Please use this identifier to cite or link to this item:
Title: Modelling, kinematics, dynamics and control design for under-actuated manipulators
Authors: Albalasie, Ahmad 
Keywords: Robots - Control systems trajectory planning algorithm; quasi-linearization; null space motion; pick and place robot;Remote control;Manipulators (Mechanism);Predictive control;Automatic control;Trajectory optimization - Mathematical models
Issue Date: 16-Sep-2016
Publisher: Technical University Berlin
Abstract: Robotic handling operations cover a diversity of applications. Pick and place, palletizing or depalletizing, loading on machines or unloading from machines, storage/retrieval, and feeding the production lines are some examples. A manifold of different applications inspires the development of industrial robot types. The advantages of industrial robots can be summarized in different aspects: worker’s protection in dangerous working conditions, higher working quality, higher productivity rate, and cost saving. Due to concerns about resource efficiency, energy consumption has become an issue for robotic development. The System Applying Momentum Transfer for Acceleration of an End Effector with the Redundant Axis (SAMARA) is a robotic prototype of an industrial robot for pick and place applications. This prototype uses redundant, under-actuated configurations and an evolutionary algorithm (EA) to minimize energy consumption. Enabling for applications with relatively large displacement tasks, higher than one meter and high payload of up to 5.5 kilograms, the effectiveness of handling can be increased. Energy saving in specified cycle time has been achieved for this robotic kinematics. Reducing the cycle time and energy consumption are conflicting goals. However, actually, the computation time for the trajectory planning is too long. PID (proportional–integral–derivative) control is not adequate for a robust under-actuated motion (UAM). The uncertainty of payload causes unacceptable effects on accuracy, repeatability, and precision of the under-actuated robot. Using the Quasi-Linearization (QL) is an approach for trajectory planning with minimizing energy consumption and reducing computation time. The QL is focused on reducing the cycle time to increase the productivity of the handling operations to achieve an optimal performance for the robot to meet the industrial requirements. The suggested control scheme uses the adaptive model predictive control (AMPC). The AMPC is classified as an advance optimal control technique; it has the ability to minimize the input torque, and the error between the actually achieved response and the desired response of the manipulator. The model has the inherent ability to deal naturally with constraints on the inputs and has the capability of updating the linearized dynamic model at each current operating point, which solves the problem of nonlinearity in dynamic equations of the robot. Evaluations for the control scheme and for the trajectory planning are tested for SAMARA prototype. The concepts have been verified using several criteria, e.g., by comparing the results between the simulation power consumption and the actual power consumption measured from the physical prototype, comparing the performance of the QL approach with EA as trajectory planning algorithm for the under-actuated motion, and comparing the performance of SAMARA with other industrial robots from several perspectives. The applicability of under-actuated robotic kinematics for practical applications has been approved by examples from food industry, and press lines industry, with their respective requirements.
Description: An approved dissertation for the academic degree Doctor of Engineering, from the Technical University Berlin, 2016
Appears in Collections:Fulltext Publications

Files in This Item:
File Description SizeFormat
albalasie_ahmad.pdf7.54 MBAdobe PDFView/Open
Show full item record

Page view(s)

checked on Jun 27, 2024


checked on Jun 27, 2024

Google ScholarTM



Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.