【Cartographer参数详解，调参，降低计算量调优笔记记录】

Cartographer学习记录

---

---

前言

最近同学都在卷论文，刷力扣，自己的导师比较松也没有一个很好的学习氛围，为了敦促自己决定开始做一些以前不会做的事，养成一些新的习惯。计划在这学期完成cartographer的学习，并在CSDN做一些记录，虽是他山之石，但也是迈出的第一步，共勉。

---

提示：以下是本篇文章正文内容，下面案例可供参考

一、cartographer参数详解

说明：分析所用的一些launch和lua文件是跟着学习的讲师所配置的，由官方下载的雷达2d.lua文件基础上配置成的，配合相应的雷达包可以直接运行，具有实际应用的参考价值。运行部分会另开一篇文章进行介绍，此篇主要是参数介绍。rviz中建图过程如下所示：
最终结果如下图所示：

二、.lua文件参数详解

1.rs16_2d_outdoor.lua

如文件名所示，该配置文件进行的是rs16多线激光雷达录制的室外bag使用时的参数配置，建立的地图为2d形式。重要的参数提出来重写，原来所在的位置：如TRAJECTORY_BUILDER_2D.submaps.num_range_data = 80，即在trajectory_builder_2d.lua的submaps函数中找num_range_data，相应的注释也在子lua文件中

-- Copyright 2016 The Cartographer Authors
--
-- Licensed under the Apache License, Version 2.0 (the "License");
-- you may not use this file except in compliance with the License.
-- You may obtain a copy of the License at
--
--      http://www.apache.org/licenses/LICENSE-2.0
--
-- Unless required by applicable law or agreed to in writing, software
-- distributed under the License is distributed on an "AS IS" BASIS,
-- WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-- See the License for the specific language governing permissions and
-- limitations under the License.

include "map_builder.lua"                   --包含的cartographer里的lua文件
include "trajectory_builder.lua"

options = {
  map_builder = MAP_BUILDER,                -- map_builder.lua的配置信息
  trajectory_builder = TRAJECTORY_BUILDER,  -- trajectory_builder.lua的配置信息
  
  map_frame = "map",                        -- 地图坐标系的名字
  tracking_frame = "imu_link",              -- 将所有传感器数据转换到这个坐标系下，如果有imu的，就写imu的link，如果没有，就写base_link或者footprint，
                                            --因为cartographer会将所有传感器进行坐标变换到tracking_fram坐标系下，每个传感器的频率不一样，imu频率远高于雷达的频率，这样做可以减少计算量
  published_frame = "footprint",            -- tf: map -> footprint 自己bag的tf最上面的坐标系的名字
  odom_frame = "odom",                      -- 里程计的坐标系名字
  provide_odom_frame = false,               -- 是否提供odom的tf, 如果为true,则tf树为map->odom->footprint
                                            -- 如果为false tf树为map->footprint
  publish_frame_projected_to_2d = false,    -- 是否将坐标系投影到平面上，没啥用
  --use_pose_extrapolator = false,            -- 发布tf时是使用pose_extrapolator的位姿还是前端计算出来的位姿，前端计算的位姿更准

  use_odometry = false,                     -- 是否使用里程计,如果使用要求一定要有odom的tf *
  use_nav_sat = false,                      -- 是否使用gps *topic形式订阅，不可订阅多个里程计/gps/landmark，要注意做重映射
  use_landmarks = false,                    -- 是否使用landmark *
  num_laser_scans = 0,                      -- 是否使用单线激光数据，可订阅多个
  num_multi_echo_laser_scans = 0,           -- 是否使用multi_echo_laser_scans数据
  num_subdivisions_per_laser_scan = 1,      -- 1帧数据被分成几次处理,一般为1
  num_point_clouds = 1,                     -- 是否使用点云数据
  
  lookup_transform_timeout_sec = 0.2,       -- 查找tf时的超时时间
  submap_publish_period_sec = 0.3,          -- 发布数据的时间间隔
  pose_publish_period_sec = 5e-3,
  trajectory_publish_period_sec = 30e-3,

  rangefinder_sampling_ratio = 1.,          -- 传感器数据的采样频率
  odometry_sampling_ratio = 1.,             --设成0.1即来10帧用1帧
  fixed_frame_pose_sampling_ratio = 1.,     --某个传感器不准，可以降低其使用频率
  imu_sampling_ratio = 1.,                  --如若不准，一般直接弃用即可
  landmarks_sampling_ratio = 1.,
}

