Design Of Autonomous Navigation System For Unmanned Vehicle Based On Multi-Sensor

With the rapid development of mobile robotics,autonomous vehicle is becoming a hotspot of robotics research in the field.After decades of experimentation,with faster computers,reliable hardware sensors and a new generation of artificial intelligence technology,vehicles can acquire a human-like ability to drive independently and safely in a complex and changing environment.In this topic,a multi-sensor unmanned vehicle platform based on GPS and four-layer LiDAR is designed,and autonomous navigation technology of unmanned vehicles is studied.Firstly,based on GPS,ultrasonic sensors and IBEO four-layer LiDAR and other sensors,a multi-sensor autonomous navigation system for unmanned vehicles is designed,and the method of multi-sensor selection,connection,system integration is introduced,as well as the use of RS232 communication protocol to compile the communication software of the upper and lower bit machines,send control commands to achieve the control of the unmanned vehicles at the PC.Secondly,autonomous navigation technology of unmanned vehicles is studied.Based on the GPS output information obtained by serial communication,the unmanned vehicle is positioned,and the global navigation of the unmanned vehicle is realized by the navigation algorithm of Baidu Map API based on the latitude and longitude information.In the dynamic navigation operation of the unmanned vehicle,the obstacle detection information is obtained through the four-layer LiDAR,and the obstacle avoidance algorithm is designed to realize the obstacle avoidance function;in the rear of the unmanned vehicle,ultrasonic sensors are used to detect the blind area of the four-wire LiDAR to assist the unmanned vehicle to avoid obstacles.Finally,based on the analysis of the characteristics of four-layer LiDAR,the method of obtaining point cloud data of four-layer LiDAR according to TCP/IP Ethernet communication protocol is introduced;the method of plotting point cloud data information using OpenGL library,combined with the SDK of IBEO company to identify some of the road obstacle information around unmanned vehicles is introduced.The effectiveness of the system design was verified through experimental testing.The multi-sensor autonomous navigation system introduced in this article is characterized by the use of four-layer LiDAR combined with GPS to achieve the autonomous navigation design of the unmanned vehicle,the system through the four-wire LiDAR can not only accurately detect the presence or absence of road obstacles,but more importantly,can identify specific types of obstacles,which lays a foundation for further improving the intelligent level of unmanned vehicle navigation control.