SCARA robot educated with artificial neural network

Key Words: SCARA robot, Robot dynamics, Trajectory, Artificial neural network Robot manipulators are used in various fields to enhance the effiency to need to get better quality to make works which are sensitive and repeatable easily for different requirements. We say that ?the simple applied for an operator which use different reflex and itelligence? for the robots. In this study the subject which SCARA manipulator which has four degrees of freedom is educated by using artificial neural network (ANN) is considered. Firstly Lagrange ? Euler dynamic and trajectory equations of the modelled robot are obtained for use in artificial neural network (ANN). The modelled robot has three rotary and one linear motion. The rotary motion is defined by cubic trajectory and linear motion has linear trajectory. The cubic and linear trajectory cefficients are acquired by using the boundary conditions in a particular time period and thus position equations are also carried out. Torque values are introduced to ANN as input and position values are obtained as output in a particular time period. The ANN architecture which has a structure at 4:11:4 used for educated by network. Performance graphic and graphics which shown position output and error values for each one link are acquired with the educated robot manipulator by working with ANN. The values at the test system which have never been seen by the ANN had been shown to the network and also the replies which were given by network had been analysed.