未知环境下全覆盖路径规划问题的研究

发布时间：2018-03-29 06:13

本文选题：生物激励神经网络　切入点：模糊逻辑　出处：《南昌大学》2016年硕士论文

【摘要】：随着智能技术与社会生产、生活的不断结合,全球制造业正在向数字化、智能化、绿色化、自动化方向发展。作为智能制造的主力军,机器人由于具备了高智能性、高效工作效率和精确性、能够代替人类完成重复或危险的任务等优点,在生产生活方面得到普及。人们对机器人的灵活性及智能也提出更高的要求,其中路径规划问题就是一项重要的研究课题,而全覆盖路径规划又是其中一种较特殊又有着广泛应用前景的形式。本文在对生物激励神经网络算法进行详细分析的基础上,对其存在的问题进行改进,并且提出将模糊逻辑算法用于障碍物边界探测中,实时更新环境地图,最后通过仿真验证。本课题主要完成下面的几点工作:1.对目前常用的地图构建方法、行走方式和全覆盖路径规划算法进行研究分析、对比各自的缺点,在此基础之上,确定了本课题所选用的算法。针对所选算法的特点,确定了与之相匹配的栅格环境建模法和往复式行走方式。2.在已知的工作环境中,通过生物激励神经网络算法进行全覆盖路径规划,对该过程进行详细阐述。针对拐点处方向不确定性的问题进行分析并提出相应的改进方法,并通过仿真实验验证改进方法的有效性。3.针对未知环境中缺少全方位信息的问题,引入了模糊逻辑算法进行障碍物边界探测,详细阐明了该过程。并且将基于生物激励神经网络的全局规划与基于模糊逻辑的局部规划相融合,实现在未知环境中的全覆盖,并通过仿真,证明该思想的可行性。
[Abstract]:With the combination of intelligent technology and social production and life, the global manufacturing industry is developing towards digitalization, intelligentization, greening and automation. As the main force of intelligent manufacturing, robot has high intelligence. The high efficiency and precision of work, which can replace the human being to complete repetitive or dangerous tasks, have been popularized in production and life. People also put forward higher demands on the flexibility and intelligence of robots. The path planning problem is an important research topic, and the full coverage path planning is one of the more special and widely applied forms. This paper analyzes the biological excitation neural network algorithm in detail. The existing problems are improved, and the fuzzy logic algorithm is used in obstacle boundary detection to update the environmental map in real time. Finally, through the simulation verification. This topic mainly completes the following several work: 1.carries on the research and analysis to the current commonly used map construction method, the walking way and the full coverage path planning algorithm, compares each shortcoming, on this basis, According to the characteristics of the selected algorithm, the corresponding grid environment modeling method and reciprocating walking mode. 2. In the known working environment, The full coverage path planning is carried out through the biological excitation neural network algorithm, and the process is described in detail. The problem of the direction uncertainty at the inflection point is analyzed and the corresponding improvement method is put forward. The effectiveness of the improved method is verified by simulation experiments. 3. Aiming at the lack of omnidirectional information in unknown environment, the fuzzy logic algorithm is introduced to detect the boundary of obstacles. The process is explained in detail, and the global planning based on biological excitation neural network is combined with the local programming based on fuzzy logic to realize full coverage in unknown environment, and the feasibility of the idea is proved by simulation.
【学位授予单位】：南昌大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TP242

【参考文献】