輪式機器人之室內定位

Indoor Positioning of a Wheeled Robot

劉不耀(Pu-Yao Liu)

指導教授 ： 徐元寶

國立虎尾科技大學/電機資訊學院/資訊工程系碩士班/碩士(2017年)

摘要

本論文融合里程計、方向感測器與雷射測距儀(Laser Range Finder, LRF)之感測資料，以擴展卡爾曼濾波器(Extended Kalman Filter, EKF)演算法校正一輪式機器人之室內定位誤差。本研究使用可編程系統單晶片(System on a Programmable Chip, SoPC) 做為硬體平台。硬體平台擷取LRF、方向感測器及馬達編碼器的數據後，經由Wi-Fi將數據傳到遠端計算平台(Remoter Computation Platform, RCP)執行運算，RCP再將運算的結果經由Wi-Fi傳回硬體平台以控制機器人。LRF的數據是透過ABD(Adaptive Breakpoint Detector)及RDP(Ramer Douglas Peucker)演算法做直線的分群與最佳化，再由分群後的直線計算室內環境的特徵，再融合里程計及方向感測器，以擴展卡爾曼濾波器，來校正定位的誤差，並畫出機器人的行走路徑。由模擬結果証實EKF演算法配合里程計，所估算出的機器人姿態，比單純根據里程計估算出來的姿態準確；而實驗所得到的結果則因為估算出的方向角不夠準確，導致定位的效果只有位置座標(X,Y)比較準確，但方向角的估計則不理想，使完成的系統只能達成部分的預期目標。

關鍵字

Extended Kalman Filter ； Ramer Douglas Peucker ； Adaptive Breakpoint Detector ； System on a Programmable Chip ； Laser Range Finder

並列摘要

This thesis fuses data from odometers, a orientation sensor and a LRF (Laser Range Finder) to correct indoor position errors of a three omni-directional wheels robot by the EKF (Extended Kalman Filter) algorithm. This work uses a SoPC (System on a Programmable Chip) platform to capture the data of the LRF, the orientation sensor and the motor encoders, then transmits the data to a RCP (Remoter Computation Platform) computer via Wi-Fi. The RCP computer then calculates and sends the results to the SoPC platform via Wi-Fi to control the robot. The ABD (Adaptive Breakpoint Detector) and RDP (Ramer Douglas Peucker) algorithm segment and optimize the data points of LRF into straight line groups. The line groups are used to calculate the features of the indoor environment, and then the odometers, the orientation sensor are combined into the EKF algorithm are combined to correct the positioning error and draw the walking path of the robot.
Simulation results showed that the EKF plus odometers method outperformed the odometers only method in the robot posture estimation. However, experimental results showed that only position coordinates (X, Y) of the robot could accurately be reached, orientation angles of the robot were not accurate enough, causing the resulting system can only meet part of the anticipated goal.

並列關鍵字

Extended Kalman Filter ； Ramer Douglas Peucker ； Adaptive Breakpoint Detector ； System on a Programmable Chip ； Laser Range Finder

參考文獻

[2] Tim Bailey and Hugh Durrant-Whyte, “TUTORIAL : Simultaneous Localization and Mapping (SLAM) :Part II,” IEEE Robotics & Automation Magazine, Vol. 13, No. 2, pp. 108-117, 2006.

[7] Hyeonwoo Cho and Sang Woo Kim, “Mobile Robot Localization Using Biased Chirp-Spread-Spectrum Ranging,” IEEE Transactions on Industrial Electronics, Vol. 57, No. 8, August 2010.

[8] Paul Michael Newman, “C4B-Mobile Robotics,” http://www.robots.ox.ac.uk/~pnewman/Teaching/C4CourseResources/C4BMobileRobots.pdf.

[10] Yassen Dobrev, Sergio Flores and Martin Vossiek, “Multi-Modal Sensor Fusion for Indoor Mobile Robot Pose Estimation,” in Proceedings of the Position, Location and Navigation Symposium (PLANS), May 2016, pp. 553 - 556

[11] Luka Tesli´c, Igor Škrjanc and Gregor Klanˇcar, “EKF-Based Localization of aWheeled Mobile Robot in Structured Environments, ” Journal of Intelligent & Robotic Systems, Vol. 62, No. 2, pp. 187–203 ,May 2011.