Research On Obstacle Avoidence Of Mobile Robot Based On Heterogeneous Sensor Information Fusion

With the continuous development of automation technology and sensor technology,robots are becoming more and more intelligent,and the role played in production and life is becoming more and more important.The research and application level of robot technology is an important symbol to measure the level of technological innovation and high-end manufacturing industry in a country.Therefore,the research of robot has a very important strategic significance.Autonomous obstacle avoidance of mobile robot is an important part of path planning in autonomous navigation.If we achieve precise obstacle avoidance,how to perceive and understand the surrounding environment is a difficult and hot issue.Compared with single sensor,multi-sensor,especially heterogeneous sensor,can provide complementary and integrated information,and describe the external environment relatively comprehensively,so as to improving the robot's scientific decision-making and planning,and the correctness and rapidity of response in navigation.This paper focuses on the obstacle detection based on depth camera and laser radar information fusion,and realizes the robot's real-time obstacle avoidance through fuzzy logic algorithm.First of all,for a single sensor can not accurately and completely describe the environment,the obstacle avoidance strategy of the robot is affected.This paper selects the depth camera and the laser radar as the sensors of the distance measuring system,and introduces the hardware platform and the software platform of the whole frame of the robot in detail.Then the obstacle recognition algorithm based on information fusion of heterogeneous sensors is studied.Through the combination of least square principle and fusion rule,we can fuse all sensor data after preprocessing to get the accurate location of obstacles.Finally,aiming at the problem of poor anti-interference ability and inflexible obstacle avoidance in the conventional obstacle avoidance algorithm,an obstacle avoidance algorithm based on fuzzy control is designed.The algorithm mainly uses the fuzzy logic controller to analyze and judge the surrounding environment information of the robot,and plans a path that avoids obstacles.Using Visual Studio 2012 and Matlab mixed programming,the fuzzy control algorithm was applied to the obstacle avoidance system,and the relevant parameters of the fuzzy system are obtained through the obstacle avoidance experiments in different scenes.Experimental results verify the effectiveness of sensor information fusion on obstacle recognition and the flexibility of robot obstacle avoidance.