Songsong Zhang

Information Fusion and its Application in Obstacle Avoidance of Mobile Robots

This paper presents an environment perception method using for mobile Robots based on Occupancy Grid Maps, which is based on ultrasonic probability grid map feature points extraction and matching. Low-cost ultrasonic sensors as the design scheme of distance measuring is adopted. Aiming to obtain the probability of grid map update effectively, an improved Bayesian formula is proposed. To realize synchronous positioning and map construction, dynamic random objects are related to the map with edge detection algorithm. Then the motion of the next step of robot is planned by improved particle swarm algorithm. The result of the numerical simulation shows that the novel particle swarm optimization is effective and can find the more optimal global solutions with high efficiency compared to the basic particle swarm optimization (PSO). The validity and reliability of this method are tested via the simulation. The result of the simulation shows that the method can solve the SLAM problem of robot in complex environment and realize real-time dynamic collision avoidance planning.

Information merging technology and its application in robot path planning

For solving the problems that random objects are feature point and obstacles, this paper presents an efficient approach for Path Planning of Robot based on Simultaneous Localization and Mapping (SLAM) algorithm. Low-cost ultrasonic sensors as the design scheme of distance measuring is adopted. Aiming to obtain the probability of grid map update effectively, an improved Bayesian formula is proposed. To realize synchronous positioning and map construction, dynamic random objects are related to the map with NN (nearest neighbor) data correlation method. Then the motion of the next step of robot is planned by improved particle swarm algorithm. The result of the numerical simulation shows that the novel particle swarm optimization is effective and can find the more optimal global solutions with high efficiency compared to the basic particle swarm optimization (PSO). The validity and reliability of this method are tested via the simulation. The result of the simulation shows that the method can solve the SLAM problem of robot in dynamic obstacles environment and realize real-time dynamic collision avoidance planning.