PERANCANGAN IN-HOUSE ENGINE CONDITION TREND MONITORING UNTUK MESIN PW-127M PADA PESAWAT ATR 72-600

• Main Author: Anderson, Filbert
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/68428
• Institution: Institut Teknologi Bandung

At of today, aircraft maintenance is dominantly done using condition-based maintenance (CBM). In order for CBM to be implemented on PW-127M Engines, MRO companies use Engine Condition Trend Monitoring (ECTM). ECTM compares the real conditions of several engine parameters with a mathematical model of the machine to determine whether maintenance action is necessary to take and what type of action is needed. PW-127M engines’ ECTM program is entirely processed by P&WC’ Danish Engine Trend Analyzing (DETA). Using DETA’s service incur large costs. By building the ECTM model independently, these costs can be avoided. The model is built with python using two different methods, namely multiple-linear regression (MLR) and neural network (NN). The model is trained by comparing the raw aircraft engine data with the output data that has been processed by DETA's ECTM program. The research conducted is declared successful if the trend displayed by the prediction model built can model the ECTM output with the same raw input data. Of the two selected methods, the model built with MLR could predict ECTM outputs quite well, while the NN model still fails to predict ECTM outputs. Suggestions for further research are to increase the amount of training data and also to further examine the data normalization process carried out by the company before being processed by ECTM.