MAP_BUILDER.use_trajectory_builder_2d = true   --之前进行include已经包含在该文件中了，这里进行重写，列出的都为比较重要的参数
--在这里进行重写就会覆盖子文件中的参数

TRAJECTORY_BUILDER_2D.use_imu_data = true
TRAJECTORY_BUILDER_2D.min_range = 0.3
TRAJECTORY_BUILDER_2D.max_range = 100.
TRAJECTORY_BUILDER_2D.min_z = 0.2
--TRAJECTORY_BUILDER_2D.max_z = 1.4
--TRAJECTORY_BUILDER_2D.voxel_filter_size = 0.02

--TRAJECTORY_BUILDER_2D.adaptive_voxel_filter.max_length = 0.5
--TRAJECTORY_BUILDER_2D.adaptive_voxel_filter.min_num_points = 200.
--TRAJECTORY_BUILDER_2D.adaptive_voxel_filter.max_range = 50.

--TRAJECTORY_BUILDER_2D.loop_closure_adaptive_voxel_filter.max_length = 0.9
--TRAJECTORY_BUILDER_2D.loop_closure_adaptive_voxel_filter.min_num_points = 100
--TRAJECTORY_BUILDER_2D.loop_closure_adaptive_voxel_filter.max_range = 50.

TRAJECTORY_BUILDER_2D.use_online_correlative_scan_matching = false
TRAJECTORY_BUILDER_2D.ceres_scan_matcher.occupied_space_weight = 1.
TRAJECTORY_BUILDER_2D.ceres_scan_matcher.translation_weight = 1.
TRAJECTORY_BUILDER_2D.ceres_scan_matcher.rotation_weight = 1.
--TRAJECTORY_BUILDER_2D.ceres_scan_matcher.ceres_solver_options.max_num_iterations = 12

--TRAJECTORY_BUILDER_2D.motion_filter.max_distance_meters = 0.1
--TRAJECTORY_BUILDER_2D.motion_filter.max_angle_radians = 0.004
--TRAJECTORY_BUILDER_2D.imu_gravity_time_constant = 1.

TRAJECTORY_BUILDER_2D.submaps.num_range_data = 80.
TRAJECTORY_BUILDER_2D.submaps.grid_options_2d.resolution = 0.1

POSE_GRAPH.optimize_every_n_nodes = 160.
POSE_GRAPH.constraint_builder.sampling_ratio = 0.3
POSE_GRAPH.constraint_builder.max_constraint_distance = 15.
POSE_GRAPH.constraint_builder.min_score = 0.48
POSE_GRAPH.constraint_builder.global_localization_min_score = 0.60

return options

2.trajectory_builder.lua

rs16_2d_outdoor.lua中包含的子lua文件

include "trajectory_builder_2d.lua"
include "trajectory_builder_3d.lua"

TRAJECTORY_BUILDER = {
  trajectory_builder_2d = TRAJECTORY_BUILDER_2D,
  trajectory_builder_3d = TRAJECTORY_BUILDER_3D,
--  pure_localization_trimmer = {
--    max_submaps_to_keep = 3,
--  },
  collate_fixed_frame = true, -- 是否将数据放入阻塞队列中
  collate_landmarks = false,  -- 是否将数据放入阻塞队列中

