空气环境下移动机器人羽流追踪方法研究

发布时间：2018-10-11 15:20

【摘要】：受自然界部分生物通过追踪散布在空气中的化学信息来搜寻配偶、觅食、躲避敌害等行为的启发,人类开始探索给机器人搭配化学烟雾传感器等模块来模拟生物追踪化学羽流的行为。希望此种机器人能够代替人类在人为灾害、自然灾害环境中完成救援与救灾任务。此外,随着全球化的推进,各国对社会安全和国家安全都安全事业关注度越来越高,因此研发能够代替人力执行搜救生命、搜寻违禁品等任务的智能机器人,具有重大的实际意义。现阶段化学羽流追踪(Chemical Plume Tracking,简写CPT)机器人成为国内外研究学者关注的重点,相关的研究机构已经取得了突出的成就。目前当CPT机器人在复杂多变的环境中执行羽流追踪任务时,伴随客观环境的复杂程度提高、羽流信息的分布更加复杂,以至于羽流追踪难度增加。所以在满足精确性的要求下,CPT机器人要完成复杂空气环境中羽流追踪任务,需要具有更强的行为决策能力、自主导航能力。因此本论文针对机器人在羽流追踪过程中的自主导航性能、速度控制精度及对机器人行为建模三个方面进行研究,其研究工作分为四部分:(1)采用模糊Petri网对CPT机器人在整个羽流追踪过程中的行为进行建模,并分析其行为的可行性与正确性。(2)为提高移动机器人的自主导航能力,根据羽流追踪过程的实际情况对传统A*算法进行修正,然后采取修正后的A*算法与机器人传感器结合循环导航的方法,用以完成整个追踪路径的导航。(3)羽流追踪过程中驱动轮电机的速度控制,会影响着CPT机器人执行羽流追踪任务的精准性。因此本文选择模糊RBF神经网络PID离线学习的方式实现对机器人驱动轮电机进行速度控制,并达到响应速度快、调整时间短的实际效果。(4)搭建羽流追踪机器人,以AS-4WD移动机器人平台为基础,与烟雾浓度传感器,超声波传感器等模块搭配构成化学羽流追踪机器人,对本文模型与方法进行验证。
[Abstract]:Inspired by nature's behavior of tracing chemical information scattered in the air to search for spouses, forage, and avoid enemies, Humans have begun to explore how robots can be fitted with modules such as chemical smoke sensors to simulate the behavior of biological tracing chemical plumes. It is hoped that this kind of robot can replace human in man-made disasters and natural disasters to complete rescue and relief missions. In addition, with the development of globalization, countries are paying more and more attention to social security and national security, so they are developing intelligent robots that can take the place of human resources to carry out search and rescue work, search for contraband and other tasks. It is of great practical significance. At present, chemical plume tracing (Chemical Plume Tracking, (CPT) robot has become the focus of researchers at home and abroad, and relevant research institutions have made outstanding achievements. At present, when CPT robot executes plume tracking task in complex and changeable environment, with the increasing complexity of objective environment, the distribution of plume information becomes more complex, which makes plume tracking more difficult. So in order to fulfill the plume tracking task in complex air environment, CPT robot needs stronger behavior decision ability and autonomous navigation ability to meet the requirement of accuracy. In this paper, the performance of autonomous navigation, the accuracy of speed control and the modeling of robot behavior in plume tracking are studied in this paper. The research work is divided into four parts: (1) using fuzzy Petri nets to model the behavior of CPT robot in the whole plume tracking process, and analyzing the feasibility and correctness of its behavior. (2) to improve the autonomous navigation ability of mobile robot, According to the actual situation of plume tracking process, the traditional A * algorithm is modified, and then the modified A * algorithm is combined with the robot sensor in the circular navigation method. (3) the speed control of driving wheel motor in plume tracking process will affect the accuracy of plume tracking task of CPT robot. Therefore, this paper chooses the fuzzy RBF neural network PID off-line learning method to realize the speed control of robot drive wheel motor, and achieves the practical effect of fast response speed and short adjustment time. (4) build plume tracking robot. Based on AS-4WD mobile robot platform, chemical plume tracking robot is constructed with smoke concentration sensor, ultrasonic sensor and other modules. The model and method in this paper are verified.
【学位授予单位】：新疆大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TP242

【参考文献】