hdl_localization試讀
- 安裝
- 實驗效果
- hdl_localization包
-
- 總覽
- launch
- apps(程式實作)
-
- globalmap_server_nodelet
-
- `globalmap_server_nodelet::onInit() `
- `globalmap_server_nodelet::initialize_params() `
- hdl_localization_nodelet
-
- `hdl_localization_nodelet::onInit()`
- `hdl_localization_nodelet::initialize_params()`
- `HdlLocalizationNodelet::globalmap_callback`
- `HdlLocalizationNodelet::imu_callback`
- `HdlLocalizationNodelet::points_callback`
- `downsample(const pcl::PointCloud::ConstPtr& cloud)`
- `publish_odometry`
- `matrix2transform`
- include(狀态估計器及ukf)
-
- hdl_localization/pose_estimator.hpp
-
- `PoseEstimator`(構造函數)
- `pose_estimator->predict`(預測)
- `pose_estimator->correct`(觀測)
- hdl_localization/pose_system.hpp
-
- `f`(系統狀态方程)
- `h` (觀測方程)
- kkl/unscented_kalman_filter.hpp
-
- `UnscentedKalmanFilterX`[^2]
- `ukf->predict`
- `ukf->correct`
- `computeSigmaPoints`
- `ensurePositiveFinite`(未實際應用)
- 總結
安裝
簡單易行,環境友好。首先附上網址:koide3/hdl_localization.
前置環境我直接用的apt-get
//安裝pcl
sudo apt-get install libpcl-dev
//安裝其他依賴(記得替換指令中的kinetic為自己的版本,一共4個地方。忘記那個自動的怎麼寫了。。。)
sudo apt-get install ros-kinetic-geodesy ros-kinetic-pcl-ros ros-kinetic-nmea-msgs ros-kinetic-libg2o
安裝好環境之後,按照readme直接走就可以。
建立工作空間→git clone hdl_localization→git clone ndt_omp→編譯1
實驗效果
總的來說效果很好,從官方給的資料集和實驗室之前的包都跑過,定位效果頗好。不用imu也可。
等會跑一個
hdl_localization包
總覽
該軟體是使用nodelet統一管理的(第一次接觸,百度一下很高端的樣子)包内檔案夾很多,apps為定義的兩個類,也就是程式實作。include内為狀态估計器和無迹卡爾曼的實作。launch不用多說。rviz内為rviz的配置檔案。data為執行個體的定位用點雲地圖。
launch
定義了幾個參數,使用nodelet運作了
velodyne_nodelet_manager
、
globalmap_server_nodelet
、
hdl_localization_nodelet
三個節點。如果隻用于仿真,可以在 arguments 前面加上。
<param name="use_sim_time" value="true"/>
apps(程式實作)
本檔案夾是隻有兩個cpp檔案,直接繼承了nodelet的類。代碼量就較少。
globalmap_server_nodelet
類
GlobalmapServerNodelet
繼承了
nodelet::Nodelet
。
關于ros,聲明了三個句柄,1個釋出,1個計時器,1個globalmap的變量。
ros::NodeHandle nh;
ros::NodeHandle mt_nh;
ros::NodeHandle private_nh;
ros::Publisher globalmap_pub;
ros::WallTimer globalmap_pub_timer;
pcl::PointCloud<PointT>::Ptr globalmap;
globalmap_server_nodelet::onInit()
globalmap_server_nodelet::onInit()
這裡是在重寫了初始化函數。同時利用計時器出發回調函數。
void onInit() override {
//定義三個節點,
nh = getNodeHandle();
mt_nh = getMTNodeHandle();
private_nh = getPrivateNodeHandle();
initialize_params();
// publish globalmap with "latched" publisher
globalmap_pub = nh.advertise<sensor_msgs::PointCloud2>("/globalmap", 5, true);
globalmap_pub_timer = nh.createWallTimer(ros::WallDuration(0.05), &GlobalmapServerNodelet::pub_once_cb, this, true, true); //20Hz
}
globalmap_server_nodelet::initialize_params()
globalmap_server_nodelet::initialize_params()
在程式
initialize_params()
中,完成了讀取地圖pcd檔案的功能,并對該地圖進行下采樣,最終的globalmap是下采樣的地圖。
void initialize_params() {
// read globalmap from a pcd file
std::string globalmap_pcd = private_nh.param<std::string>("globalmap_pcd", "");
globalmap.reset(new pcl::PointCloud<PointT>());
pcl::io::loadPCDFile(globalmap_pcd, *globalmap);
globalmap->header.frame_id = "map";
//TODO:這個實際上是沒有到這裡來的,初步想法是沒有utm檔案。類似于經緯度的坐标檔案。
std::ifstream utm_file(globalmap_pcd + ".utm");
if (utm_file.is_open() && private_nh.param<bool>("convert_utm_to_local", true)) {
std::cout << "now utf_file is open" << std::endl;
double utm_easting;
double utm_northing;
double altitude;
utm_file >> utm_easting >> utm_northing >> altitude;
