Support Vector with ROC Optimization Method Based Fuel Consumption Modeling for Civil Aircraft

This paper is to present a simplified model to estimate aircraft fuel consumption using support vector algorithm. The method developed here can be implemented in fast-time airspace and airfield simulation application. A representative support vector network aided fuel consumption model is developed using data given in the route date and aircraft performance manual, support vector machine is trained to estimate fuel consumption of a certain aircraft. Also Receiver Operating Characteristic Curve is introduced to the performance evaluation of trained model. This methodology can be extended to any type of aircraft including piston and turboprop type with confidence. The data used in this study is applicable to the Boeing 737-800 aircraft which powered by CFM56 engines. Model outputs were compared to the actual performance provided in the aircraft performance manual and found to be accurate for implementation in fast-time simulation models. The results of this study illustrate that a support vector model with ROC optimization can accurately represent complex aircraft fuel consumption functions for full flight phase.