3.trajectory_builder_2d.lua

trajectory_builder.lua中包含的子lua文件，如更改是否使用imu数据也需要在此处进行修改

TRAJECTORY_BUILDER_2D = {
  use_imu_data = true,            -- 是否使用imu数据
  min_range = 0.,                 -- 雷达数据的最远最近滤波, 保存中间值
  max_range = 30.,
  min_z = -0.8,                   -- 雷达数据的最高与最低的过滤, 保存中间值
  max_z = 2.,
  missing_data_ray_length = 5.,   -- 超过最大距离范围的数据点（NaN点）或反射较弱的点，即表明周围较空，用这个距离代替
  num_accumulated_range_data = 1, -- 几帧有效的点云数据进行一次扫描匹配
  voxel_filter_size = 0.025,      -- 体素滤波的立方体的边长

  -- 使用固定的voxel滤波之后, 再使用自适应体素滤波器
  -- 体素滤波器 用于生成稀疏点云 以进行 扫描匹配
  adaptive_voxel_filter = {
    max_length = 0.5,             -- 尝试确定最佳的立方体边长, 边长最大为0.5
    min_num_points = 200,         -- 如果存在 较多点 并且大于'min_num_points', 则减小体素长度以尝试获得该最小点数
    max_range = 50.,              -- 距远离原点超过max_range 的点被移除
  },

  -- 闭环检测的自适应体素滤波器, 用于生成稀疏点云 以进行 闭环检测
  loop_closure_adaptive_voxel_filter = {
    max_length = 0.9,
    min_num_points = 100,
    max_range = 50.,
  },

  -- 是否使用 real_time_correlative_scan_matcher 为ceres提供先验信息
  -- 计算复杂度高 , 但是很鲁棒 , 在odom或者imu不准时依然能达到很好的效果
  --如只使用单线激光雷达，频率又很低，建图效果不好总是叠图，可以打开，在扫描匹配前就提供先验信息
  use_online_correlative_scan_matching = false,
  real_time_correlative_scan_matcher = {
    linear_search_window = 0.1,             -- 线性搜索窗口的大小
    angular_search_window = math.rad(20.),  -- 角度搜索窗口的大小
    translation_delta_cost_weight = 1e-1,   -- 用于计算各部分score的权重
    rotation_delta_cost_weight = 1e-1,
  },

  -- ceres匹配的一些配置参数，优化残差相关
  ceres_scan_matcher = {
    occupied_space_weight = 1.,
    translation_weight = 10.,
    rotation_weight = 40.,
    ceres_solver_options = {
      use_nonmonotonic_steps = false,
      max_num_iterations = 20,
      num_threads = 1,
    },
  },

  -- 为了防止子图里插入太多数据, 在插入子图之前之前对数据进行过滤
  motion_filter = {
    max_time_seconds = 5.,
    max_distance_meters = 0.2,
    max_angle_radians = math.rad(1.),
  },

  -- TODO(schwoere,wohe): Remove this constant. This is only kept for ROS.
  imu_gravity_time_constant = 10.,

  -- 位姿预测器
  pose_extrapolator = {
    use_imu_based = false,
    constant_velocity = {
      imu_gravity_time_constant = 10.,
      pose_queue_duration = 0.001,
    },
    imu_based = {
      pose_queue_duration = 5.,
      gravity_constant = 9.806,
      pose_translation_weight = 1.,
      pose_rotation_weight = 1.,
      imu_acceleration_weight = 1.,
      imu_rotation_weight = 1.,
      odometry_translation_weight = 1.,
      odometry_rotation_weight = 1.,
      solver_options = {
        use_nonmonotonic_steps = false;
        max_num_iterations = 10;
        num_threads = 1;
      },
    },
  },

