温室移动机器人复合栅格地图构建方法研究.

A mobile robot based on a normal grid map could not obtain the comprehensive environmental information in a greenhouse because of the less number and limited location of carried sensors. An incorrect planning path may arise. The system adopted a way to build a composite grid map based on the comprehensive environmental information in greenhouse. Firstly, wireless sensor network regularly collected data of temperature and humidity and sent them to the robot. These data had heavy impacts on mobile robot’s passing through in greenhouse. Secondly, when the rate of change of temperature or humidity data reached the threshold, the system used the method of threshold segmentation and interpolation to establish the grid maps for temperature and humidity respectively. Finally, the system combined the grid map for temperature, the grid map for humidity and the grid map for interior material barrier in a greenhouse to build a dynamically updated composite grid map. After testing, when the system adopted the A*algorithm, the rate of success to avoid area of high temperature and high humidity based on the composite grid map at threshold of ± 10% was 2. 5 times to normal grid map. The time of completion was 1. 05 times to normal grid map. The results show that the composite grid map can improve the success rate of the path planning and the robot does not meet the problem of slow response of the robot due to the dynamic updating of the composite grid map. The real-time performance satisfies the actual requirements of the system. [ABSTRACT FROM AUTHOR]