IMPLEMENTASI LEARNING VECTOR QUANTIZATION UNTUK KONTROL GERAK ROBOT DALAM MENGAMATI OBJEK YANG TERBUNGKUS
Daftar Isi:
- One of the task that is usually delegated lo a robot is to follow an object with the aim of knowing shape of objects that have been determined which are circle ob1ect. square ob1ects and tnangular objects .. In this research, the logic of neural networks has been chosen using the learning Vector Quanti=ation (LVQ) algorithm as a pattern recognition methot that can used to classify object shapes. This method was chosen for research because it has an automatic learning response to perform classijicat,on of the given input vector. , in this research, the robot uses several hardware modules which are 3 infrared distance sensors on the nght side, 3 infrared distance sensors on /he /eji side, 2 DC motors and use a minimum system ATMega8535. Reliability Jes/ of learning Vector Quanti=ation logic is using a program that was created with the CJC ++ programming language. As the results of val,dation of Method /he percentage of success for pallern recognition oflhe Jes/ using the program based on the result of 200 sensor data on the circle object by 6-1,5%. 200 sensor data on the square objeCI by ./3% and also 200 sensor data on the triangular object by 75%.