for(auto& pt : globalmap->points) {
pt.getVector3fMap() -= Eigen::Vector3f(utm_easting, utm_northing, altitude);
}
ROS_INFO_STREAM("Global map offset by UTM reference coordinates (x = "
<< utm_easting << ", y = " << utm_northing << ") and altitude (z = " << altitude << ")");
}
//endTODO
// downsample globalmap
double downsample_resolution = private_nh.param<double>("downsample_resolution", 0.1);
boost::shared_ptr<pcl::VoxelGrid<PointT>> voxelgrid(new pcl::VoxelGrid<PointT>());
voxelgrid->setLeafSize(downsample_resolution, downsample_resolution, downsample_resolution);
voxelgrid->setInputCloud(globalmap);
pcl::PointCloud<PointT>::Ptr filtered(new pcl::PointCloud<PointT>());
voxelgrid->filter(*filtered);
globalmap = filtered;
}
同時,每隔0.05s釋出一次。(onInit定義的)
void pub_once_cb(const ros::WallTimerEvent& event) {
globalmap_pub.publish(globalmap);
}
完了
hdl_localization_nodelet
類
HdlLocalizationNodelet
繼承了
nodelet::Nodelet
。
這次我要先看初始化了。
hdl_localization_nodelet::onInit()
hdl_localization_nodelet::onInit()
void onInit() override {
//依然是三個句柄
nh = getNodeHandle();
mt_nh = getMTNodeHandle();
private_nh = getPrivateNodeHandle();
//這裡的時間用了boost庫裡的 circular_buffer<double>。感興趣的可以自己百度一下,畢竟……我也沒用過。
processing_time.resize(16);
//這些參數,又來了。
initialize_params();
//這個預設的base_link, launch裡覆寫了。實際上是velodyne。參數類的在launch裡改寫了一部分,這裡就不一一贅述了。可以自己對比來看,比較容易。
odom_child_frame_id = private_nh.param<std::string>("odom_child_frame_id", "base_link");
//是否使用imu
use_imu = private_nh.param<bool>("use_imu", true);
//imu是否倒置
invert_imu = private_nh.param<bool>("invert_imu", false);
if(use_imu) {//如果使用imu,則定義訂閱函數。
NODELET_INFO("enable imu-based prediction");
imu_sub = mt_nh.subscribe("/gpsimu_driver/imu_data", 256, &HdlLocalizationNodelet::imu_callback, this);
}
//點雲資料、全局地圖、初始位姿的訂閱。initialpose_sub隻是用于rviz劃點用的。
points_sub = mt_nh.subscribe("/velodyne_points", 5, &HdlLocalizationNodelet::points_callback, this);
globalmap_sub = nh.subscribe("/globalmap", 1, &HdlLocalizationNodelet::globalmap_callback, this);
initialpose_sub = nh.subscribe("/initialpose", 8, &HdlLocalizationNodelet::initialpose_callback, this);
//釋出裡程計資訊,以及對齊後的點雲資料。
pose_pub = nh.advertise<nav_msgs::Odometry>("/odom", 5, false);
aligned_pub = nh.advertise<sensor_msgs::PointCloud2>("/aligned_points", 5, false);
}
hdl_localization_nodelet::initialize_params()
hdl_localization_nodelet::initialize_params()
初始化參數
void initialize_params() {
// intialize scan matching method
double downsample_resolution = private_nh.param<double>("downsample_resolution", 0.1);
std::string ndt_neighbor_search_method = private_nh.param<std::string>("ndt_neighbor_search_method", "DIRECT7");
double ndt_resolution = private_nh.param<double>("ndt_resolution", 1.0);
boost::shared_ptr<pcl::VoxelGrid<PointT>> voxelgrid(new pcl::VoxelGrid<PointT>());
voxelgrid->setLeafSize(downsample_resolution, downsample_resolution, downsample_resolution);
downsample_filter = voxelgrid;
//定義了ndt和glcp
pclomp::NormalDistributionsTransform<PointT, PointT>::Ptr ndt(new pclomp::NormalDistributionsTransform<PointT, PointT>());
pclomp::GeneralizedIterativeClosestPoint<PointT, PointT>::Ptr gicp(new pclomp::GeneralizedIterativeClosestPoint<PointT, PointT>());
//ndt參數與搜尋方法。預設DIRECT7,作者說效果不好可以嘗試改為DIRECT1.
