室内AGV的同时定位与建图方法研究

发布时间：2018-01-13 11:02

本文关键词：室内AGV的同时定位与建图方法研究　出处：《青岛科技大学》2015年硕士论文　论文类型：学位论文

【摘要】：自动导航搬运车(Automated Guided Vehicle, AGV)也被称作搬运机器人,是现代智能物流系统中的重要环节,然而目前主流的AGV导引技术并不能使其在工作中从真正意义上实现完全自主。同时定位与建图(Simultaneous Localization and Mapping, SLAM)方法是最早由Smith Self和Cheese man提出的解决移动机器人导航关键问题的方法,由于其巨大的理论研究和实际应用价值在被提出以来的几十年间发展迅速,已在室外、室内、水下和航空等多种环境得到了成功应用。使用SLAM思想解决AGV导航问题,保证了AGV的安全导航和精确定位,也是其高效规划路径、决定控制决策的基础。本文对室内AGV的同时定位与建图方法进行了研究,首先简要介绍了AGV目前的常用导航方式,并分别论述了它们存在的优势和缺陷；然后对本文引入到AGV导航领域的同时定位与建图算法进行了详细分析；通过一系列模型建立,搭建了用于AGV的SLAM方案的基础；最后重点研究了目前的各种SLAM算法,包括基于非线性系统的扩展卡尔曼滤波(EKF)算法、基于概率的粒子滤波(PF)算法,基于RBPF的FastSLAM算法,并针对传统方法的不足,提出了一种融合改进粒子群优化算法的粒子滤波方法。该方法在是基于Rao-Blackwellized(RBPF)的FastSLAM算法基础上,引入粒子群优化算法更新AGV预测位姿,考虑了个体粒子和群体粒子的最优解对结果的影响,并引入遗传算法中的变异操作,使用变异后的新粒子替代原粒子集中权值较小的粒子,对求得的粒子集进一步优化调整。仿真结果表明,新算法有效地提高了AGV的定位精度,保持了粒子的多样性并提高了算法的全局搜索能力。最后一章对全文进行了总结,并且对今后进一步的研究方向进行了展望。
[Abstract]:Automated Guided vehicle (AGV) is also called the moving robot, which is an important part of modern intelligent logistics system. However, the current mainstream AGV guidance technology can not make it completely autonomous in the real sense of its work. At the same time, positioning and mapping (. Simultaneous Localization and Mapping. Slam) method is the first method proposed by Smith Self and Cheese man to solve the key problem of mobile robot navigation. Because of its great theoretical research and practical application value in the decades since it was put forward, it has developed rapidly in the outdoor, indoor. Many environments such as underwater and aviation have been successfully applied. Using SLAM to solve the problem of AGV navigation ensures the safe navigation and accurate positioning of AGV and is also its efficient planning path. In this paper, the simultaneous positioning and mapping methods of indoor AGV are studied. Firstly, the common navigation methods of AGV are briefly introduced. Their advantages and disadvantages are discussed respectively. Then, the algorithm of simultaneous location and mapping in the field of AGV navigation is analyzed in detail. The foundation of SLAM scheme for AGV is built through a series of models. Finally, we focus on various SLAM algorithms, including extended Kalman filter (EKF) algorithm based on nonlinear system and particle filter algorithm based on probability. The FastSLAM algorithm based on RBPF, and aiming at the shortcomings of the traditional method. In this paper, a particle filter method based on Rao-Blackwelle lizedRBP (RBP) is proposed, which integrates the improved particle swarm optimization (PSO) algorithm. The particle swarm optimization (PSO) algorithm is introduced to update the AGV prediction position, and the influence of the optimal solution of the individual particle and the swarm particle on the result is considered, and the mutation operation in the genetic algorithm is introduced. The new particle after mutation is used to replace the particle with small weight in the original particle set, and the particle set is further optimized and adjusted. The simulation results show that the new algorithm can effectively improve the positioning accuracy of AGV. The diversity of particles is maintained and the global search ability of the algorithm is improved. In the last chapter, the full text is summarized, and the future research directions are prospected.
【学位授予单位】：青岛科技大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TP242

【相似文献】