整合基因演算與模糊控制法於 自走式機器人之路徑規劃

Path Planning of An Autonomous Mobile Robot by Integrating Genetic Algorithm and Fuzzy Logic Control

王興仁(Xing-Ren Wang)

指導教授 ： 許政行 ； 賴玲瑩

中原大學/工學院/機械工程研究所/碩士(2003年)

https://doi.org/10.6840/cycu200300020

摘要

本論文探討如何規劃自走式機器人的路徑，讓機器人能夠避開環境中的障礙物並到達指定的目地。首先，分析自走式機器人的運動方程式，其次，依環境中障礙物個數的不同分析其避障行為模式，利用模糊控制的觀念，提出一避障路徑規畫演算法；並以目標與障礙物之間的夾角及自走式機器人與障礙物間的距離作為模糊控制器的輸入來修正自走式機器人行進的方向。在控制器最佳化部分，以基因演算法進行參數搜尋的工作，將適應函數的設定為路徑總長，同時將5x5模糊規則庫、三角型歸屬函數之底端參數與尺度因子納入搜尋；針對不同的障礙物環境，個別得到最佳避障模糊控制器。並由套裝軟體Matlab模擬。模擬顯示，在較複雜環境下設計的最佳避障模糊控制器，亦可適用於較簡單環境下之避障規劃。

關鍵字

適應函數 ； 路徑規劃 ； 最佳避障模糊控制器 ； 基因演算法 ； 模糊控制 ； 自走式機器人

並列摘要

The present study concerned about how to guide an autonomous mobile robot (AMR) moving in obstructive environments to avoid obstacles and reach the goal. First, the dynamic equations of the AMR were analyzed. Next, according to the number of obstacles the avoiding behavior were studied, and a path-planning algorithm based on fuzzy control was also proposed. A fuzzy controller was used to modify the moving direction of the AMR. The angle between the obstacle and the goal, and the distance between the obstacle and the AMR were inputs of the controller. A genetic algorithm was used for optimization searching of parameters in design of the controller. The searching parameters included the 5x5 consequent variables of the control rule table, the vertexes of the triangular-shaped membership functions and scaling factors. The fitnessfunction was set as the total traveling length. Optimal controllers were found for various obstructive environments through Matlab simulations.The simulation results showed that the optimal controller obtained for the most complex environment was also fit for the simpler ones.

並列關鍵字

Path-Planning ； Fuzzy Control ； Genetic Algorithm ； Autonomous Mobile Robot ； Fitness Function ； Optimal Controller

參考文獻

[1] S. Hayati, "The Rocky 7 Rover：A mars sciencecraft prototype," IEEE International Conference on Robotics and Automation,Vol.3 pp.2458-2464,1997.

[3] C.T.C. Arsene , A.M.S. Zalzala ,”Control of autonomous robots usingfuzzy logic controllers tuned by genetic algorithms ,” Proceedings of

Evolutionary Computation , Vol. 1,pp.428-435,1999.

[5] T.H.S. Li , Ming-Sheng Chiang and Sheng-Sung Jian,” Motion planning of an autonomous mobile robot by integrating GAs and fuzzy logic control

,”The Ninth IEEE International Conference of Fuzzy Systems,vol.2,pp. 933 –936,2000.