摘要
航班的燃油加载量决定了航空公司运行成本。本文提出了基于飞机性能的燃油消耗估计模型及其动态修正方法。以飞机性能手册参数为依据,根据各飞行阶段的性能特点,建立基于神经网络的燃油消耗原始模型。利用积累的飞行数据(QAR数据)修正模型,消除因飞机性能衰减对燃油消耗的影响,弥补飞机性能手册中飞行气象条件样本量较少的不足。以某次北京—成都航班飞行计划数据为验证数据源,按照签派流程进行了模型的验证测试,对比本次航班实际燃油消耗量,误差为2.38%,达到现阶段国内外航空公司现有商业软件估计精度水平。
Flights fuel loading determines the airlines operating costs. In this paper, a fuel consumption model based on aircraft performance and its dynamic correction method is presented. According to aircraft performance manual parameters and the performance characteristics of each flight stage, the fuel consumption of the original model based on neural network is established. Meanwhile, considering the aircraft performance decay during the operating process, this paper uses flight data for a period of time (QAR data) to correct model dynamically and greatly improve the accuracy of the model. Also this paper gives an idea to achieve fuel consumption evaluation for dispatch business. The verification test is made by taking one Beijing - Chengdu flight plan data as validation data source. Comparing with this flight actual fuel consumption, there is an error of 2.38% which reaches the estimation precision level of large commercial software commonly used by domestic and foreign airlines.
出处
《中国民航飞行学院学报》
2015年第1期24-28,共5页
Journal of Civil Aviation Flight University of China
基金
国家自然科学基金资助项目(61079003
61179066)
关键词
燃油消耗
评估模型
飞机性能
飞行计划
Fuel consumption Evaluation model Aircraft performance Flight plan