Enhanced cruise range prediction for narrow-body turbofan commercial aircraft based on QAR data

Abstract

Aircraft performance parameters play a critical role in maintaining economic and environmental sustainability in aviation. Furthermore, the ability to calculate aircraft performance parameters accurately for the cruise range contributes to aviation in areas such as the preliminary design of aircraft and air traffic management. This study is focused on cruise range performance, as this is critical to both the evaluation and understanding of the economic and environmental impacts of commercial aircraft. Quick Access Recorders (QAR) data were used for more accurate analysis of the cruise range. The QAR data used in this study included 6,574 short-distance domestic flights by narrow-body turbofan commercial aircraft between 31 different city pairs. To obtain a more accurate cruise range equation, parameters affecting the cruise range performance were determined and studied. First, the drag polar model was improved to take the cambered profile, compressibility effects and cruise airspeeds of commercial aircraft into consideration using the real flight data. Second, Thrust-Specific Fuel Consumption (TSFC) models were compared and the most suitable one for the cruise phase was selected. After these steps, cruise range values were calculated using the Breguet range equation with these improved parameters. When the results of this enhanced range model were compared with the real flight data, the mean absolute percentage error (MAPE) was found to be 2.5% for all the Aircraft and Engine Type Groups (AETGs) considered in the data. This figure corresponds to a 7.9% smaller error than provided by previous range models based on simple parabolic drag polar and TSFC models. According to these results, the application of a simple parabolic drag polar and TSFC is not appropriate for cruise range calculations.

Analysis of Abnormal Flight and Controllers Data Based on DBSCAN Method

“…e samples of the same category are closely connected, and a cluster category is obtained by dividing the closely connected samples into one category [12]. By dividing all closely connected samples of each group into different categories, the final results of all clustering categories are obtained [13].…”