Paper Title
DEVELOPMENT OF A SURFACE ROUGHNESS PREDICTION MODEL DURING PRINTING PROCESS OF 3D PRINTER USING FREQUENCY -SERIES DATA

Abstract
Prototyping and low-cost manufacturing widely utilize a type of 3D printing called Fused Deposition Modeling (FDM). However, they have limited conditions;the distance between the nozzle and printer bed, and the abnormal state of the belt which is a factor that affects the quality of the prototype. In this article, we investigated frequency-series data in a 3D printing process. The goal is to use this information to predict the roughness of the 3D printer. In this research, we used 6 algorithms to compare the efficiency: autoregressive(AR), autoregressive integrated moving average (ARIMA), XGBoost, support vector regression (SVR), long short-term memory (LSTM), recurrent neural, and network (RNN). Modeling with 20 percent of the total dataset: accelerometer RMSE of 6.35 µm and runtime of 0.12 seconds, and acoustic emission (AE) sensor RMSE of 6.58 µm and runtime of 0.07 seconds. Keyword - additive manufacturing (AM), fused deposition modeling (FDM), predictive modeling, surface roughness, frequency series