  -- 子图相关的一些配置
  submaps = {
    num_range_data = 90,          -- 一个子图里插入雷达数据的个数的一半
    grid_options_2d = {
      grid_type = "PROBABILITY_GRID", -- 地图的种类, 还可以是tsdf格式
      resolution = 0.05,          --地图的分辨率，重要
    },
    range_data_inserter = {
      range_data_inserter_type = "PROBABILITY_GRID_INSERTER_2D",
      -- 概率占用栅格地图的一些配置
      probability_grid_range_data_inserter = {
        insert_free_space = true,
        hit_probability = 0.55,
        miss_probability = 0.49,
      },
      -- tsdf地图的一些配置
      tsdf_range_data_inserter = {
        truncation_distance = 0.3,
        maximum_weight = 10.,
        update_free_space = false,
        normal_estimation_options = {
          num_normal_samples = 4,
          sample_radius = 0.5,
        },
        project_sdf_distance_to_scan_normal = true,
        update_weight_range_exponent = 0,
        update_weight_angle_scan_normal_to_ray_kernel_bandwidth = 0.5,
        update_weight_distance_cell_to_hit_kernel_bandwidth = 0.5,
      },
    },
  },
}

3.trajectory_builder_3d.lua

trajectory_builder.lua中包含的子lua文件，以上都是前端一些扫描匹配的参数

MAX_3D_RANGE = 60.
INTENSITY_THRESHOLD = 40

TRAJECTORY_BUILDER_3D = {
  min_range = 1.,
  max_range = MAX_3D_RANGE,
  num_accumulated_range_data = 1,
  voxel_filter_size = 0.15,

  -- 在3d slam 时会有2个自适应体素滤波, 一个生成高分辨率点云, 一个生成低分辨率点云
  high_resolution_adaptive_voxel_filter = {
    max_length = 2.,
    min_num_points = 150,
    max_range = 15.,
  },

  low_resolution_adaptive_voxel_filter = {
    max_length = 4.,
    min_num_points = 200,
    max_range = MAX_3D_RANGE,
  },

  use_online_correlative_scan_matching = false,
  real_time_correlative_scan_matcher = {
    linear_search_window = 0.15,
    angular_search_window = math.rad(1.),
    translation_delta_cost_weight = 1e-1,
    rotation_delta_cost_weight = 1e-1,
  },

  ceres_scan_matcher = {
    -- 在3D中,occupied_space_weight_0和occupied_space_weight_1参数分别与高分辨率和低分辨率滤波点云相关
    occupied_space_weight_0 = 1.,
    occupied_space_weight_1 = 6.,
    intensity_cost_function_options_0 = {
        weight = 0.5,
        huber_scale = 0.3,
        intensity_threshold = INTENSITY_THRESHOLD,
    },
    translation_weight = 5.,
    rotation_weight = 4e2,
    only_optimize_yaw = false,
    ceres_solver_options = {
      use_nonmonotonic_steps = false,
      max_num_iterations = 12,
      num_threads = 1,
    },
  },

  motion_filter = {
    max_time_seconds = 0.5,
    max_distance_meters = 0.1,
    max_angle_radians = 0.004,
  },

  rotational_histogram_size = 120,

  -- TODO(schwoere,wohe): Remove this constant. This is only kept for ROS.
  imu_gravity_time_constant = 10.,
  pose_extrapolator = {
    use_imu_based = false,
    constant_velocity = {
      imu_gravity_time_constant = 10.,
      pose_queue_duration = 0.001,
    },
    -- TODO(wohe): Tune these parameters on the example datasets.
    imu_based = {
      pose_queue_duration = 5.,
      gravity_constant = 9.806,
      pose_translation_weight = 1.,
      pose_rotation_weight = 1.,
      imu_acceleration_weight = 1.,
      imu_rotation_weight = 1.,
      odometry_translation_weight = 1.,
      odometry_rotation_weight = 1.,
      solver_options = {
        use_nonmonotonic_steps = false;
        max_num_iterations = 10;
        num_threads = 1;
      },
    },
  },

  submaps = {
    -- 2种分辨率的地图
    high_resolution = 0.10,           -- 用于近距离测量的高分辨率hybrid网格 for local SLAM and loop closure, 用来与小尺寸voxel进行精匹配
    high_resolution_max_range = 20.,  -- 在插入 high_resolution map 之前过滤点云的最大范围
    low_resolution = 0.45,
    num_range_data = 160,             -- 用于远距离测量的低分辨率hybrid网格 for local SLAM only, 用来与大尺寸voxel进行粗匹配
    range_data_inserter = {
      hit_probability = 0.55,
      miss_probability = 0.49,
      num_free_space_voxels = 2,
      intensity_threshold = INTENSITY_THRESHOLD,
    },
  },

