Artificial Neural Network for Learning
İnverse Kinematics of Humanoid Robot Arms
Mechatronics Master (With thesis), MSc. Thesis, 2015
Assoc. Prof. Dr. Kemalettin Erbatur (Thesis Advisor),
Prof. Dr. Mustafa Ünel,
Assist. Prof. Dr. Meltem Elitaş,
Prof. Dr. Erkay Savaş,
Assist. Prof. Dr. Murat Yeşiloğlu,
Date & Time: June 04th, 2015 – 2:00 PM
Place: FENS G035
Keywords : Artificial Neural Networks, Bio-inspired Learning, Infants Developments, Inverse Kinematics, Humanoid Robots, URDF Model, Humanoid Robot's Arm Design, Multilayer Perceptron, Goal Babbling, Motor Babbling
Nowadays, many humanoid teen sized robot platforms have been developed by different research groups. The idea is either to conduct research or to produce a specific task fulfilling machine. This imposes many challenges on the design of algorithms for different actions like walk or reaching some target. There are many sophisticated humanoid research platforms available, but one crucial thing to look at is the developmental cost associated with them. As the name describes, the Humanoid robots are the ones that resemble humans in their design as well as their way of performance.
In the development of humanoid robots, many design for the arm of a humanoid robot has been developed. We have developed an arm with 5 degrees of freedom for a humanoid robot using dynamixel servo motors. We used 3D plastic printing for part manufacturing. This arm with multiple degrees of freedom enables the robot to have free movement around the body. Later we also designed a simulator model of a robot that works with the advanced simulators available today.
A great number of approaches and algorithms have been implemented to solve the problem of inverse kinematics. The research carried out in this thesis takes the early learning in infants as the basis. Infants in their early age of development move their arm to reach new goals that they have not seen yet and with the help of the visual feedback they learn the solution. We have used this idea to develop a learning algorithm that eventually enables the robot to reach goals in 3D space accurately.
This algorithm is advantageous in the sense that it is faster than the parent approach done by Matthias Rolf, “Goal babbling with unknown ranges: a direction-sampling approach”, and no prior knowledge of the arm model is required to learn the inverse solution for correct positioning. The algorithm starts with the knowledge of only one goal in the 3D space, explores more goals in the 3D space and the learning enables the algorithm to grasp the solution of inverse positioning of the arm.
The results obtained are comparable to the results generated by the aforementioned work with the advantage that the learning is faster with our algorithm. In current research in the field of cognitive and developmental robotics, one aim is to develop robots based on biological beings present on our planet. Humanoid robots can be considered as an exemplary development in this sense. Similarly, the researchers are trying to move the mathematically computational solutions more towards bio-inspired computational solutions. Therefore, exploring bio-inspired learning which was achieved by taking advantage of Artificial Neural Networks (ANNs) is another advantage associated with this work.