The Ceiling Projector Based Robot Accessory Navigation System In Intelligent Space

Because the home environment has the characteristic of uncertainty, the realization of fast navigation in the complex home environment is important for home service robot. This thesis concentrates on the study of using accessory navigation system which is based on ceiling projector to realize home service robot's faster navigation in home environment.In this thesis, A ceiling projector based accessory navigation system was designed, which include the ceiling projector's path planning, communication with upper computer and motion control algorithm. AS-RⅡrobot was used in the related experiments to realize track-following navigation task. The corresponding works are summarized as follows:(1) First of all, we studied the working principle and working steps for the ceiling projector based accessory navigation system, the design principles was proposed according to the required performance of the system, and the selection for parts of hardware which were used in system was done.(2) In this section particle swarm optimization(PSO) algorithm based path planning was studied. Firstly, according to the grid map of indoor environment, we got danger degree map. Then we used uniform distribution of particles based and non-uniform distribution of particles based PSO path planning algorithm to do correlative experiments separately, the experiments proved that using non-uniform distribution of particles based PSO path planning algorithm, we could get smoother and safer path. After that, using coordinate transformation, the path could be converted from image coordinates to world coordinates and then be stored in the intelligent space server.(3) In this section we studied wireless communication, motion control algorithm. Firstly, we established kinematics model of the projector, based on this model we converted the path information in server to angle displacement of the projector, then through ZigBee wireless network in the intelligent space we transmitted it to the projector as control input, finally the path which was used to navigate the mobile robot could be projected to the ground.(4) In this section we designed experiments about track-following navigation. Firstly, through color threshold segmentation and fast positioning, robot identified laser point which was projected to the ground by projector, then tracked the moving laser point, in the tracking process, Psd and Sonar sensor were used to ensure real-time obstacle avoidance. At last, based on experimental results, we summarized the shortages of this system and gave corresponding improvements.