ndt->setTransformationEpsilon(0.01);
ndt->setResolution(ndt_resolution);
if(ndt_neighbor_search_method == "DIRECT1") {
NODELET_INFO("search_method DIRECT1 is selected");
ndt->setNeighborhoodSearchMethod(pclomp::DIRECT1);
registration = ndt;
} else if(ndt_neighbor_search_method == "DIRECT7") {
NODELET_INFO("search_method DIRECT7 is selected");
ndt->setNeighborhoodSearchMethod(pclomp::DIRECT7);
registration = ndt;
} else if(ndt_neighbor_search_method == "GICP_OMP"){
NODELET_INFO("search_method GICP_OMP is selected");
registration = gicp;
}
else {
if(ndt_neighbor_search_method == "KDTREE") {
NODELET_INFO("search_method KDTREE is selected");
} else {
NODELET_WARN("invalid search method was given");
NODELET_WARN("default method is selected (KDTREE)");
}
ndt->setNeighborhoodSearchMethod(pclomp::KDTREE);
registration = ndt;
}
// initialize pose estimator
//設定起點用。
if(private_nh.param<bool>("specify_init_pose", true)) {
NODELET_INFO("initialize pose estimator with specified parameters!!");
pose_estimator.reset(new hdl_localization::PoseEstimator(registration,
ros::Time::now(),
Eigen::Vector3f(private_nh.param<double>("init_pos_x", 0.0), private_nh.param<double>("init_pos_y", 0.0), private_nh.param<double>("init_pos_z", 0.0)),
Eigen::Quaternionf(private_nh.param<double>("init_ori_w", 1.0), private_nh.param<double>("init_ori_x", 0.0), private_nh.param<double>("init_ori_y", 0.0), private_nh.param<double>("init_ori_z", 0.0)),
private_nh.param<double>("cool_time_duration", 0.5)
));
}
}
-----------下面就是回調函數的處理---------
實際使用的回調函數就是
HdlLocalizationNodelet::imu_callback
、
HdlLocalizationNodelet::points_callback
以及
HdlLocalizationNodelet::globalmap_callback
三個。分别訂閱了
"/gpsimu_driver/imu_data"
、
"/velodyne_points"
以及
"/globalmap"
兩個話題。
首先看
HdlLocalizationNodelet::globalmap_callback
。完成了對全局地圖的訂閱以及從ros消息到點雲的轉化。
HdlLocalizationNodelet::globalmap_callback
HdlLocalizationNodelet::globalmap_callback
void globalmap_callback(const sensor_msgs::PointCloud2ConstPtr& points_msg) {
NODELET_INFO("globalmap received!");
pcl::PointCloud<PointT>::Ptr cloud(new pcl::PointCloud<PointT>());
pcl::fromROSMsg(*points_msg, *cloud);
globalmap = cloud;
//釋出出來的全局地圖用作配準用的目标點雲。這裡就是globalmap_server_nodelet發出來的。
registration->setInputTarget(globalmap);
}
HdlLocalizationNodelet::imu_callback
HdlLocalizationNodelet::imu_callback
接收imu并扔到imu_data裡去。會在
HdlLocalizationNodelet::points_callback
裡用到。
void imu_callback(const sensor_msgs::ImuConstPtr& imu_msg) {
std::lock_guard<std::mutex> lock(imu_data_mutex);
imu_data.push_back(imu_msg);
}
HdlLocalizationNodelet::points_callback
HdlLocalizationNodelet::points_callback
輸入點雲輸出位姿。
void points_callback(const sensor_msgs::PointCloud2ConstPtr& points_msg) {
//加鎖
std::lock_guard<std::mutex> estimator_lock(pose_estimator_mutex);
if(!pose_estimator) {//等待位姿估計器初始化
NODELET_ERROR("waiting for initial pose input!!");
return;
}
if(!globalmap) {//等待全局地圖
NODELET_ERROR("globalmap has not been received!!");
return;
}
//将ros消息轉化為點雲
const auto& stamp = points_msg->header.stamp;
pcl::PointCloud<PointT>::Ptr pcl_cloud(new pcl::PointCloud<PointT>());
pcl::fromROSMsg(*points_msg, *pcl_cloud);
//檢查
if(pcl_cloud->empty()) {
NODELET_ERROR("cloud is empty!!");
return;
}
//将點雲轉換到odom_child_frame_id 坐标系。
//但是這個tf是自己發的,這一個還要再看一下。TODO。雞生蛋蛋生雞的問題一直搞不太懂。
// transform pointcloud into odom_child_frame_id
