查询结果:   刘家学,尹鹏.改进深度信念网络在飞机下降段油耗估计中的应用[J].计算机应用与软件,2019,36(8):69 - 74.
中文标题
改进深度信念网络在飞机下降段油耗估计中的应用
发表栏目
应用技术与研究
摘要点击数
78
英文标题
APPLICATION OF IMPROVED DBN IN FUEL CONSUMPTION ESTIMATION OF AIRCRAFT DOWN SECTION
作 者
刘家学 尹鹏 Liu Jiaxue Yin Peng
作者单位
中国民航大学电子信息与自动化学院 天津 300300     
英文单位
College of Electronic Information and Automation, Civil Aviation University of China, Tianjin 300300,China     
关键词
下降段油耗估计 油耗影响因素 改进深度信念网络 高斯-伯努利受限玻尔兹曼机 自适应步长
Keywords
Fuel consumption estimation of down section Influence factor of fuel consumption Improved deep belief network Gauss-Bernoulli restricted boltzmann machine Adaptive step
基金项目
国家科技支撑计划项目(2012BAC20B0304);中美绿色航线项目(GH201661279);民航节能减排监测与报告方法研究项目(DPDSR0061)
作者资料
刘家学,教授,主研领域:民航数据分析与维修仿真。尹鹏,硕士生。 。
文章摘要
为准确估计飞机油耗量,针对飞机下降段油耗影响因素众多且与油耗呈非线性关系的特点,提出一种基于改进深度信念网络(DBN)模型的飞机下降阶段油耗估计方法。通过引入高斯-伯努利受限玻尔兹曼机(GBRBM),解决传统DBN模型中受限玻尔兹曼机(RBM)处理连续油耗输入数据时信息丢失问题;采用自适应步长(AS)策略加快收敛速度。该方法利用改进DBN模型的深层非线性网络结构实现油耗影响因素与油耗复杂非线性函数关系的逼近,并通过模型顶层连接的BP网络进行油耗估计。实验结果表明,改进DBN模型在复杂非线性估计方面有较大优势,油耗估计精度和拟合度较高,验证了该方法在飞机下降段油耗估计领域具有可行性。
Abstract
In order to accurately estimate the fuel consumption of aircraft, aiming at the many factors affecting the fuel consumption of the descending section of the aircraft and its non-linear relationship with fuel consumption, this paper proposed a fuel consumption estimation method based on the improved deep belief network (DBN) model. We solved the problem of information loss when the restricted Boltzmann machine (RBM) processed the continuous fuel consumption input data in the traditional DBN model by introducing the Gauss-Bernoulli restricted Boltzmann machine (GBRBM). The adaptive step (AS) strategy was used to speed up convergence. This method used the deep nonlinear network structure of the improved DBN model to achieve the approximation of the complex nonlinear function relationship between fuel consumption factors and fuel consumption, and the fuel consumption was estimated by BP network connected at the top of the model. The experimental results show that the improved DBN model has great advantages in complex nonlinear estimation, and the fuel consumption estimation accuracy and fitting degree are high. It is proved that the method is feasible in the fuel consumption estimation of aircraft descending section.
下载PDF全文   

根据该篇关键词查找到本刊已发表相关论文供参考
序号
文  章  标  题
作者1
发表栏目
页码
摘要
1
改进深度信念网络在飞机下降段油耗估计中的应用
刘家学
应用技术与研究
2019
8
69
[摘要]
2
一种小规模多种群布谷鸟算法
郑巧燕
算法
2014
10
278
[摘要]