划，清除
      bool gotPlan = n.ok() && makePlan(temp_goal, *planner_plan_);//此处调用机器人的全局路径规划===========>调用全局规划

      if(gotPlan){
        ROS_DEBUG_NAMED("move_base_plan_thread","Got Plan with %zu points!", planner_plan_->size());
        //pointer swap the plans under mutex (the controller will pull from latest_plan_)
        std::vector* temp_plan = planner_plan_;

        lock.lock();
        planner_plan_ = latest_plan_;
        latest_plan_ = temp_plan;
        last_valid_plan_ = ros::Time::now();
        new_global_plan_ = true;//新旧plan的赋值

        ROS_DEBUG_NAMED("move_base_plan_thread","Generated a plan from the base_global_planner");

        //make sure we only start the controller if we still haven't reached the goal
        if(runPlanner_)
          state_ = CONTROLLING;
        if(planner_frequency_ <= 0)
          runPlanner_ = false;
        lock.unlock();
      }
      //if we didn't get a plan and we are in the planning state (the robot isn't moving)
      else if(state_==PLANNING){
        ROS_DEBUG_NAMED("move_base_plan_thread","No Plan...");
        ros::Time attempt_end = last_valid_plan_ + ros::Duration(planner_patience_);

        //check if we've tried to make a plan for over our time limit
        lock.lock();
        if(ros::Time::now() > attempt_end && runPlanner_){//如果已经路径规划等待超时并且机器人未在运动，那么就更新障碍物层，机器人的速度设置为0，等待重新的规划。
          //we'll move into our obstacle clearing mode
          state_ = CLEARING;
          publishZeroVelocity();
          recovery_trigger_ = PLANNING_R;
        }
        lock.unlock();
      }

      //take the mutex for the next iteration
      lock.lock();

      //setup sleep interface if needed
      if(planner_frequency_ > 0){
        ros::Duration sleep_time = (start_time + ros::Duration(1.0/planner_frequency_)) - ros::Time::now();
        if (sleep_time > ros::Duration(0.0)){
          wait_for_wake = true;
          timer = n.createTimer(sleep_time, &MoveBase::wakePlanner, this);//定时器，多久没有规划路径，就通知一次路径规划。
        }
      }
    }
  }

  void MoveBase::executeCb(const move_base_msgs::MoveBaseGoalConstPtr& move_base_goal)//action等待goal
  {
    if(!isQuaternionValid(move_base_goal->target_pose.pose.orientation)){//对目标机器人位姿的四元数进行可行性判断
      as_->setAborted(move_base_msgs::MoveBaseResult(), "Aborting on goal because it was sent with an invalid quaternion");
      return;
    }

    geometry_msgs::PoseStamped goal = goalToGlobalFrame(move_base_goal->target_pose);

    //we have a goal so start the planner
    boost::unique_lock lock(planner_mutex_);
    planner_goal_ = goal;
    runPlanner_ = true;
    planner_cond_.notify_one();//通知路径规划线程
    lock.unlock();

    current_goal_pub_.publish(goal);
    std::vector global_plan;

    ros::Rate r(controller_frequency_);
    if(shutdown_costmaps_){
      ROS_DEBUG_NAMED("move_base","Starting up costmaps that were shut down previously");
      planner_costmap_ros_->start();
      controller_costmap_ros_->start();//全局路径规划和局部路径规划所需要的costmap的准备工作，各个层完成添加。
    }