pcl::PointCloud<PointT>::Ptr cloud(new pcl::PointCloud<PointT>());
if(!pcl_ros::transformPointCloud(odom_child_frame_id, *pcl_cloud, *cloud, this->tf_listener)) {
NODELET_ERROR("point cloud cannot be transformed into target frame!!");
return;
}
//對點雲下采樣。這裡用到的是同一檔案下的函數。後面會放上。
auto filtered = downsample(cloud);
// predict
if(!use_imu) {//不用imu則用0。
pose_estimator->predict(stamp, Eigen::Vector3f::Zero(), Eigen::Vector3f::Zero());
} else {
std::lock_guard<std::mutex> lock(imu_data_mutex);
auto imu_iter = imu_data.begin();
//利用imu資料疊代。
for(imu_iter; imu_iter != imu_data.end(); imu_iter++) {
//若目前點雲時間早于imu雷達,則跳出。imu做預測,點雲觀測。
if(stamp < (*imu_iter)->header.stamp) {
break;
}
//讀取線加速度和角速度。判斷是否倒置imu
const auto& acc = (*imu_iter)->linear_acceleration;
const auto& gyro = (*imu_iter)->angular_velocity;
double gyro_sign = invert_imu ? -1.0 : 1.0;
//利用imu資料做位姿的預測。這裡用了pose_estimator→predict,進一步調用了ukf進行估計。還沒具體看ukf。 TODO
pose_estimator->predict((*imu_iter)->header.stamp, Eigen::Vector3f(acc.x, acc.y, acc.z), gyro_sign * Eigen::Vector3f(gyro.x, gyro.y, gyro.z));
}
//删除用過的imu資料
imu_data.erase(imu_data.begin(), imu_iter);
}
// correct
auto t1 = ros::WallTime::now();
//用pose_estimator 來矯正點雲。pcl庫配準。擷取到結果後利用ukf矯正位姿。
auto aligned = pose_estimator->correct(filtered);
auto t2 = ros::WallTime::now();
processing_time.push_back((t2 - t1).toSec());
double avg_processing_time = std::accumulate(processing_time.begin(), processing_time.end(), 0.0) / processing_time.size();
// NODELET_INFO_STREAM("processing_time: " << avg_processing_time * 1000.0 << "[msec]");
//如果有訂閱才釋出
if(aligned_pub.getNumSubscribers()) {
aligned->header.frame_id = "map";
aligned->header.stamp = cloud->header.stamp;
aligned_pub.publish(aligned);
}
//釋出裡程計。時間戳為目前幀雷達時間,裡程計位姿為ukf校正後位姿。同時也會釋出從map到odom_child_frame_id的tf
publish_odometry(points_msg->header.stamp, pose_estimator->matrix());
}
----------主要流程到此結束,下面是其他的一些功能函數-----------
downsample(const pcl::PointCloud<PointT>::ConstPtr& cloud)
downsample(const pcl::PointCloud<PointT>::ConstPtr& cloud)
目前幀點雲資料下采樣
pcl::PointCloud<PointT>::ConstPtr downsample(const pcl::PointCloud<PointT>::ConstPtr& cloud) const {
//在函數initialize_params()裡聲明了。0.1,0.1,0.1網格
if(!downsample_filter) {
return cloud;
}
pcl::PointCloud<PointT>::Ptr filtered(new pcl::PointCloud<PointT>());
downsample_filter->setInputCloud(cloud);
downsample_filter->filter(*filtered);
filtered->header = cloud->header;
return filtered;
}
publish_odometry
publish_odometry
釋出裡程計的tf和msg。輸入為目前幀點雲時間戳與pose_estimator的結果矩陣。這裡還用到了
matrix2transform
這個函數,用于做eigen矩陣到tf的轉化(取值)。
void publish_odometry(const ros::Time& stamp, const Eigen::Matrix4f& pose) {
// broadcast the transform over tf
geometry_msgs::TransformStamped odom_trans = matrix2transform(stamp, pose, "map", odom_child_frame_id);
pose_broadcaster.sendTransform(odom_trans);
// publish the transform
nav_msgs::Odometry odom;
odom.header.stamp = stamp;
odom.header.frame_id = "map";
odom.pose.pose.position.x = pose(0, 3);
odom.pose.pose.position.y = pose(1, 3);
odom.pose.pose.position.z = pose(2, 3);
odom.pose.pose.orientation = odom_trans.transform.rotation;
odom.child_frame_id = odom_child_frame_id;
odom.twist.twist.linear.x = 0.0;
odom.twist.twist.linear.y = 0.0;
odom.twist.twist.angular.z = 0.0;
pose_pub.publish(odom);
}
matrix2transform
matrix2transform
matrix2transform(const ros::Time& stamp, const Eigen::Matrix4f& pose, const std::string& frame_id, const std::string& child_frame_id)
