【作者】 吴轩；

【导师】 葛召炎；

【作者基本信息】 湖南大学 ， 控制科学与工程， 2011， 硕士

【摘要】 随着社会经济的迅速发展,交通运输量的不断上升,桥梁负荷量迅速增大,桥梁坍塌事故有增多的趋势。实践表明,桥梁的大部分缺陷产生在桥梁的底面,目前,普遍采用的桥底缺陷探测方法是:采用传统桥梁检测车工作臂将检查人员送至桥底进行检测作业,该方法效率低,检测质量差且不安全;因此,最近国内外学者开始进行了智能桥梁检测车的研究,试图实现自主检测桥底缺陷,以促进桥梁检测技术的发展。本文在现有研究成果的基础上,重点研究了桥梁检测车工作臂的定位与避障。这是实现自动检测作业的关键技术;工作臂在精确地自身定位的同时,承载摄像机沿桥底进行仿形运动,有效地回避桥底结构障碍、并始终和桥梁底面保持一定的距离,本文的研究有利于解决桥底检测难的问题,本文主要有以下内容:首先,从总体上设计了工作臂定位与避障系统,阐述了其工作原理,并说明了工作臂的运动过程。其次,在对多种传感器分类和归纳的基础上,设计了有探测外部环境和工作臂自身定位功能的感知子系统,采用最小值法实现信息融合,提高其精确性和处理速度。再次,研究了桥梁检测车的定位和工作臂的定位。根据能否接收GPS定位信号,分别为车辆定位建立了两种不同的参考坐标系,车辆定位子系统采用了基于模糊自适应扩展卡尔曼滤波的GPS/DR组合定位方式,在Matlab中进行了仿真试验,试验结果表明该方法是有效的。然后,分析了机械臂的坐标转换关系,研究了工作臂的位姿定位问题,该定位系统的设计为自动检测桥底提供了重要的前提条件。最后,根据工作臂的特征,提出了探测式的仿形避障方案;采用基于神经网络的电液避障方法,并结合桥梁底面结构来开展仿形运动,通过仿真试验,证明了该方法的可行性和优越性。更多还原

【Abstract】 As the rapid economic development ,the continuous rise in traffic load of bridge augment rapidly, and bridge collapse accidences is also increasing. Practice has proved that most of bridge defects appeared in the bridge-bottom.At present, the traditional method is sending inspectors to the bridge bottom by working arm to carry on manual test operations. The method have several disadvantages that are low efficiency, that are low effiency, poor detection quality and unsafety. Therefore, to study and design a kind of intelligent bridge-detection vehicle to detect the bridge-bottom defects independently, to make progress in bridge detection technology.Based on the existed fruits,the research in localization and obstacle-avoidance control of the bridge-detection vehicle’working arm is carried in this paper.This is a key technology for achieving the automatic detection fution. The working arm does self-localization accurately,while carrying the camera load in the working arm along the bridge bottom to do imitative movement.To effectively avoiding structural obstacles of the bridge-bottom, The working arms always keep a certain distance away from the bridge-bottom, that is tosay, the difficult problem of the bridge-bottom detection is solved.Firstly, the working arm’localization and the obstacle-avoidance control system is designed for overall, their system working principles are elaborate and the movement process of working arm is described in detail.Secondly, based on the classification and induction of varieties of sensors, a sub-detection system with the fuction of detecting external environment and self-localization of the working arm is designed. The method of the minimum numeric to achive the aim of information’fusion in this paper,to improve their accuracy and processing speed.Thirdly, the localization of the bridge-detection vehicle and the working arm.According to receiving GPS locational signals or not, two different reference coordinate system are established.Integrated localization GPS/DR based on fuzzy-adapted Kalman filtering is used in the Vehicle localization subsystem,and a lot of simulation in Matlab are tested.The results show that the method is effective. Then, the relation of the mechanical arms’coordinate transformation is analyzed ,the posture localization of the working arm is explored.The design of whole localization system provide important pre-condition for the automatical detection of the bridge-bottom defects.Finally,according to the characteristics of the working arm, a tracing and imitative movement plan of obstacle-avoidance is designed.And imitative movement is carried on which combined with the bridge-bottom’structure.Also,a method of the electron-hydraulic control based on neutral net is lodged.Through some simulation experiments,the method is proved the feasibility and superiority.更多还原