  -- When setting use_intensites to true, the intensity_cost_function_options_0
  -- parameter in ceres_scan_matcher has to be set up as well or otherwise
  -- CeresScanMatcher will CHECK-fail.
  use_intensities = false,
}

4.map_builder.lua

rs16_2d_outdoor.lua中包含的子lua文件

include "pose_graph.lua"

MAP_BUILDER = {
  use_trajectory_builder_2d = false,     --2d和3d一定要有一个为true，且只能有一个，可以在1中的主lua里重写配置
  use_trajectory_builder_3d = false,
  num_background_threads = 4,
  pose_graph = POSE_GRAPH,
  collate_by_trajectory = false,

5.pose_graph.lua

map_builder.lua中包含的子lua文件，主要为后端参数设置

-- Copyright 2016 The Cartographer Authors
--
-- Licensed under the Apache License, Version 2.0 (the "License");
-- you may not use this file except in compliance with the License.
-- You may obtain a copy of the License at
--
--      http://www.apache.org/licenses/LICENSE-2.0
--
-- Unless required by applicable law or agreed to in writing, software
-- distributed under the License is distributed on an "AS IS" BASIS,
-- WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-- See the License for the specific language governing permissions and
-- limitations under the License.

POSE_GRAPH = {
  -- 每隔多少个节点执行一次后端优化
  optimize_every_n_nodes = 90, --一般设置成2倍的submaps.num_range_data

  -- 约束构建的相关参数
  constraint_builder = {
    sampling_ratio = 0.3,                 -- 对局部子图进行回环检测时的计算频率, 数值越大, 计算次数越多
    max_constraint_distance = 15.,        -- 对局部子图进行回环检测时能成为约束的最大距离
    min_score = 0.55,                     -- 对局部子图进行回环检测时的最低分数阈值
    global_localization_min_score = 0.6,  -- 对整体子图进行回环检测时的最低分数阈值
    loop_closure_translation_weight = 1.1e4,
    loop_closure_rotation_weight = 1e5,
    log_matches = true,                   -- 打印约束计算的log
    
    -- 基于分支定界算法的2d粗匹配器
    fast_correlative_scan_matcher = {
      linear_search_window = 7.,
      angular_search_window = math.rad(30.),
      branch_and_bound_depth = 7,
    },

    -- 基于ceres的2d精匹配器
    ceres_scan_matcher = {
      occupied_space_weight = 20.,
      translation_weight = 10.,
      rotation_weight = 1.,
      ceres_solver_options = {
        use_nonmonotonic_steps = true,
        max_num_iterations = 10,
        num_threads = 1,
      },
    },

    -- 基于分支定界算法的3d粗匹配器
    fast_correlative_scan_matcher_3d = {
      branch_and_bound_depth = 8,
      full_resolution_depth = 3,
      min_rotational_score = 0.77,
      min_low_resolution_score = 0.55,
      linear_xy_search_window = 5.,
      linear_z_search_window = 1.,
      angular_search_window = math.rad(15.),
    },

    -- 基于ceres的3d精匹配器
    ceres_scan_matcher_3d = {
      occupied_space_weight_0 = 5.,
      occupied_space_weight_1 = 30.,
      translation_weight = 10.,
      rotation_weight = 1.,
      only_optimize_yaw = false,
      ceres_solver_options = {
        use_nonmonotonic_steps = false,
        max_num_iterations = 10,
        num_threads = 1,
      },
    },
  },

  matcher_translation_weight = 5e2,
  matcher_rotation_weight = 1.6e3,

  -- 优化残差方程的相关参数
  optimization_problem = {
    huber_scale = 1e1,                -- 值越大,（潜在）异常值的影响就越大
    acceleration_weight = 1.1e2,      -- 3d里imu的线加速度的权重
    rotation_weight = 1.6e4,          -- 3d里imu的旋转的权重
    
