Reaearch Of Mobile Robot's Environmental Percepon Based On Multi-Sensor Information Fusion

With the fast enlargement of mobile robot's application field,their work environment will be more and more complicated. The environmental perception system of mobile robot take charge for the obtaining of inner state and exterior environment information, which is the basic of mobile robot's decision-making. Many researchers in and abroad are committing themselves to the study of mobile robot's environmental perception based on multi-sensor information fusion. Some sensors have been put on the mobile robots, obtain a consistent environment description by multi-sensor information fusion. With the development of the technology of computer and sensors, the research will be more and more embedded and much more fruits will come out.Supported by the Hebei Natural Science Foundation (HBNSF) "Target identification and location of mobile manipulator for public safety"(NO. E2006000030), This paper makes embedded systemic research about the environmental perception of mobile robot based on multi-sensor information fusion, The object of research is HEBUT-Ⅱmobile robot developed by our lab. What follows are the main contents of this paper:Firstly, the Hardware architecture and the bodywork configuration of HEBUT-Ⅱmobile robot are introduced, the driving system and the kinematics characteristic are also analyzed. And then, the environmental perception tactic of HEBUT-Ⅱmobile robot is made,which is based on the summarization of the sensors on the robot.Secondly, the modeling of road-tracking camera and the conversion between image coordinate and vehicle coordinate is completed. A method of image segmentation based on HSI color threshold is used for mobile robot's road image segmentation. The image segmentation method based on HSI color threshold has many advantages: strong anti-noise ability, using more information during image processing, perfect segmentation result without being affected by illuminative condition, and simple arithmetic. All these advantages improve the segmentation speed and veracity of complicated road image and make the road information revert from the image space. Then,A fuzzy controller fit for HEBUT-Ⅱmobile robot has been designed based the process of fuzzy controller designing. The road recognition and tracking experiment has proved the feasibility of above way.Thirdly,the multi-sensor information fusion ways are presented. After comparing their characters, information fusion based on fuzzy neural network is applied in avoiding obstacles of HEBUT-Ⅱmobile robot to fuse the information of CCD cameras and ultrasonic sensors. According to result of experiment, this network is valid and the robot can move well in avoiding obstacles.