Obstacle Avoidance Method Based on Multi-source Information Fusion Intended for Robots Operating in Different Environments
Li Xiaomin, Lixue Zhu, Zhang Rihong, Huang Baiyu, Pan Haizhang, Chen Haohan · 2019
Obstacle avoidance is a precondition for a smooth robot moving. However, current methods for obstacle avoidance are deficient in different information and sensitive to environmental conditions. In order to overcome the two mentioned shortcomings, this paper proposes an automatic obstacle avoidance method based on multi-source information fusion. A multi-source information platform is constructed based on the multi-core embedded system, and an artificial intelligence algorithm is used to construct the environment factor acquisition model. Also, a multi-source information fusion method based on fuzzy theory is proposed. Finally, a prototype test platform is constructed to conduct practical experiments on obstacle avoidance for different environments. Experimental results show that the presented method has the advantage of high obstacle identification rate.