Integration of Intelligent Systems and Sensors in Quadcopter Path Planning and Obstacle Avoidance Control Application (II)

瀏覽統計 Email 通知 RSS Feed

簡歷

基本資料

Project title

Code/計畫編號

MOST107-2221-E019-046

Translated Name/計畫中文名

智慧型系統與感測器整合於旋翼機之路徑規劃與避障控制的應用(II)

Project Coordinator/計畫主持人

Jih-Gau Juang

Funding Organization/主管機關

National Science and Technology Council

Department/Unit

Department of Communications, Navigation and Control Engineering

Website

https://www.grb.gov.tw/search/planDetail?id=12675170

Year

2018

Start date/計畫起

01-08-2018

Expected Completion/計畫迄

31-07-2019

Bugetid/研究經費

910千元

ResearchField/研究領域

電子電機工程

This is a two-year project that integrates artificial intelligence, control theory, sensors, path planning, network communications, GPS, and image processing techniques in quadcopter control system design. Quadcopter can be applied to collect disaster information and assist rescue mission. In addition, it can be used to monitor objects, remote surveillance, check mountain condition and sea wall condition. It also can be used to perform vessel check and replace the labor work and reduce cost for shipping company. In quadcopter flight control, there are two main control issues, one is the flight control and another is the mission planning. In fight control, the key problems are attitude control and position control. The attitude control focus on flight stability and tracking desired heading angle. The position control is for tracking trajectory. In mission planning, the key problems are flight mission and path planning. This project integrates evolutionary algorithm and fuzzy sliding mode control to improve chattering phenomena in control signals. In path planning, first year is for known environment and the second year is for unknown environment. The path is designed for the quadcopter tracking, but the planning time and optimal path needed to be considered in the design process. This project applies an improved A* algorithm, Dijkstra algorithm and ant colony optimization with artificial potential field to plan path and avoid obstacle for known obstacle condition and unknown condition. In GPS path planning, there are two problems need to be considered. One is the accuracy of the GPS and another is pattern recognition. We will apply a GNSS system that can reduce the error to 10 cm. In addition, vision guidance will be applied to assist path tracking control. The first year of this project is to inspect vessel and check whether there is rust or damage area or not. In this mission, we will apply a simple and fast algorithm, connected-component labeling, to recognize object pattern. The second year of this project is the rescue mission. In this mission, we will apply the thinning algorithm to skeletonize the object in order to identify the object is human. The purpose of the proposed project is to apply intelligent system to quadcopter automatic guidance and control. With the advantages of IC industry and precision mechanical industry in Taiwan, we hope that the proposed quadcopter system can have competitive power in the world market. 本計畫為2年期計畫(106/8-108/7)，結合人工智慧、控制理論、感測元件、路徑規劃、網路通訊、衛星定位、影像處理等技術，應用於監測型無人四旋翼機之系統整合與控制。監測型旋翼機除了可應用於災防情資蒐集及協助搜救外，亦可應用於目標物檢視，如遠端遙控勘查災區，定期記錄山區情況，或檢查海堤是否有損壞。另外，船舶靠港後，無人飛行載具可執行船舶外體檢視，取代人力檢視，可有效降低船公司人力成本及人員垂吊船外之風險，提供船舶港灣管理附加價值。無人旋翼機的控制主要分為兩個層次：飛行控制及任務規劃。飛行控制主要是姿態控制及位置控制，其中姿態控制是針對飛行穩定及追蹤期望頭向角，位置控制則是針對軌跡追隨；任務規劃主要是定義飛行任務及設定飛行路徑的規劃。本計畫擬結合演化計算及模糊滑動模式控制，以改善控制命令chattering現象。在路徑規劃方面，第一年(106/8-107/7)是針對已知環境，第二年(107/8-108/7)是針對未知環境，主要是將路徑規畫出來供旋翼機追跡之用，其中要考量的是規劃所花的時間及是否能找出最佳化路徑。在本計畫中，將分別以改良型A*演算法與Dijkstra演算法及螞蟻演算法結合人工勢能場，分別針對已知障礙物(第一年計畫)及未知障礙物(第二年計畫)，進行路徑規劃及避障控制，期望設計出計算時間短且避障效果佳的路徑規劃方法。在GPS路徑規劃上，有2個問題值得探討，一個是衛星定位的精確度，一個是影像辨識。本計畫擬使用本系自行研發之GNSS系統，使定位誤差減至10公分內。另外，亦將以視覺導引輔助路徑追隨控制，故影像處理亦是研究重點，第一年計畫的初期標的物是船體外部的鏽蝕或破損，本計畫擬先以較簡單且計算快速的Connected-component Labeling演算法進行圖形識別，在第二年計畫的搜救受難人員任務中，則以簡易的Thinning演算法進行骨架識別程序處理。本計畫的重點即為利用現有之智慧型系統的研究，應用於四旋翼機自動導引控制的實現，以台灣在IC電子製造業及精密機械加工業的優勢，期望將來能整合出具有市場競爭力之四旋翼機系統。

Keyword(s)

避障控制
路徑規劃
影像處理
演化理論
人工勢能場
衛星定位系統
自動導航
collision avoidance control
path planning
image process
evolutionary algorithm
artificial repulsive force
GPS
automatic guidance

DSpace CRIS

Integration of Intelligent Systems and Sensors in Quadcopter Path Planning and Obstacle Avoidance Control Application (II)

基本資料

Description