從matrix 到 geometry_msgs::TransformStamped。
geometry_msgs::TransformStamped matrix2transform(const ros::Time& stamp, const Eigen::Matrix4f& pose, const std::string& frame_id, const std::string& child_frame_id) {
Eigen::Quaternionf quat(pose.block<3, 3>(0, 0));
quat.normalize();
geometry_msgs::Quaternion odom_quat;
odom_quat.w = quat.w();
odom_quat.x = quat.x();
odom_quat.y = quat.y();
odom_quat.z = quat.z();
geometry_msgs::TransformStamped odom_trans;
odom_trans.header.stamp = stamp;
odom_trans.header.frame_id = frame_id;
odom_trans.child_frame_id = child_frame_id;
odom_trans.transform.translation.x = pose(0, 3);
odom_trans.transform.translation.y = pose(1, 3);
odom_trans.transform.translation.z = pose(2, 3);
odom_trans.transform.rotation = odom_quat;
return odom_trans;
}
include(狀态估計器及ukf)
apps裡的兩個cpp大緻内容均為以上。可以看到在
points_callback
裡使用了
pose_estimator
作為位姿的估計。而該類又使用了ukf作為位姿的解算。二者的實作都在include檔案夾内。
hdl_localization/pose_estimator.hpp
該檔案定義了類
PoseEstimator
。
PoseEstimator
(構造函數)
PoseEstimator
首先看構造函數。可以看到在初始化之後,最重要的是進入了ukf的處理。
PoseEstimator(pcl::Registration<PointT, PointT>::Ptr& registration, const ros::Time& stamp, const Eigen::Vector3f& pos, const Eigen::Quaternionf& quat, double cool_time_duration = 1.0)
: init_stamp(stamp),
registration(registration),
cool_time_duration(cool_time_duration)
{
//機關陣初始化,随後給過程噪聲。
process_noise = Eigen::MatrixXf::Identity(16, 16);
process_noise.middleRows(0, 3) *= 1.0;
process_noise.middleRows(3, 3) *= 1.0;
process_noise.middleRows(6, 4) *= 0.5;
process_noise.middleRows(10, 3) *= 1e-6;
process_noise.middleRows(13, 3) *= 1e-6;
//測量噪聲,機關陣
Eigen::MatrixXf measurement_noise = Eigen::MatrixXf::Identity(7, 7);
measurement_noise.middleRows(0, 3) *= 0.01;
measurement_noise.middleRows(3, 4) *= 0.001;
//權重平均的位姿。
Eigen::VectorXf mean(16);
mean.middleRows(0, 3) = pos;
mean.middleRows(3, 3).setZero();
mean.middleRows(6, 4) = Eigen::Vector4f(quat.w(), quat.x(), quat.y(), quat.z());
mean.middleRows(10, 3).setZero();
mean.middleRows(13, 3).setZero();
//初始化協方差
Eigen::MatrixXf cov = Eigen::MatrixXf::Identity(16, 16) * 0.01;
//聲明posesystem。
PoseSystem system;
//初始化ukf
ukf.reset(new kkl::alg::UnscentedKalmanFilterX<float, PoseSystem>(system, 16, 6, 7, process_noise, measurement_noise, mean, cov));
}
pose_estimator->predict
(預測)
pose_estimator->predict
另外在
hdl_localization.cpp
中用到的
pose_estimator->predict
等也在本檔案進行了解釋。
void predict(const ros::Time& stamp, const Eigen::Vector3f& acc, const Eigen::Vector3f& gyro) {
//目前與初始化的時間間隔小于設定的時間,或prev_stamp(上次更新時間)為0(未更新),或prev_stamp等于目前時間。則更新prev_stamp并跳出。
if((stamp - init_stamp).toSec() < cool_time_duration || prev_stamp.is_zero() || prev_stamp == stamp) {
prev_stamp = stamp;
return;
}
//正常處理,首先計算dt,更新prev_stamp。
double dt = (stamp - prev_stamp).toSec();
prev_stamp = stamp;
//對ukf設定噪聲和處理間隔。
ukf->setProcessNoiseCov(process_noise * dt);
ukf->system.dt = dt;
//利用imu資料定義控制量
Eigen::VectorXf control(6);
control.head<3>() = acc;
control.tail<3>() = gyro;
//利用ukf預測。
ukf->predict(control);
}
pose_estimator->correct
(觀測)
pose_estimator->correct
pcl::PointCloud<PointT>::Ptr correct(const pcl::PointCloud<PointT>::ConstPtr& cloud) {
//機關陣來初始化
Eigen::Matrix4f init_guess = Eigen::Matrix4f::Identity();
init_guess.block<3, 3>(0, 0) = quat().toRotationMatrix();
init_guess.block<3, 1>(0, 3) = pos();
