【作者】 张锐；

【导师】 储昭碧；

【作者基本信息】 合肥工业大学 ， 控制科学与工程， 2019， 硕士

【摘要】 塔式起重机在作业过程中存在视觉盲区、操作误差、对操作员的要求高等特点及隐患,同时在负载吊动过程中会产生很多摆动。如何排除这些隐患、降低塔机吊物的摆动等问题逐渐的被大众所关注。塔式起重机的控制日益复杂化,目前的控制系统和方式渐渐地难以满足要求,因此,探索一套集解决视觉盲区、定点与防摆控制于一体的控制策略,对于该方向的发展有着实质上的意义。首先,本文开篇陈述了该课题的研究背景、课题的来源以及研究该课题的意义,详细了阐述塔机控制的常用方法与工业视觉理论的现在情况以及未来的发展。针对了不同类型的控制策略,阐述了其控制原理,介绍了目前控制领域对于塔机控制所取得的控制效果与存在的问题。其次,建立了塔式起重机的空间坐标系,利用深度学习的YOLO算法,实现了对塔式起重机吊钩的识别.通过标定、矫正、立体匹配等过程,获取吊钩的深度信息,以得出吊物在此坐标系中的具体位置。第三,分析了塔式起重机作为典型欠驱动模型的模型特点,利用拉格朗日方法,给出了建模的详细过程与推导。抽象出了塔式起重机的物理学模型,并最终得到了用于实际控制的控制模型。对于得到的控制模型,针对性地采用了滑模控制方法,将吊物过程分为三个过程,分过程地进行控制。此外还在Simulink对该被控对象的模型结合控制算法进行了仿真,验证了该方法可以在较短时间内实现塔式起重机的定点控制和防摆。第四,根据实际被控对象塔机的功能要求,提出了一种基于DSP控制器的控制结构,并给出了整体框图与硬件整体框图。给出了DSP最小系统、JTAG接口电路、485通信模块、光电耦合电路、放大电路、报警电路等硬件电路设计,并给出了滑模控制算法软件实现,Modbus通讯实现等。最后,对未来本课题的研究与发展进行了展望。更多还原

【Abstract】 Tower crane in the process of operation there are visual blind areas,operating errors,high requirements for the operator characteristics and hidden dangers,at the same time in the process of load lifting will produce a lot of swing.How to eliminate these hidden dangers,reduce the swing of the tower crane and other issues have gradually been concerned by the public.The control of tower crane is becoming more and more complicated,and the current control system and mode are gradually unable to meet the requirements.Therefore,exploring a set of control strategy that integrates solving visual blind area,fixed-point control and anti-pendulum control is of great significance for the development of this direction.First of all,this dissertation begins with a statement of the research background of the subject,the source of the subject and the significance of the study of the subject,a detailed description of the tower crane control methods and industrial vision theory of the current situation and future development.Aiming at different types of control strategies,the control principle is expounded,and the control effect and existing problems of tower crane control in current control field are introduced.Secondly,the spatial coordinate system of the tower crane is established,and the deep learning YOLO algorithm is used to realize the identification of the hook of the tower crane.Thirdly,the characteristics of tower crane as a typical under-actuated model are analyzed,and the detailed modeling process and derivation are given by using Lagrange method.The physical model of tower crane is abstracted and the control model for actual control is obtained.For the control model,the sliding mode control method is adopted,and the lifting process is divided into three processes.In addition,Simulink simulates the model combined with the control algorithm of the controlled object,which verifies that the method can realize fixed-point control and anti-pendulum of tower crane in a relatively short time.Fourth,according to the functional requirements of the actual controlled object tower crane,a control structure based on DSP controller is proposed,and the overall block diagram and the overall hardware block diagram are given.The hardware circuit design of DSP minimum system,JTAG interface circuit,485 communication module,photoelectric coupling circuit,amplifying circuit,alarm circuit and so on is given.Lastly,the future research and development of this subject are prospected.更多还原