基于目标模型的遗传算法在走廊辨识中的应用  

Application of Genetic Algorithm Based on Objective Model to Corridor Recognition

作　　者：高峰[1] 黄玉美[1] 刘鸿雁[1] 见浪護[2] 

机构地区：[1]西安理工大学机械与精密仪器工程学院,陕西西安710048 [2]日本福井大学智能工学部

出　　处：《信息与控制》2007年第1期68-73,共6页Information and Control

摘　　要：为了保证移动机器人能够沿着走廊自动行走并具有一定的智能水平,工作环境的辨识就显得至关重要.同时考虑到控制的实时性,提出了一种基于目标模型的模型匹配的走廊辨识方法.利用遗传算法对未处理的灰度图像中的走廊影像进行匹配,从而使移动机器人自动地适应复杂环境,实现了模型对灯光条件变化以及包括干扰物等环境噪声的鲁棒性,并经实际走廊图像试验验证了辨识效果.It＇s very important to recognize the working environment for a mobile robot when it travels in the corridor automatically with its possible intelligence. Taking real-time control of mobile robot into consideration, a corridor recognition method with model matching based on objective model is presented. The image of the corridor in unpro- cessed gray image is matched using genetic algorithm by which the mobile robot can automatically adapt to the complex environ-ment and the proposed model with high robustness against noise such as lighting condition variations and obstacles is realized. The recognition effect is verified with real corridor images.

关 键 词：遗传算法 走廊辨识 模型匹配 

分 类 号：TP24[自动化与计算机技术—检测技术与自动化装置]