    //we want to make sure that we reset the last time we had a valid plan and control
    last_valid_control_ = ros::Time::now();
    last_valid_plan_ = ros::Time::now();
    last_oscillation_reset_ = ros::Time::now();

    ros::NodeHandle n;
    while(n.ok())
    {
      if(c_freq_change_)
      {
        ROS_INFO("Setting controller frequency to %.2f", controller_frequency_);
        r = ros::Rate(controller_frequency_);
        c_freq_change_ = false;
      }
//如果目标，goal取消，判断出有新的目标那么重新开始一次操作步骤
      if(as_->isPreemptRequested()){
        if(as_->isNewGoalAvailable()){
          //if we're active and a new goal is available, we'll accept it, but we won't shut anything down
          move_base_msgs::MoveBaseGoal new_goal = *as_->acceptNewGoal();

          if(!isQuaternionValid(new_goal.target_pose.pose.orientation)){
            as_->setAborted(move_base_msgs::MoveBaseResult(), "Aborting on goal because it was sent with an invalid quaternion");
            return;
          }

          goal = goalToGlobalFrame(new_goal.target_pose);

          //we'll make sure that we reset our state for the next execution cycle
          recovery_index_ = 0;
          state_ = PLANNING;

          //we have a new goal so make sure the planner is awake
          lock.lock();
          planner_goal_ = goal;
          runPlanner_ = true;
          planner_cond_.notify_one();//在线程被阻塞时，该函数会自动调用 lck.unlock() 释放锁，使得其他被阻塞在锁竞争上的线程得以继续执行。
          lock.unlock();

          //publish the goal point to the visualizer
          ROS_DEBUG_NAMED("move_base","move_base has received a goal of x: %.2f, y: %.2f", goal.pose.position.x, goal.pose.position.y);
          current_goal_pub_.publish(goal);

          //make sure to reset our timeouts
          last_valid_control_ = ros::Time::now();
          last_valid_plan_ = ros::Time::now();
          last_oscillation_reset_ = ros::Time::now();
        }
        else {
          //if we've been preempted explicitly we need to shut things down
          resetState();

          //notify the ActionServer that we've successfully preempted
          ROS_DEBUG_NAMED("move_base","Move base preempting the current goal");
          as_->setPreempted();

          //we'll actually return from execute after preempting
          return;
        }
      }

      //we also want to check if we've changed global frames because we need to transform our goal pose
      if(goal.header.frame_id != planner_costmap_ros_->getGlobalFrameID()){//将goal的坐标系转化为全局坐标系下
        goal = goalToGlobalFrame(goal);

        //we want to go back to the planning state for the next execution cycle
        recovery_index_ = 0;
        state_ = PLANNING;

        //we have a new goal so make sure the planner is awake
        lock.lock();
        planner_goal_ = goal;
        runPlanner_ = true;
        planner_cond_.notify_one();
        lock.unlock();

        //publish the goal point to the visualizer
        ROS_DEBUG_NAMED("move_base","The global frame for move_base has changed, new frame: %s, new goal position x: %.2f, y: %.2f", goal.header.frame_id.c_str(), goal.pose.position.x, goal.pose.position.y);
        current_goal_pub_.publish(goal);

        //make sure to reset our timeouts
        last_valid_control_ = ros::Time::now();
        last_valid_plan_ = ros::Time::now();
        last_oscillation_reset_ = ros::Time::now();
      }

      //for timing that gives real time even in simulation
      ros::WallTime start = ros::WallTime::now();

      //the real work on pursuing a goal is done here
      bool done = executeCycle(goal, global_plan);//这个函数是执行机器人到达目标的函数的调用=============》核心函数调用

      //if we're done, then we'll return from execute
      if(done)
        return;

      //check if execution of the goal has completed in some way

      ros::WallDuration t_diff = ros::WallTime::now() - start;
      ROS_DEBUG_NAMED("move_base","Full control cycle time: %.9f\n", t_diff.toSec());

      r.sleep();
      //make sure to sleep for the remainder of our cycle time，根据规划是否超出时间判断是否要打印警告信息。
      if(r.cycleTime() > ros::Duration(1 / controller_frequency_) && state_ == CONTROLLING)
        ROS_WARN("Control loop missed its desired rate of %.4fHz... the loop actually took %.4f seconds", controller_frequency_, r.cycleTime().toSec());
    }

    //wake up the planner thread so that it can exit cleanly
    lock.lock();
    runPlanner_ = true;
    planner_cond_.notify_one();
    lock.unlock();

