基于遗传算法优化BP神经网络的飞机油耗预测方法

Aircraft fuel consumption prediction method based on BP neural network optimized by genetic algorithm

飞机油耗的精准预测可以有效减少环境污染、节约燃油能源、为航空公司降低运营成本。为了提高飞机油耗的预测精度,本文采用主成分分析方法从QAR数据中选择对飞机油耗影响较大的地速、纵向加速度、垂直加速度、风速、风向、倾斜角、空速、气压高度作为BP神经网络的输入变量,提出了基于遗传算法优化反向传播神经网络的飞机油耗预测方法。通过Matlab仿真软件建立了预测模型,以某航空公司飞机下降阶段QAR数据为基础进行验证实验。实验结果显示,该模型的预测精度优于传统的BP神经网络模型,预测性能更好。

Accurate prediction of aircraft fuel consumption can effectively reduce environmental pollution,save fuel energy,and reduce operating costs for airlines.In order to improve the prediction accuracy of aircraft fuel consumption,the ground speed,longitudinal acceleration,vertical acceleration,wind speed,wind direction,tilt angle,airspeed and air pressure height with greater influence on aircraft fuel consumption are selected from the QAR data by principal component analysis method,and an aircraft fuel consumption prediction method based on genetic algorithm optimization backpropagation neural network is proposed.A predictive model is established through Matlab simulation software,and a verification experiment is conducted based on the QAR data of the aircraft descent phase of an airline.Experimental results show that compared with the traditional BP neural network,the model predicts the effect more accurately and the prediction performance is better.