Bir tripteron robotun yapay sinir ağları ile kontrolü

Developing technologies of today is aimed to be designed in new ways of control algorithms such as artificial neural networks. In this study, a tripteron robot was designed and manufactured to be controlled by an artificial neural network algorithm. When the study considered in two main different sections, the first part of it would be the CAD design part in SolidWorks. After all the researches about the robot, the last design was built which was easy to assemble, cheap enough to be manufactured and sufficient to be used to test the developed algorithms. The second section of the study includes all the software of the robot. Two main programs being used were Matlab and Arduino IDE. The Matlab was firstly responsible to apply the determined neural network algorithm to convert the end effector position data to the linear guide position data. The determined artificial neural network algorithm was simple adaptive linear combiner. Secondly, the Matlab was responsible to calculate the required simultaneous stepper motor driving data by the other developed algorithms. Arduino microcontroller was used to drive the stepper motors according to the data which calculated on Matlab and stored in a micro SD card. While the Arduino can drive digital signals through only serial ports, the stepper motors were successfully driven simultaneously to achieve the most realistic robotic motion.