機器人體驗營筆記（二）基礎

全文内容來源于國外權威資料彙總整理，具體資訊請查閱文末參考文獻。

For specific information, please refer to the reference at the end of the article.

概念 Concept

1. 行為模式 Behavior

機器人行為由精心設計的算法和表示建構的機制。Autonomous robot behaviors are mechanisms constructed from carefully designed algorithms and representations.

基礎技術：Underlying technologies:

有限狀态機、事件觸發結構。State machines; event-based architectures.

2. 視覺感覺 Visual Perception

機器人使用複雜但不完美的計算機視覺算法來從像素陣列推導出真實世界的對象表示。Robots use sophisticated but imperfect computer vision algorithms to deduce real world object representations from arrays of pixels.

基礎技術：Underlying technologies:

Hough變換; 人臉檢測算法; ArUco标記等。Hough transforms; face detection algorithms; ArUco markers; much more

3. 定位 Localization

機器人使用裡程計、視覺地标和其他類型的傳感器資訊來估計他們在世界上的位置。Robots estimate their position in the world using a combination of odometry, visual landmarks, and other types of sensor information.

基礎技術：Underlying technologies:

粒子濾波器; SLAM（同時定位和地圖建構）算法。 Particle filters; SLAM (Simultaneous Localization and Mapping) algorithms.

4. 路徑規劃 Path Plan

機器人使用路徑規劃器在世界中導航，以搜尋障礙物周圍的路線以達到目标。Robots navigate through the world using a path planner to search for routes around obstacles to reach their goal.

基礎技術：Underlying technologies:

路徑規劃算法，如波前算法或RRT（快速探索随機樹）。Path planning algorithms such as wavefront algorithms or RRTs (Rapidly-exploring Random Trees)

5. 控制 Control

機器人本體被描述為運動樹，并使用運動學求解器在關節角度和本體坐标之間進行變換。Robots describe their bodies as kinematic trees and use kinematics solvers to translate between joint angles and body coordinates.

基礎技術：Underlying technologies:

運動學描述檔案; Denavit-Hartenberg慣例; 正向和反向運動學求解器。Kinematic description files; Denavit-Hartenberg conventions; forward and inverse kinematics solvers.

6-10等更多内容參考文獻，不一一列舉。

軟體開發工具包 SDK

簡要介紹Cozmo SDK，Introduction to the Cozmo SDK：

$ git clone https://github.com/anki/cozmo-python-sdk

通過USB連接配接蘋果或安卓手機并配置好。具體内容參考之前博文。Connect your Apple or Android phone via USB and configure it. For details, please refer to the previous blog post.

測試 test：

運作 Hello World示例，Run the Hello World demo by doing:

$ cd ~/cozmo-python-sdk/examples/tutorials/01_basics

$ ./01_hello_world.py

不同電腦依據配置稍有不同，但基本内容一緻。

嘗試一下遠端遙控應用？ Try the Cozmo remote control app by doing：

$ cd ~/cozmo-python-sdk/examples/apps

$ ./remote_control_cozmo.py

工具包 Tools

簡要介紹Cozmo tools，Introduction to Cozmo-Tools：

下載下傳：Download a private copy of the cozmo-tools package by doing:

$ git clone https://github.com/touretzkyds/cozmo-tools

配置環境：Add the cozmo-tools directory to your search path by putting the following lines in

your .bashrc and then doing "source .bashrc":

export PATH=~/cozmo-tools:$PATH

export PYTHONPATH=~/cozmo-tools

運作，Run the simple_cli command line interpreter and display the camera viewer by doing：

$ simple_cli

等待完成，再運作，then, doing：

C> show viewer

C> show worldmap_viewer

C> show particle_viewer

在環境中遙控機器人，檢視各視窗中示範的結果。Using the WASD keys in the particle viewer window, turn Cozmo so he can see a cube. What happens in the worldmap viewer?

Vector機器人資料，請自主查找。更多功能，更多内容。

持續更新完善，時間标簽：2019-07-25。

參考文獻references：

Anki文檔：http://cozmosdk.anki.com/docs/

cozmopedia：https://github.com/touretzkyds/cozmopedia/wiki

認知機器人學：https://blog.csdn.net/ZhangRelay/article/details/86736743