//點雲的配準。ndt
pcl::PointCloud<PointT>::Ptr aligned(new pcl::PointCloud<PointT>());
registration->setInputSource(cloud);
registration->align(*aligned, init_guess);
//讀取資料
Eigen::Matrix4f trans = registration->getFinalTransformation();
Eigen::Vector3f p = trans.block<3, 1>(0, 3);
Eigen::Quaternionf q(trans.block<3, 3>(0, 0));
if(quat().coeffs().dot(q.coeffs()) < 0.0f) {
q.coeffs() *= -1.0f;
}
//填充至觀測矩陣observation
Eigen::VectorXf observation(7);
observation.middleRows(0, 3) = p;
observation.middleRows(3, 4) = Eigen::Vector4f(q.w(), q.x(), q.y(), q.z());
//ukf更新
ukf->correct(observation);
return aligned;
}
----------還有一些簡單的函數不再說明了。直接怼上,很好了解。-----------
/* getters */
Eigen::Vector3f pos() const {
return Eigen::Vector3f(ukf->mean[0], ukf->mean[1], ukf->mean[2]);
}
Eigen::Vector3f vel() const {
return Eigen::Vector3f(ukf->mean[3], ukf->mean[4], ukf->mean[5]);
}
Eigen::Quaternionf quat() const {
return Eigen::Quaternionf(ukf->mean[6], ukf->mean[7], ukf->mean[8], ukf->mean[9]).normalized();
}
Eigen::Matrix4f matrix() const {
Eigen::Matrix4f m = Eigen::Matrix4f::Identity();
m.block<3, 3>(0, 0) = quat().toRotationMatrix();
m.block<3, 1>(0, 3) = pos();
return m;
}
hdl_localization/pose_system.hpp
本檔案定義了完成了類
PoseSystem
的實作。主要是實作了ukf裡 矩陣
f(定義了系統)
和
h(觀測)
代碼實作。這是要扔到ukf中去的。
系統狀态量16位,分别是位姿(3)、速度(3)、四元數(4)、加速度偏差(3)、陀螺儀偏差(3)。另還有6位控制量,加速度(3)和陀螺儀(3)。
狀态量 | 表示 |
---|---|
位置 | p t pt pt = [ p x , p y , p z ] T px, \quad py, \quad pz]^{T} px,py,pz]T |
速度 | v t vt vt = [ v x , v y , v z ] T vx, \quad vy, \quad vz]^{T} vx,vy,vz]T |
四元數 | q t qt qt = [ q w , q x , q y , q z ] T qw, \quad qx, \quad qy, \quad qz]^{T} qw,qx,qy,qz]T |
加速度偏差 | a c c _ b i a s acc\_bias acc_bias = [ a c c _ b i a s x , a c c _ b i a s y , a c c _ b i a s z ] T acc\_bias_x, \quad acc\_bias_y, \quad acc\_bias_z]^{T} acc_biasx,acc_biasy,acc_biasz]T |
陀螺儀偏差 | g y r o _ b i a s gyro\_bias gyro_bias = [ g y r o _ b i a s x , g y r o _ b i a s y , g y r o _ b i a s z ] T gyro\_bias_x, \quad gyro\_bias_y, \quad gyro\_bias_z]^{T} gyro_biasx,gyro_biasy,gyro_biasz]T |
控制量 | 表示 |
---|---|
加速度 | r a w _ a c c raw\_acc raw_acc = [ r a w _ a c c x , r a w _ a c c y , r a w _ a c c z ] T raw\_acc_x, \quad raw\_acc_y, \quad raw\_acc_z]^{T} raw_accx,raw_accy,raw_accz]T |
陀螺儀 | r a w _ g y r o raw\_gyro raw_gyro = [ g y r o _ b i a s x , g y r o _ b i a s y , g y r o _ b i a s z ] T gyro\_bias_x, \quad gyro\_bias_y, \quad gyro\_bias_z]^{T} gyro_biasx,gyro_biasy,gyro_biasz]T |
f
(系統狀态方程)
f
VectorXt f(const VectorXt& state, const VectorXt& control) const {
VectorXt next_state(16);
Vector3t pt = state.middleRows(0, 3); //位置
Vector3t vt = state.middleRows(3, 3); //速度
Quaterniont qt(state[6], state[7], state[8], state[9]);
qt.normalize(); // 歸一化四元數
Vector3t acc_bias = state.middleRows(10, 3); //加速度偏差
Vector3t gyro_bias = state.middleRows(13, 3); //陀螺儀偏差
Vector3t raw_acc = control.middleRows(0, 3); //加速度控制
Vector3t raw_gyro = control.middleRows(3, 3); //陀螺儀控制
//下一時刻狀态
// position 。 首先更新位置
next_state.middleRows(0, 3) = pt + vt * dt; //
// velocity。 更新速度,實際上并沒有利用加速度矯正速度,原因是認為加速度噪聲較大,對最終的精度并沒有貢獻。
Vector3t g(0.0f, 0.0f, -9.80665f);
Vector3t acc_ = raw_acc - acc_bias;
Vector3t acc = qt * acc_;
next_state.middleRows(3, 3) = vt; // + (acc - g) * dt; // acceleration didn't contribute to accuracy due to large noise
// orientation。首先完成了陀螺儀的增量計算并歸一化(直接轉化為四元數形式),将其轉換為下一時刻的四元數。