    //if the node is killed then we'll abort and return
    as_->setAborted(move_base_msgs::MoveBaseResult(), "Aborting on the goal because the node has been killed");
    return;
  }

  double MoveBase::distance(const geometry_msgs::PoseStamped& p1, const geometry_msgs::PoseStamped& p2)
  {
    return hypot(p1.pose.position.x - p2.pose.position.x, p1.pose.position.y - p2.pose.position.y);
  }

  bool MoveBase::executeCycle(geometry_msgs::PoseStamped& goal, std::vector& global_plan){//进行局部路径规划，每次有goal发布的时候，都会去调用executeCb,executeCb会去调用executeCycle进行local plan,
    boost::recursive_mutex::scoped_lock ecl(configuration_mutex_);
    //we need to be able to publish velocity commands
    geometry_msgs::Twist cmd_vel;//需要发布的那个速度

    //update feedback to correspond to our curent position
    tf::Stamped global_pose;
    planner_costmap_ros_->getRobotPose(global_pose);
    geometry_msgs::PoseStamped current_position;
    tf::poseStampedTFToMsg(global_pose, current_position);

    //push the feedback out
    move_base_msgs::MoveBaseFeedback feedback;
    feedback.base_position = current_position;
    as_->publishFeedback(feedback);
    ros::NodeHandle N;
    mb_sound_ = N.advertise("/move_base_file",2);

    //check to see if we've moved far enough to reset our oscillation timeout
    if(distance(current_position, oscillation_pose_) >= oscillation_distance_)
    {
      last_oscillation_reset_ = ros::Time::now();
      oscillation_pose_ = current_position;

      //if our last recovery was caused by oscillation, we want to reset the recovery index 
      if(recovery_trigger_ == OSCILLATION_R)
        recovery_index_ = 0;
    }

    //check that the observation buffers for the costmap are current, we don't want to drive blind
    if(!controller_costmap_ros_->isCurrent()){
      ROS_WARN("[%s]:Sensor data is out of date, we're not going to allow commanding of the base for safety",ros::this_node::getName().c_str());
      publishZeroVelocity();
      return false;
    }

    //if we have a new plan then grab it and give it to the controller
    if(new_global_plan_){
      //make sure to set the new plan flag to false
      new_global_plan_ = false;

      ROS_DEBUG_NAMED("move_base","Got a new plan...swap pointers");

      //do a pointer swap under mutex
      std::vector* temp_plan = controller_plan_;

      boost::unique_lock lock(planner_mutex_);
      controller_plan_ = latest_plan_;
      latest_plan_ = temp_plan;
      lock.unlock();
      ROS_DEBUG_NAMED("move_base","pointers swapped!");
//下面调用局部路径规划
      if(!tc_->setPlan(*controller_plan_)){//将全局路径规划的到的传递给局部路径规划，完成了局部的初始化等各项预备工作，controller_plan_数据还是全局路径
        //ABORT and SHUTDOWN COSTMAPS
        ROS_ERROR("Failed to pass global plan to the controller, aborting.");
        resetState();

        //disable the planner thread
        lock.lock();
        runPlanner_ = false;
        lock.unlock();

        as_->setAborted(move_base_msgs::MoveBaseResult(), "Failed to pass global plan to the controller.");
        return true;
      }

      //make sure to reset recovery_index_ since we were able to find a valid plan
      if(recovery_trigger_ == PLANNING_R)
        recovery_index_ = 0;
    }

    //the move_base state machine, handles the control logic for navigation
    switch(state_){//move_base的状态机，
      //if we are in a planning state, then we'll attempt to make a plan
      case PLANNING:
        {
          boost::mutex::scoped_lock lock(planner_mutex_);
          runPlanner_ = true;
          planner_cond_.notify_one();
        }
        ROS_DEBUG_NAMED("move_base","Waiting for plan, in the planning state.");
        break;

      //if we're controlling, we'll attempt to find valid velocity commands
      case CONTROLLING:
        ROS_DEBUG_NAMED("move_base","In controlling state.");

