【作者】 杨帅；

【导师】 邵磊；

【作者基本信息】 天津理工大学 ， 工程硕士（专业学位）， 2019， 硕士

【摘要】 自主移动智能机器人在军用和民用上的应用越来越广泛,定位技术是机器人运动规划中重要的组成部分。大多数机器人运动路径为二维曲线,而管内爬壁机器人运行路径为三维曲线。针对这一问题,本文提出了应用于大管径管道内壁的管内爬壁机器人定位系统。本文描述了系统的总体结构,介绍了系统运行的原理,解算出了机器人运动姿态角,消除了解算的结果中存在累计误差的问题,引入误差修正因子进一步消除误差。最后通过对姿态角进行仿真分析和机器人运动轨迹跟踪验证了定位系统的可靠性。本文首先在已有的定位方法的基础上提出了一种应用于管道内壁曲面的定位方法,该方法具有简单、精确、快速的特点。获取了陀螺仪读取的机器人运动数据,解决了机器人沿管道摆放方向直线行走这一关键问题,消除了由于机器人摆放误差和管道摆放两种人为误差。其次在获取到的机器人运动数据为基础,使用四元数法求得机器人的运动姿态阵,并使用四阶龙格库塔法求解四元数微分方程,得到了机器人运动姿态角。由于机器人定位系统中使用的陀螺仪存零漂误差,在短时间内机器人解算的姿态角会有很大误差。针对这一问题,提出了卡尔曼滤波算法、扩展卡尔曼滤波算法来消除累计误差。同时分析了机器人在运动过程中所受外力的影响,引入修正因子。改进了扩展卡尔曼滤波算法,进一步消除误差。再次介绍了系统各个模块的选型分析和电路分析,主控芯片选择飞思卡尔MK10系列单片机,使用了结合MEMS陀螺仪和加速度计的传感器模块,解决了陀螺仪和加速度计时间轴不统一的问题,使实验结果更精确。本文采用了步进电机作为履带电机驱动机器人运动,选用了LM2576和AMS1117为电源芯片,在保证稳定的前提下尽量降低了功耗。机器人与上位机通信方式为WIFI,机器人所携带WIFI模块与上位机共同连至同一WIFI网络即可实现通信,为进一步开发(例如手机APP)的组网工作提供便捷。最后介绍了管内爬壁机器人的定位系统的仿真设计过程,对经过滤波的姿态角解算结果进行仿真分析,对机器人运动时的轨迹进行轨迹跟踪,验证了定位系统的有效性。更多还原

【Abstract】 Autonomous mobile intelligent robot is widely used in military and civil applications.Localization technology is an important part of robot motion planning.The motion path of most robots is two-dimensional curve,while that of the wall-climbing robot in the pipe is three-dimensional curve.In order to solve this problem,this paper puts forward a robot positioning system for wall climbing in large diameter pipe.The overall structure of the system is described,the principle of system operation is introduced,the movement information of the robot is solved,and the robot movement attitude Angle is obtained to eliminate the problem of accumulated errors in the calculation results,and the error correction factor is introduced to further eliminate the error of the calculation results.Finally,simulation and trajectory tracking simulation are carried out to verify the reliability of the positioning system.First of all,a positioning method applied to the inner surface of pipeline based on the existing positioning methods has been proposed.It is simple,accurate and fast.The motion data of the robot read by the gyro are obtained,which solves the key problem that the robot walks in a straight line in the direction of pipeline placement and eliminates two kinds of human errors caused by robot placement error and pipeline placed error.Secondly,the robot’s motion attitude matrix by using the quaternion method has been obtained based on the obtained robot motion data.The fourth order Runge-Kutta method is used to solve the quaternion differential equation.The robot motion attitude Angle is obtained.Because the gyroscope used in the robot positioning system has zero drift error,the attitude Angle of the robot will have great error in a short time.Aiming at this problem,the kalman filter algorithm and extended kalman filter algorithm are proposed to eliminate the cumulative error.This paper also analyzes the influence of external force on the robot in the process of motion and introduces a correction factor.The extended kalman filter algorithm is improved to further eliminate errors.And then,the type selection analysis and circuit analysis of each module of the system are introduced.The main control chip is freesckel MK10 series single chip microcomputer,and the sensor module combining MEMS gyro and accelerometer is used to solve the problem of inconsistent time axis between gyro and accelerometer,so that the experimental results are more accurate.The stepper motor is used as the crawler motor to drive the robot,LM2576 and AMS1117 are selected as the power chip to reduce the power consumption as far as possible on the premise of ensuring stability.The communication mode between the robot and the host computer is WIFI.The WIFI module carried by the robot and the host computer are connected to the same WIFI network to achieve communication,which provides convenience for the networking work of further development(such as mobile APP).Finally,the simulation design process of the positioning system of the wall-climbing robot in the pipe is introduced,and the result of attitude Angle calculation after filtering is simulated and analyzed.The trajectory tracking of the robot during movement is carried out,which verifies the effectiveness of the positioning system.更多还原