Vector3t gyro = raw_gyro - gyro_bias;
Quaterniont dq(1, gyro[0] * dt / 2, gyro[1] * dt / 2, gyro[2] * dt / 2);
dq.normalize();
Quaterniont qt_ = (qt * dq).normalized();
next_state.middleRows(6, 4) << qt_.w(), qt_.x(), qt_.y(), qt_.z();
//将目前控制量傳入下一時刻的狀态向量。認為加速度和角速度上偏差不變
next_state.middleRows(10, 3) = state.middleRows(10, 3); // constant bias on acceleration
next_state.middleRows(13, 3) = state.middleRows(13, 3); // constant bias on angular velocity
return next_state;
}
h
(觀測方程)
h
觀測方程直接将目前輸入狀态量作為觀測量。這裡的輸入是在更新階段(correct)生成的帶誤差方差的(
error variances
)的擴充狀态空間下的(
extended state space
)狀态量,也就是
ext_sigma_points
。
// observation equation
VectorXt h(const VectorXt& state) const {
VectorXt observation(7);
observation.middleRows(0, 3) = state.middleRows(0, 3);
observation.middleRows(3, 4) = state.middleRows(6, 4).normalized();
return observation;
}
kkl/unscented_kalman_filter.hpp
本檔案中主要的函數也就構造函數、預測、矯正、計算sigma點、使協方差矩陣正有限(不太清楚)五個。
UnscentedKalmanFilterX
2
UnscentedKalmanFilterX
首先,構造函數。可以看到輸入了一系列包括待估計系統、狀态向量次元、輸入次元、觀測次元、兩個噪聲、參數等等。完成了初始化操作。
UnscentedKalmanFilterX(const System& system, int state_dim, int input_dim, int measurement_dim, const MatrixXt& process_noise, const MatrixXt& measurement_noise, const VectorXt& mean, const MatrixXt& cov)
: state_dim(state_dim),
input_dim(input_dim),
measurement_dim(measurement_dim),
N(state_dim),
M(input_dim),
K(measurement_dim),
S(2 * state_dim + 1),
mean(mean),
cov(cov),
system(system),
process_noise(process_noise),
measurement_noise(measurement_noise),
lambda(1),
normal_dist(0.0, 1.0)
{
//設定長度。
weights.resize(S, 1);
sigma_points.resize(S, N);
ext_weights.resize(2 * (N + K) + 1, 1);
ext_sigma_points.resize(2 * (N + K) + 1, N + K);
expected_measurements.resize(2 * (N + K) + 1, K);
// initialize weights for unscented filter
weights[0] = lambda / (N + lambda);
for (int i = 1; i < 2 * N + 1; i++) {
weights[i] = 1 / (2 * (N + lambda));
}
// weights for extended state space which includes error variances
ext_weights[0] = lambda / (N + K + lambda);
for (int i = 1; i < 2 * (N + K) + 1; i++) {
ext_weights[i] = 1 / (2 * (N + K + lambda));
}
}
ukf->predict
ukf->predict
通過
pose_estimator->predict調用
。
void predict(const VectorXt& control) {
// calculate sigma points. ukf的sigma點
ensurePositiveFinite(cov);
computeSigmaPoints(mean, cov, sigma_points);
//sigma_points更新。用在posesystem中定義的f函數來進行。
for (int i = 0; i < S; i++) {
sigma_points.row(i) = system.f(sigma_points.row(i), control);
}
/*----至此,sigma_points裡存儲的就是目前時刻的由ukf輸出的系統狀态。-----*/
//過程噪聲,即ukf中的矩陣R
const auto& R = process_noise;
// unscented transform。定義目前的平均狀态和協方差矩陣,并設定為0矩陣。
VectorXt mean_pred(mean.size());
MatrixXt cov_pred(cov.rows(), cov.cols());
mean_pred.setZero();
cov_pred.setZero();
//權重平均,預測狀态
for (int i = 0; i < S; i++) {
mean_pred += weights[i] * sigma_points.row(i);
}
//根據狀态預測協方差。
for (int i = 0; i < S; i++) {
VectorXt diff = sigma_points.row(i).transpose() - mean_pred;
cov_pred += weights[i] * diff * diff.transpose();
}
//附加過程噪聲R,在pose_estimator中給出初值
cov_pred += R;
//更新mean和cov
mean = mean_pred;
cov = cov_pred;
}
ukf->correct
ukf->correct
通過
pose_estimator->correct調用
。
void correct(const VectorXt& measurement) {
//N-狀态方程次元。K-觀測次元
// create extended state space which includes error variances
VectorXt ext_mean_pred = VectorXt::Zero(N + K, 1);
MatrixXt ext_cov_pred = MatrixXt::Zero(N + K, N + K);
//左上角N行1列
