Research On Computer Aided Crane Selection And Lifting Planning

Since1980s, the technology of Computer Aided Lift PlAn Design (CALPAD) has emerged as the times require and received extensive attention from many researchers. However, at present, most studies of crane selection did not consider some engineering factors, such as the ground pressure, the clearance between the lifted object and the boom. Furthermore, few studies have focused on crane selection for crawler cranes with lattice boom. In lift path planning research, almost all studies assumed that the bottom of crane is fixed, without considering the traveling of crane with load. Among the research groups working on lift simulation, only a few are focusing on simulation of cooperative operations for dual-crane lifts. And the parameters were numerous and difficult to be determined in the existing simulation approaches for dual-crane lifts. Few existing achievements have been used in current practice because of lacking consideration of engineering constraints. Therefore, this dissertation does research on the following three aspects:intelligent crane selection based on the multiple constraints, lift path planning of single crane considering traveling, and simulation for dual-crane lifts. A computer system was implemented to integrate all the computational methods developed in the research. This system referred to as CALPADS, which stands for "Computer Aided Lift Plan Design System", is aimed to help engineers to design lift plan. In practice, the computer system has been used in several heavy lift projects of Sinopec. The main contents of this dissertation are summarized as follows:(1) An algorithm of suboptimal crane selection with multiple constraints for mobile cranes is proposed. Firstly, we construct a mathematical model for crane selection under the multiple constraints, and design the general framework of crane selection algorithm. Secondly, we take crawler crane with lattice boom for example to provide the approaches of handling the constraints of lifting capacity, clearance between the lifted object and the boom, ground pressure. Finally, the effectiveness and usability of the developed algorithm are verified by using an actual heavy lift project. Comparing with the existing crane selection algorithms, the proposed algorithm reduces the difficulty of handling the clearance constraint by transforming the complex3D spatial distance computation into two-dimensional geometric calculation. The accuracy of crane selection is raised for the consideration of ground pressure. The proposed algorithm is more fit for engineering projects and is conveniently used since it does not need to create the3D models of cranes and lifted objects. (2) A novel path planning algorithm, named as RRT-Connect++, is proposed based on priori information. For the drawback that the path returned by RRT-Connect is not optimal, we use prior information to guide the tree grows towards high quality region. Three modifications on the original RRT-Connect algorithm are made:constructing sampling pools with those promising vertices of trees and picking random state from them; avoiding sampling from the explored regions; adding the middle vertices during the Connection operation and testing regression of vertices to guarantee the quality of trees. Results from the several simulation experiments show that the RRT-Connect++can significantly improve the quality of path and keep the high efficiency of the original RRT-Connect. Also, RRT-Connect++generally out-performs several other RRT-based algorithms.(3) An efficient lift path planning approach for crawler crane is proposed, in which the mobility of crawler crane and its nonholonomic kinematics are considered. Firstly, we formulate the problem of lift path planning for a crawler crane, and elaborate the overview of lift path planning approach based on RRT-Connect++. Then, the definition of the C-Space for crawler crane, metric in C-Space and representation of the nonholonomic kinematics constraint for crawler crane are provided. Finally, the results of these experiments showed that the proposed approach can find a feasible collision-free lift path and the obtained path satisfies collision constraint, lifting capacity constraint and nonholonomic constraint of the mobility of crawler crane. The lift path is more natural and can be used to control the crane since the nonholonomic kinematics constraint is incorporated into the proposed approach. A metric of two configurations in C-space and a path length function are developed without any weight coefficients for each degree of freedom.(4) A model of dual-crane lift system for classic cooperative erections is established and a simulation framework is provided. Firstly, the two cranes and the lifted object are modeled as a whole complicated system (called dual-crane lift system), and we study the collaborative operations between the dual cranes in classic erections from the view of dual-crane lift system; then, the model of dual-crane lift system is constructed; finally, a simulation framework based on space geometry constraints is presented. The effectiveness and usability of the presented simulation framework is verified by using an actual heavy lift project. This simulation approach is more accurate and easier to be manipulated since the strategies of collaborative operations between the dual cranes have been incorporated into the basic actions of the dual-crane lift system. Besides, the concept of dual-crane lift system provides a novel view for other research of dual-crane cooperative lifts.(5) The forward kinematics model for dual cranes cooperative lift is modeled and a simulation framework is provided. For general dual-crane lifts, we construct the kinematics model of the dual-crane lifting system. After that, we utilize the principle of minimum potential energy to transform the problem of forward kinematics into an optimization problem with geometric and physical constraints, and propose a numerical method to solve the slopes and tensions of lifting cables. On above foundation, a general simulation approach for dual-crane lifts based on forward kinematics is presented. Through an actual case, the simulation approach is proven as an effective tool for designing and previewing the lifting processes of dual-crane lifts. Compared with the existing simulation methods based on dynamics, this approach can accurately and fast determine the lifted object pose and tensions of lifting cables, just inputting the weight and relative gravity center of the lifted object. This approach can be easily embedded lifting simulation software to achieve real-time simulation for dual-crane lifts, with the characteristics of fewer parameters, real-time, and so on.