Research on Automatic Delivery Robot Control Based on Multi-sensor Fusion

Abstract

With the explosive growth of E-commerce markets such as online shopping consumption and live delivery, various problems in the express industry have seriously affected the good experience of users to some extent, especially the problem of the last kilometer is becoming increasingly prominent. Especially in the peak period of e-commerce package volume such as “double 11” and “6.18”, the problem of manpower shortage in the express industry is more serious, resulting in the lag of "last kilometer" distribution from distribution outlets to customers. Aiming at the problems of "last mile" distribution lag and high logistics cost in China's e-commerce industry, this paper takes the automatic delivery robot that completes the express terminal distribution task in the university campus as the research object, studies the robot control scheme, multi-sensor information fusion algorithm and obstacle detection method, and puts forward the multi-sensor information fusion technology using adaptive weighted fusion algorithm, The problem of one sidedness of single sensor data is effectively solved, and the static and dynamic obstacle recognition can be realized more accurately and reliably. The research results lay a foundation for the obstacle avoidance research of delivery robot in the future.