        //check to see if we've reached our goal
        if(tc_->isGoalReached()){//actionserver判断goal是否已经完成
          ROS_DEBUG_NAMED("move_base","Goal reached!");
          resetState();

          //disable the planner thread
          boost::unique_lock lock(planner_mutex_);
          runPlanner_ = false;
          lock.unlock();

          as_->setSucceeded(move_base_msgs::MoveBaseResult(), "Goal reached.");
          return true;
        }

        //check for an oscillation condition
        if(oscillation_timeout_ > 0.0 &&
            last_oscillation_reset_ + ros::Duration(oscillation_timeout_) < ros::Time::now())//如果机器人的震荡时间已经超过了额定的设定，机器人速度置零。确定震荡，进行机器人不动，障碍物清除模式+recovery机制将原来的障碍物进行清除。
        {
          publishZeroVelocity();
          state_ = CLEARING;
          recovery_trigger_ = OSCILLATION_R;
        }
        
        {
         boost::unique_lock lock(*(controller_costmap_ros_->getCostmap()->getMutex()));
        
        if(tc_->computeVelocityCommands(cmd_vel)){//此处发布局部路径规划产生的速度
          ROS_DEBUG_NAMED( "move_base", "Got a valid command from the local planner: %.3lf, %.3lf, %.3lf",
                           cmd_vel.linear.x, cmd_vel.linear.y, cmd_vel.angular.z );
          last_valid_control_ = ros::Time::now();
          //make sure that we send the velocity command to the base
          vel_pub_.publish(cmd_vel);//开始publish速度
          if(recovery_trigger_ == CONTROLLING_R)
            recovery_index_ = 0;
        }
        else {
          ROS_DEBUG_NAMED("move_base", "The local planner could not find a valid plan.");
          ros::Time attempt_end = last_valid_control_ + ros::Duration(controller_patience_);

          //check if we've tried to find a valid control for longer than our time limit
          if(ros::Time::now() > attempt_end){//recovery后，机器人也旋转了，但是机器人还是不能继续沿着规划出的路径前进。
            //we'll move into our obstacle clearing mode
            publishZeroVelocity();//下面的散步操作和上面（//check for an oscillation condition）一致
            state_ = CLEARING;
            recovery_trigger_ = CONTROLLING_R;
          }
          else{//如果时间还未产生合法速度，尝试重新进行局部规划
            //otherwise, if we can't find a valid control, we'll go back to planning
            last_valid_plan_ = ros::Time::now();
            state_ = PLANNING;
            publishZeroVelocity();

            //enable the planner thread in case it isn't running on a clock
            boost::unique_lock lock(planner_mutex_);
            runPlanner_ = true;
            planner_cond_.notify_one();
            lock.unlock();
          }
        }
        }

        break;

      //we'll try to clear out space with any user-provided recovery behaviors
      case CLEARING:
        ROS_DEBUG_NAMED("move_base","In clearing/recovery state");
        //we'll invoke whatever recovery behavior we're currently on if they're enabled
        if(recovery_behavior_enabled_ && recovery_index_ < recovery_behaviors_.size()){
          ROS_DEBUG_NAMED("move_base_recovery","Executing behavior %u of %zu", recovery_index_, recovery_behaviors_.size());
          recovery_behaviors_[recovery_index_]->runBehavior();

          //we at least want to give the robot some time to stop oscillating after executing the behavior
          last_oscillation_reset_ = ros::Time::now();

          //we'll check if the recovery behavior actually worked
          ROS_DEBUG_NAMED("move_base_recovery","Going back to planning state");
          state_ = PLANNING;

          //update the index of the next recovery behavior that we'll try
          recovery_index_++;
        }
        else{//已经recovery过了，但是还没有效果
          ROS_DEBUG_NAMED("move_base_recovery","All recovery behaviors have failed, locking the planner and disabling it.");

          mb_msg.data = "recovery_failed.wav";
          mb_sound_.publish(mb_msg);