ext_mean_pred.topLeftCorner(N, 1) = VectorXt(mean);
//左上角N行N列
ext_cov_pred.topLeftCorner(N, N) = MatrixXt(cov);
//右下角K行K列。初始化為在pose_estimator輸入的噪聲。位置噪聲0.01,四元數0.001
ext_cov_pred.bottomRightCorner(K, K) = measurement_noise;
/*---------------- 經過以上操作,現在擴充狀态變量前N項為mean,擴充協方差左上角為N*N的cov,右下角為K*K的觀測噪聲--------------*/
//驗證并計算
ensurePositiveFinite(ext_cov_pred);
//利用擴充狀态空間的參數計算sigma點
computeSigmaPoints(ext_mean_pred, ext_cov_pred, ext_sigma_points);
// unscented transform
//這裡使用了 ukf 的h 函數來計算觀測。
//ext_sigma_points、expected_measurements是(2 * (N + K) + 1, K)的矩陣
//沒太看明白 TODO
//取左上角前N個量,加上右下角K個量。
expected_measurements.setZero();
for (int i = 0; i < ext_sigma_points.rows(); i++) {
expected_measurements.row(i) = system.h(ext_sigma_points.row(i).transpose().topLeftCorner(N, 1));
expected_measurements.row(i) += VectorXt(ext_sigma_points.row(i).transpose().bottomRightCorner(K, 1));
}
//權重平均。同predict函數相似。
VectorXt expected_measurement_mean = VectorXt::Zero(K);
for (int i = 0; i < ext_sigma_points.rows(); i++) {
expected_measurement_mean += ext_weights[i] * expected_measurements.row(i);
}
MatrixXt expected_measurement_cov = MatrixXt::Zero(K, K);
for (int i = 0; i < ext_sigma_points.rows(); i++) {
VectorXt diff = expected_measurements.row(i).transpose() - expected_measurement_mean;
expected_measurement_cov += ext_weights[i] * diff * diff.transpose();
}
// calculated transformed covariance
//轉換方差。用于計算sigama,進而計算卡爾曼增益
MatrixXt sigma = MatrixXt::Zero(N + K, K);
for (int i = 0; i < ext_sigma_points.rows(); i++) {
auto diffA = (ext_sigma_points.row(i).transpose() - ext_mean_pred);
auto diffB = (expected_measurements.row(i).transpose() - expected_measurement_mean);
sigma += ext_weights[i] * (diffA * diffB.transpose());
}
kalman_gain = sigma * expected_measurement_cov.inverse();
const auto& K = kalman_gain;
//更新最後的ukf。
VectorXt ext_mean = ext_mean_pred + K * (measurement - expected_measurement_mean);
MatrixXt ext_cov = ext_cov_pred - K * expected_measurement_cov * K.transpose();
mean = ext_mean.topLeftCorner(N, 1);
cov = ext_cov.topLeftCorner(N, N);
}
computeSigmaPoints
computeSigmaPoints
通過mean和cov計算sigma點。思路是将cov做Cholesky分解,用下三角矩陣L對mean做處理。得到一系列sigma_points.
void computeSigmaPoints(const VectorXt& mean, const MatrixXt& cov, MatrixXt& sigma_points) {
const int n = mean.size();
assert(cov.rows() == n && cov.cols() == n);
//llt分解。求Cholesky分解A=LL^*=U^*U。L是下三角矩陣
Eigen::LLT<MatrixXt> llt;
llt.compute((n + lambda) * cov);
MatrixXt l = llt.matrixL();
//mean是列向量。這裡會自動轉置處理。
sigma_points.row(0) = mean;
for (int i = 0; i < n; i++) {
sigma_points.row(1 + i * 2) = mean + l.col(i); //奇數1357
sigma_points.row(1 + i * 2 + 1) = mean - l.col(i); //偶數2468
}
}
ensurePositiveFinite
(未實際應用)
ensurePositiveFinite
保證協方差的正有限。未實際應用。
void ensurePositiveFinite(MatrixXt& cov) {
return;
//就到這裡了,在上面就return掉了。
const double eps = 1e-9;
Eigen::EigenSolver<MatrixXt> solver(cov);
MatrixXt D = solver.pseudoEigenvalueMatrix(); //特征值
MatrixXt V = solver.pseudoEigenvectors(); //特征向量
for (int i = 0; i < D.rows(); i++) {
if (D(i, i) < eps) {
D(i, i) = eps;
}
}
cov = V * D * V.inverse();
}
總結
其實看到這裡我已經忘了整個的架構了。是以再捋一遍。搞完再加。
- 可能會出現
運作下面指令即可No rule to make target '/usr/lib/x86_64-linux-gnu/libproj.so
↩︎sudo ln -s /usr/lib/x86_64-linux-gnu/libproj.so.9 /usr/lib/x86_64-linux-gnu/libproj.so
- 關于ukf,我看的這個部落客的内容。https://blog.csdn.net/l2014010671/article/details/93305871 ↩︎