    -- 前端结果残差的权重
    local_slam_pose_translation_weight = 1e5,
    local_slam_pose_rotation_weight = 1e5,
    -- 里程计残差的权重
    odometry_translation_weight = 1e5,
    odometry_rotation_weight = 1e5,
    -- gps残差的权重
    fixed_frame_pose_translation_weight = 1e1,
    fixed_frame_pose_rotation_weight = 1e2,
    fixed_frame_pose_use_tolerant_loss = false,
    fixed_frame_pose_tolerant_loss_param_a = 1,
    fixed_frame_pose_tolerant_loss_param_b = 1,

    log_solver_summary = false,
    use_online_imu_extrinsics_in_3d = true,
    fix_z_in_3d = false,
    ceres_solver_options = {
      use_nonmonotonic_steps = false,
      max_num_iterations = 50,
      num_threads = 7,
    },
  },

  max_num_final_iterations = 200,   -- 在建图结束之后执行一次全局优化, 不要求实时性, 迭代次数多
  global_sampling_ratio = 0.003,    -- 纯定位时候查找回环的频率
  log_residual_histograms = true,
  global_constraint_search_after_n_seconds = 10., -- 纯定位时多少秒执行一次全子图的约束计算

  --  overlapping_submaps_trimmer_2d = {
  --    fresh_submaps_count = 1,
  --    min_covered_area = 2,
  --    min_added_submaps_count = 5,
  --  },
}

三、调参总结

1.调参总结

将常用的调整参数列写如下：

--雷达的最大最小距离
TRAJECTORY_BUILDER_2D.min_range = 0.3
TRAJECTORY_BUILDER_2D.max_range = 100.
--使用多高以上的点云，单线的时候不要设置，多线防止打到地面上的点干扰建图
TRAJECTORY_BUILDER_2D.min_z = 0.2 -- / -0.8
--体素滤波参数
TRAJECTORY_BUILDER_2D.voxel_filter_size = 0.02
--ceres地图的扫描，平移，旋转的权重，影响建图效果，其他基本上是影响计算量等
TRAJECTORY_BUILDER_2D.ceres_scan_matcher.occupied_space_weight = 10.--扫描匹配点云和地图匹配程度，值越大，点云和地图匹配置信度越高
 --残差平移、旋转分量，值越大，越不相信和地图匹配的效果，而是越相信先验位姿的结果
TRAJECTORY_BUILDER_2D.ceres_scan_matcher.translation_weight = 1. 
--如果imu不好，接入后地图旋转厉害，可以将这里的旋转权重减小
TRAJECTORY_BUILDER_2D.ceres_scan_matcher.rotation_weight = 1.

--一个子图插入多少个节点，optimize_every_n_nodes=80*2
TRAJECTORY_BUILDER_2D.submaps.num_range_data = 80.
TRAJECTORY_BUILDER_2D.submaps.grid_options_2d.resolution = 0.1 -- / 0.02
POSE_GRAPH.optimize_every_n_nodes = 160. -- 2倍的num_range_data以上
POSE_GRAPH.constraint_builder.sampling_ratio = 0.3
POSE_GRAPH.constraint_builder.max_constraint_distance = 15.
--回环检测阈值，如果不稳定有错误匹配，可以提高这两个值，场景重复可以降低或者关闭回环
POSE_GRAPH.constraint_builder.min_score = 0.48
POSE_GRAPH.constraint_builder.global_localization_min_score = 0.60

2.wheeltec调参实例

wheeltec大车出现建图效果很差的情况，可以通过调整一下参数进行调整，同时还可以通过降低调整旋转的权重改善旋转时候偏差过大；同时在走廊里进行建图时，因为走廊中场景重复度较高，回环检测导致误匹配，此时关了回环检测用纯激光建图即可。

3.降低延迟与减少计算量、降低纯定位时错误重定位的概率