          //disable the planner thread
          boost::unique_lock lock(planner_mutex_);
          runPlanner_ = false;
          lock.unlock();

          ROS_DEBUG_NAMED("move_base_recovery","Something should abort after this.");

          if(recovery_trigger_ == CONTROLLING_R){
            ROS_ERROR("Aborting because a valid control could not be found. Even after executing all recovery behaviors");
            as_->setAborted(move_base_msgs::MoveBaseResult(), "Failed to find a valid control. Even after executing recovery behaviors.");
          }
          else if(recovery_trigger_ == PLANNING_R){
            ROS_ERROR("Aborting because a valid plan could not be found. Even after executing all recovery behaviors");
            as_->setAborted(move_base_msgs::MoveBaseResult(), "Failed to find a valid plan. Even after executing recovery behaviors.");
          }
          else if(recovery_trigger_ == OSCILLATION_R){
            ROS_ERROR("Aborting because the robot appears to be oscillating over and over. Even after executing all recovery behaviors");
            as_->setAborted(move_base_msgs::MoveBaseResult(), "Robot is oscillating. Even after executing recovery behaviors.");
          }
          resetState();
          return true;
        }
        break;
      default:
        ROS_ERROR("This case should never be reached, something is wrong, aborting");
        resetState();
        //disable the planner thread
        boost::unique_lock lock(planner_mutex_);
        runPlanner_ = false;
        lock.unlock();
        as_->setAborted(move_base_msgs::MoveBaseResult(), "Reached a case that should not be hit in move_base. This is a bug, please report it.");
        return true;
    }

    //we aren't done yet
    return false;
  }

  bool MoveBase::loadRecoveryBehaviors(ros::NodeHandle node){
    XmlRpc::XmlRpcValue behavior_list;
    if(node.getParam("recovery_behaviors", behavior_list)){
      if(behavior_list.getType() == XmlRpc::XmlRpcValue::TypeArray){
        for(int i = 0; i < behavior_list.size(); ++i){
          if(behavior_list[i].getType() == XmlRpc::XmlRpcValue::TypeStruct){
            if(behavior_list[i].hasMember("name") && behavior_list[i].hasMember("type")){
              //check for recovery behaviors with the same name
              for(int j = i + 1; j < behavior_list.size(); j++){
                if(behavior_list[j].getType() == XmlRpc::XmlRpcValue::TypeStruct){
                  if(behavior_list[j].hasMember("name") && behavior_list[j].hasMember("type")){
                    std::string name_i = behavior_list[i]["name"];
                    std::string name_j = behavior_list[j]["name"];
                    if(name_i == name_j){
                      ROS_ERROR("A recovery behavior with the name %s already exists, this is not allowed. Using the default recovery behaviors instead.", 
                          name_i.c_str());
                      return false;
                    }
                  }
                }
              }
            }
            else{
              ROS_ERROR("Recovery behaviors must have a name and a type and this does not. Using the default recovery behaviors instead.");
              return false;
            }
          }
          else{
            ROS_ERROR("Recovery behaviors must be specified as maps, but they are XmlRpcType %d. We'll use the default recovery behaviors instead.",
                behavior_list[i].getType());
            return false;
          }
        }

        //if we've made it to this point, we know that the list is legal so we'll create all the recovery behaviors
        for(int i = 0; i < behavior_list.size(); ++i){
          try{
            //check if a non fully qualified name has potentially been passed in
            if(!recovery_loader_.isClassAvailable(behavior_list[i]["type"])){
              std::vector classes = recovery_loader_.getDeclaredClasses();
              for(unsigned int i = 0; i < classes.size(); ++i){
                if(behavior_list[i]["type"] == recovery_loader_.getName(classes[i])){
                  //if we've found a match... we'll get the fully qualified name and break out of the loop
                  ROS_WARN("Recovery behavior specifications should now include the package name. You are using a deprecated API. Please switch from %s to %s in your yaml file.",
                      std::string(behavior_list[i]["type"]).c_str(), classes[i].c_str());
                  behavior_list[i]["type"] = classes[i];
                  break;
                }
              }
            }

            boost::shared_ptr behavior(recovery_loader_.createInstance(behavior_list[i]["type"]));

