摘要
为了准确预测正常状态下民航发动机的滑油消耗量,以某型号民航发动机的快速存取记录器Quick Access Recorder(QAR)数据建立能够预测正常状态下滑油消耗的模型并预测。利用遗传算法对输入数据进行筛选并优化网络的权值和阈值,建立BP网络。在此基础上对遗传算法的遗传算子进行改进,建立新的优化BP网络。将单BP网络的仿真结果分别与两种优化过的网络仿真结果对比,结果表明优化过的BP网络提高了预测的准确率,并且改进后的遗传算法优化的BP网络准确率更高。由此证明改进遗传算法优化的神经网络在预测滑油消耗上具有很强的实用性。
In order to predict the normal consumption of civil aviation engine lubricating oil,a lubricating oil consumption model for civil aviation engine in normal working condition was set up based on a civil aviation engine Quick Access Recorder（QAR）.Genetic algorithm was used to filter the inputted data and optimize the network weights and thresholds,and then the BP neural network was built.On this basis,the genetic operators of genetic algorithm was improved,and a new optimization BP network was established.The simulation result of single BP network was compared with that of two optimized networks.The results show that the optimized GA-BP network improves the accuracy of forecasting lubricating oil consumption.
出处
《机械工程与自动化》
2017年第2期9-10,13,共3页
Mechanical Engineering & Automation
基金
中央高校基本科研业务费资助项目(Y16-02)
关键词
滑油消耗量
遗传算法
BP神经网络
民航发动机
lubricating oil consumption
genetic algorithm
BP neural network
civil aviation engine
