一种基于逆向求解的双机吊装动作规划研究

发布时间：2018-09-07 13:23

【摘要】：由于大型工程的发展促进吊装行业走向复杂化、集成化,两台起重机协同吊装(双机吊装系统)趋向普遍化,同时多台起重机协同吊装(多机吊装系统)逐步常见,这也使得保证吊装安全性成为首要问题。吊装方案的制作是保证吊装安全的重要保障,而合理高效的起重机吊装动作规划对吊装方案的制作有着重要指导作用和现实意义。目前,对双机吊装系统动作规划的研究,存在规划效率较低、动作序列不优、建模误差较大等问题,而对多机吊装系统动作规划研究较少。因此,本文针对双机吊装系统中封闭链约束及非完整运动学约束,提出基于被吊物及起重机下车驱动的系统建模方法,采用改进的RRT-Connect++动作规划算法,通过三个涉及起重机行走、转弯、变幅、起升等动作的案例来验证该求解思路的可行性和规划的高效性。对于多机吊装系统,采用相同的建模思路和规划算法,通过对四台起重机协同吊装管架被吊物的案例进一步验证该求解思路对多机吊装系统的适用性及规划的高效性。文中首先通过分析双机吊装动作规划的重要性,和对双机吊装建模及其动作规划的研究现状总结,确定双机吊装动作规划的难点在于系统的封闭链约束及非完整运动学约束问题,并论述本课题研究意义。其次,对于系统的封闭链约束,结合目前研究现状,提出基于被吊物及起重机下车驱动的建模方法,同时给出逆向方程求解过程；对于系统非完整运动学约束,根据履带与轮式移动小车相似性建立起重机下车模型,进而得到求解方程。接着,根据双机吊装系统建模定义动作规划的位姿空间、动作输入、位姿转换矩阵、距离计算等,同时论述RRT-Connect++算法流程及对其改进,并通过对双机翻转被吊物、双机行走越障平移被吊物、双机转弯避障平移被吊物三个案例动作规划,验证本文求解思路可行性、规划高效性。最后,根据对双机吊装系统动作规划的研究,尝试研究多机吊装系统动作规划。文中具体分析四台起重机协同吊装管架被吊物的建模,包括该系统中多重约束的处理、管架模型位姿定义、吊点计算等。在定义系统位姿空间及动作输入后,通过对四台起重机起升、行走平移、翻转管架案例的动作规划,说明本文求解思路对多机吊装系统具有适用性,同样在动作规划中具有较高效率并能规划较优的动作序列。
[Abstract]:Because of the development of large-scale engineering, the hoisting industry is becoming more and more complicated and integrated, so the cooperative hoisting system of two cranes (double crane hoisting system) tends to be generalized, while the cooperative hoisting system of multiple cranes (multi-machine hoisting system) is becoming more and more common. This also makes the safety of hoisting become the top issue. The production of hoisting scheme is an important guarantee to ensure the safety of hoisting, and reasonable and efficient crane hoisting action planning has an important guiding role and practical significance in the production of hoisting scheme. At present, there are some problems in the research of action planning of double hoisting system, such as low planning efficiency, poor action sequence and large modeling error, but less research on action planning of multi-machine hoisting system. Therefore, aiming at the closed chain constraint and nonholonomic kinematics constraint in the double crane hoisting system, this paper presents a system modeling method based on the lifting object and the crane driver, and adopts the improved RRT-Connect motion planning algorithm. The feasibility of the solution and the efficiency of the planning are verified by three cases involving crane walking, turning, amplitude changing, lifting and so on. For the multi-crane hoisting system, the same modeling idea and planning algorithm are adopted, and the applicability of the solution to the multi-machine hoisting system and the high efficiency of the planning are further verified by a case study of four cranes. In this paper, firstly, the importance of hoisting action planning is analyzed, and the present situation of modeling and action planning of hoisting is summarized. It is determined that the difficulty of hoisting action planning lies in the closed chain constraint and nonholonomic kinematics constraint of the system, and the significance of this research is discussed. Secondly, for the closed chain constraints of the system, combined with the current research situation, a modeling method based on the lifting object and the crane driver is proposed, and the inverse equation solution process is given, and for the system non-holonomic kinematics constraints, According to the similarity between crawler and wheeled mobile trolley, the model of crane getting off is established, and the solving equation is obtained. Then, according to the modeling of dual-machine hoisting system, the position and pose space, motion input, position and attitude conversion matrix, distance calculation and so on are defined. At the same time, the flow of RRT-Connect algorithm and its improvement are discussed, and the lifting object is overturned by two machines. It is proved that the method is feasible and efficient to solve the problem. Finally, according to the research of action planning of double-machine hoisting system, the action planning of multi-machine hoisting system is studied. This paper analyzes the modeling of the hoisting object of the four cranes, including the treatment of multiple constraints in the system, the definition of the position and orientation of the pipe frame model, the calculation of the lifting point, etc. After defining the position and pose space of the system and the action input, through the action planning of four cranes, such as lifting, walking translation and turning over the pipe frame, it is shown that the solution of this paper is applicable to the multi-machine hoisting system. It also has higher efficiency and better action sequence in action planning.
【学位授予单位】：大连理工大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TH21

【参考文献】