            //shouldn't be possible, but it won't hurt to check
            if(behavior.get() == NULL){
              ROS_ERROR("The ClassLoader returned a null pointer without throwing an exception. This should not happen");
              return false;
            }

            //initialize the recovery behavior with its name
            behavior->initialize(behavior_list[i]["name"], &tf_, planner_costmap_ros_, controller_costmap_ros_);
            recovery_behaviors_.push_back(behavior);
          }
          catch(pluginlib::PluginlibException& ex){
            ROS_ERROR("Failed to load a plugin. Using default recovery behaviors. Error: %s", ex.what());
            return false;
          }
        }
      }
      else{
        ROS_ERROR("The recovery behavior specification must be a list, but is of XmlRpcType %d. We'll use the default recovery behaviors instead.", 
            behavior_list.getType());
        return false;
      }
    }
    else{
      //if no recovery_behaviors are specified, we'll just load the defaults
      return false;
    }

    //if we've made it here... we've constructed a recovery behavior list successfully
    return true;
  }

  //we'll load our default recovery behaviors here
  void MoveBase::loadDefaultRecoveryBehaviors(){
    recovery_behaviors_.clear();
    try{
      //we need to set some parameters based on what's been passed in to us to maintain backwards compatibility
      ros::NodeHandle n("~");
      n.setParam("conservative_reset/reset_distance", conservative_reset_dist_);
      n.setParam("aggressive_reset/reset_distance", circumscribed_radius_ * 4);

      //first, we'll load a recovery behavior to clear the costmap                      //各种清理costmap的插件的产生
      boost::shared_ptr cons_clear(recovery_loader_.createInstance("clear_costmap_recovery/ClearCostmapRecovery"));
      cons_clear->initialize("conservative_reset", &tf_, planner_costmap_ros_, controller_costmap_ros_);
      recovery_behaviors_.push_back(cons_clear);

      //next, we'll load a recovery behavior to rotate in place
      boost::shared_ptr rotate(recovery_loader_.createInstance("rotate_recovery/RotateRecovery"));
      if(clearing_rotation_allowed_){
        rotate->initialize("rotate_recovery", &tf_, planner_costmap_ros_, controller_costmap_ros_);
        recovery_behaviors_.push_back(rotate);
      }

      //next, we'll load a recovery behavior that will do an aggressive reset of the costmap
      boost::shared_ptr ags_clear(recovery_loader_.createInstance("clear_costmap_recovery/ClearCostmapRecovery"));
      ags_clear->initialize("aggressive_reset", &tf_, planner_costmap_ros_, controller_costmap_ros_);
      recovery_behaviors_.push_back(ags_clear);

      //we'll rotate in-place one more time
      if(clearing_rotation_allowed_)
        recovery_behaviors_.push_back(rotate);
    }
    catch(pluginlib::PluginlibException& ex){
      ROS_FATAL("Failed to load a plugin. This should not happen on default recovery behaviors. Error: %s", ex.what());
    }

    return;
  }

  void MoveBase::resetState(){
    // Disable the planner thread
    boost::unique_lock lock(planner_mutex_);
    runPlanner_ = false;
    lock.unlock();

    // Reset statemachine
    state_ = PLANNING;
    recovery_index_ = 0;
    recovery_trigger_ = PLANNING_R;
    publishZeroVelocity();

    //if we shutdown our costmaps when we're deactivated... we'll do that now
    if(shutdown_costmaps_){
      ROS_DEBUG_NAMED("move_base","Stopping costmaps");
      planner_costmap_ros_->stop();
      controller_costmap_ros_->stop();
    }
  }
};

 

                                                                                                                                     wxw 2018.05.09 SH