Trajectory Planning And Motion Control Techniques Of Parallel Mechanism For Cabins Docking

Cabins docking is the key process of assembly,test and experiment of aerospace products,It has strict operational requirements in terms of technique and security.Cabins docking technology is directly related to the production quality and efficiency of aerospace products,and has become a research focus in the fields of aerospace and manufacturing.In regards to the problems of low efficiency,poor accuracy and quality consistency of manual docking technology,a 6-UPU type parallel mechanism is developed as a cabin docking equipment,and the related key techniques of cabins docking are studied in this paper.The aim is to provide a systematic and effective method to improve the accuracy and efficiency of cabins docking.This research is of great significance for improving the production quality and shortening the production cycle of aerospace products.Basic researches are carried out on the 6-UPU parallel mechanism.Based on the 3D model and structural parameters design of the mechanism,the inverse kinematics model is established by using closed-loop vector chain,and the inverse dynamics model is established by using the principle of virtual work.Simulation analysis of translation,rotation and composite motion is implemented to verify the correctness of kinematics and dynamics models.In order to make efficient use of parallel mechanism to achieve cabin pose alignment,a singular-free trajectory planning method of parallel mechanism under multi-objective and multi-constraint conditions is proposed.B-spline curves are employed to parameterize the motion trajectories of parallel mechanism and branch chains.After combining the requirements of efficiency,stability and energy consumption of the cabins docking task and considering the performance and motion constraints of the parallel mechanism,the multi-objective trajectory optimization model of the parallel mechanisim is established,and an improved Immune Cloning Selection Algorithm(IICSA)is developed to realize the multi-objective trajectory optimization of the parallel mechanism.In regards to the pose measurement of cabins without mark points,a pose fitting method based on Key Characteristics(KCs)is proposed.In order to improve the fitting accuracy of point set matching algorithm,a quaternion pose fitting method based on multi-feature point measurement is proposed.In order to obtain the relative pose relation in cabin docking,a method of relative pose relation measurement and equivalent transformation is studied,and the relative pose with higher accuracy is fitted by the Nelder-Mead Algorithm(NMA).Aiming at the target pose conversion of the mobile platform,several reliable and effective methods such as pose vector conversion and coordinate relation conversion are studied.To track the planned trajectory of the parallel mechanism accurately and stably,an improved control method based on the improved Dynamic-recursive Fuzzy Neural Network(DFNN)and incremental PID is proposed.FNN is improved to DFNN by recursive units so as to improve the time varying property and dynamic mapping ability,thus the control precision and stability of DNFF-PID can also be improved.In order to reduce the effect of network parameter initial value on trajectory tracking accuracy and overcome the limitation of BP algorithm,a network parameter optimization learning algorithm based on the improved Fish-swarm Algorithm(IFA)is proposed.A cabins docking simulation system is designed,a cabins docking motion control software is developed,and a cabins docking experiment platform is built.By using the above methods,the parallel mechanism is controlled to realize the automatic docking of cabins.