Modeling and Docking Trajectory Optimization of LNG Loading Arm Based on Modified D-H Method

Abstract

Aiming at the low accuracy and poor stability of manual docking for 5-axis LNG loading arms, this paper focuses on its kinematic modeling and docking trajectory planning optimization. First, the hydraulic-driven LNG loading arm was simplified into a 5-joint robotic arm model with a virtual joint, and its kinematic model was established based on the Modified Denavit-Hartenberg (M-DH) method. A D-H parameter table was constructed, and the safe rotation range of each joint was defined via motion interference analysis. Considering the large mass-span ratio of the LNG loading arm, joint space trajectory planning was adopted. Four interpolation algorithms (cubic/quintic spline, cubic/quintic polynomial) were compared and simulated by Matlab Robotics Toolbox, with the motion characteristics of the end effector and joints evaluated. The results show that the quintic polynomial interpolation algorithm is optimal: it generates a continuously differentiable spatial arc trajectory in Cartesian space with C⁴ continuity of position, velocity and acceleration in the time domain. The joint motion is smooth without abrupt changes, the jerk index meets engineering requirements, and the trajectory has high fault tolerance. This study provides a reliable kinematic model and optimal trajectory scheme for the automatic docking of LNG loading arms, offering theoretical and technical support for its automation transformation and high-precision controller design.