Rancang Bangun Detektor Standart Preform Botol Minuman Menggunakan Metode Jaringan Saraf Tiruan

Abstract

Preform is a semi-finished material from a bottle before cooking in the blowing process. Standards form most, same shapes and colors in one production. However, it does not have to close in one production which requires several preforms that have different colors and weights than other preforms so that they are not included in the standard and must be rejected. In this case a standard detector and color of the preform drink bottle were made using backpropagation neural network method where hardware that loaded arduino uno, photodiode sensor, load cell and HX 711 module and LCD i2c 16 x 2. Photodiode sensors can be used in blue preform together with load cell which is translated directly preform which is directly converted by the HX711 module. Two input data is then processed in the Arduino UNO module. Data output from Arduino UNO is approved on the LCD and processed in the Artificial Neural Network in Matlab on the laptop. The final output of the research results will be displayed in the command window matlab column containing rich "YES" or "NO". In this study backpropagation artificial neural networks as a method to provide accurate assessment by displaying the test results with 19 grams, color density 8 with a voltage of 0.038 Volts and output data is 1 with error data -4.75E13.