PREDICTION OF NOx AND FUEL FLOW OF COMMERCIAL HIGH BYPASS AIRCRAFT ENGINES BASED ON CSA-SVR MODEL

Due to the negative effects of emissions caused by fossil fuels used by aircraft engines on the environment and human health, and the fact that fuel consumption is a high cost input for airlines, the aviation community has many studies on both issues. In order to overcome these problems, much space has been devoted to modeling, prediction and optimization studies on emissions and fuel consumption in the literature. Within the scope of this study, a model was created to predict the NOx emission values and fuel flow of high by-pass turbofan engines, which are also used in today's commercial air transportation. 165 different turbofan data taken from the International Civil Aviation Organization (ICAO) emission databank were used for modeling, and the specified parameters were modeled according to the by-pass ratio (BPR), overall pressure ratio (OPR) and rated thrust input parameters. In this context, the Cuckoo search algorithm-support vector regression (CSA-SVR) method for the Landing and Take-off (LTO) cycle, which includes the idle, take-off (T/O), climb out (C/O) and approach (App) phases, was used for the first time in the literature for the above-mentioned purpose. As a result of the error analysis methods, the minimum R2 value for 4 phases in FF estimation was found to be 0.972763. This value for NOX was 0.6745 in the idle phase. However, the fact that this value was found to be 0.861497, 0.884984 and 0.792779 for T/O, C/O and App, respectively, shows the success of the model in estimating actual data.