查詢結果分析
相關文獻
- 以實驗設計方法求解競賽輪型機器人最佳化設計之研究
- Distributed Broadcasting Algorithms in Rotator Graphs
- Machining Parameters Selection for Stock Removal Turning in Process Plans Using a Float Encoding Genetic Algorithm
- A Modified Multistart Method for Multimodal Optimization
- 線性軸幅路網接駁系統最適整合區位、路線與排班模式之研究
- Bayesian Estimation for the Optimum in Single Factor Quadratic Regression
- 衛星探空資料(SATEM)在中央氣象局有限區域預報系統之客觀分析模組的應用
- 二次元靜態連續體結構之最佳化:應用族群概念之基因演算法
- Optimization of Order-packing Travel Route in a Finished Goods Warehouse
- 田口法應用於最佳化設計問題
頁籤選單縮合
題名 | 以實驗設計方法求解競賽輪型機器人最佳化設計之研究=A Research on Solving the Optimization Design for a Wheeled Contest Robot by Applying Experimental Design Methods |
---|---|
作者姓名(中文) | 林邦傑; 李政鋼; 黃國豐; 羅崑原; | 書刊名 | 全球商業經營管理學報 |
卷期 | 8 2016.09[民105.09] |
頁次 | 頁189-198 |
分類號 | 448.94 |
關鍵詞 | 輪型競賽機器人; 自走車; 避障競賽; 最佳化; 均勻實驗設計法; 拉丁超立方抽樣法; Wheeled contest robot; Self-propelled vehicle; Obstacle avoidance contest; Optimization; Uniform experimental design method; Latin hypercube sampling method; |
語文 | 中文(Chinese) |
中文摘要 | 智慧工廠與工業 4.0 的概念,已成為全球產業界的重要議題,智慧機器人更是未來工程人才必備之核心能力之一。服務型機器人(Service Robots),有別於工業機器人侷限於工廠內,其種類眾多且應用範圍廣泛,更需具備對環境的感測,以自行決定行動的智慧化功能,所以又稱為「智慧型機器人」。智慧型機器人依外形可區分輪型機器人、雙足機器人、多足機器人(如機器狗、六足機器人)、飛行機器人等多種造型。本研究旨在設計一部Speed Car 參加自走車避障競賽項目並求取Speed Car 的程式控制參數之最佳解,以期能在自走車避障競賽項目取得較佳的競賽成績。Speed car 自走車避障主要依據5 個超音波感測器的回饋數值,以程式控制5 個超音波及2 馬達,控制參數包含左側近、左側遠、左邊、中間、右邊、右側近、右側遠、左轉、直走、右轉共10 項參數,透過控制10 項參數調整進行避障。為了搜尋控制參數之最佳化數據,本研究應用之研究方法為均勻實驗設計法及拉丁超立方抽樣法。首先,以均勻實驗設計法規劃十因子的均勻實驗,然後以模擬場地進行實驗並記錄成果。其次以拉丁超立方抽樣,抽出100 組實驗,再次以模擬場地進行實驗並記錄實驗成果。兩次實驗中由拉丁超立方抽樣,轉換成的程式參數有較好的表現,因此對上下限進行調整,再以拉丁超立方抽樣做第二次的抽樣,抽出100 組實驗,以模擬場地進行實驗,求得可行解4 組。最後以參加競賽,運用實驗求得的4 組解及1 組的人工思維經驗解,運用其競賽。本研究顯示,透過系統化的抽樣方法能找到控制參數的優良解,將此優良解運用於競賽中可大幅縮短競賽作品賽前調整的時間,可以在練習時有更快、更容易對不同的競賽場地,找到最合適的參數進行競賽,並得取較佳的競賽成績。 |
英文摘要 | Smart factory and industry 4.0 have become an important issue in the global industry. Knowing how to manipulate smart robot is one of the core competencies for future engineering talented persons. Unlike industrial robots operated in factories, service robots have many kinds and wide range of applications. Service robots have to sense environment to determine their own intelligence capabilities of action. Service robots are also known "smart robot." According to their shapes, smart robots can be classified as wheel robot, two-legged robot and multi-legged robots such as robotic dog, six-legged robot, and flying robot, etc. This study aims to design a high grade Speed Car to participate in a self-propelled vehicle obstacle avoidance contest. Speed car, controlled by program, has 5 ultrasonic sensors and 2 motors. Speed car can avoid obstacles according to 10 control parameters: left near side, left far side, left side, middle region, right side, right near side, right far side, left turn, going straight, and right turn. To solve the optimal solutions for control parameters, this study applies uniform experimental design method and Latin hypercube sampling method. Firstly, uniform design is applied to design uniform experiments with 10 factors. Next, Latin hypercube sampling method is applied to plan 100 experiments. The results of all experiments are obtained by using the simulated contest filed. Finally, 4 feasible solutions can be obtained. Besides, one more experienced solution is also considered. Therefore, there are 5 solutions applied in the real contest. This study shows that the application of systematic sampling methods can get the good solutions for control parameters. Applying the good solutions for parameters in real contest can significantly shorten the adjustment time and get better race results. |
本系統之摘要資訊系依該期刊論文摘要之資訊為主。