In the current era of technological development, developments in computer systems are very fast, computerized systems are needed at this time, including to diagnose electrodes in the welding process, frame thermostats on soulplate, because sometimes it's mechanical. Mechanics find it difficult to detect damage to machines due to the absence of a good computerized system. The use of SMAW (Shielded Metal Arc Welding) in the industrial world is quite widely used. With this machine humans are greatly helped by the need to make a metal object. So that with the frequent use of these tools, it will also be more vulnerable to damage to these tools. The machine technicians supplied by the company are not proportional to the number of machines. Therefore, to help solve the problem. The purpose of this research is to make it easier for mechanics to detect engine damage. This study uses the Hebb Rule method with the simplest learning method concept. In this method learning is done by fixing the weight values in such a way that if there are 2 connected neurons, and both are 'on' at the same time, then the weight between the two is increased. If the data is represented in a bipolar manner, then the weight is improved. Therefore this method is very useful in solving a problem that occurs. The final result of this research is 1, in which the network can understand the intended pattern. It has a value of 1 because it uses a binary number pattern, not bipolar.
Copyrights © 2022