基于神经网络的垂尾飞行载荷模型研究

针对现代飞机结构设计复杂、传力路径多和大载荷结构变形大等因素,提出用神经网络方法建立载荷模型.利用垂尾载荷校准试验数据建立基于神经网络的垂尾根部剪力和弯矩的载荷模型,并对多点进行载荷预测.通过与多元线性回归载荷预测结果进行对比分析,结果表明,在大载荷下神经网络预测精度优于多元线性回归,从而为测量垂尾飞行载荷提供了一种可...     

小波神经网络在板形板厚控制系统中的应用

带材轧制是一个复杂的非线性过程,针对板形和板厚控制相互耦合等特点,本文提出了一种基于小波神经网络的自适应控制新算法。文中系统由两个小波神经网络组成,分别实现综合系统的模型辨识和控制。由于小波变换的紧支性及神经网络的非线性映射能力,模型辨识能准确地辨识板形板厚系统的动态特性,控制器能产生较为复杂的控制规律。仿真结果证明,该板形和板厚控制系统具有良好的自适应跟随和抗扰性能,其控制效果优于传统的解耦PID控制。     

石英晶振老化的建模与模型求解

为预测石英晶振的频率老化，提出了一种非线性时变模型。在一阶马尔可夫假设下，利用多层前向神经网络进行迭代逼近求解。对四种晶体的全部或部分实验数据进行了学习和预测试验。结果表明该模型不仅能拟合函数，而且可在存在野值时，根据以往的数据预测目前的频率。与其他线性模型相比，该模型的预测能力更佳，适用于多种类型的晶振老化模式，预测误差小，有很好的抗噪声能力。     

分数阶Hopfield型神经网络的全局渐近稳定性

研究了分数阶Hopfield型神经网络的全局渐近稳定性,通过LMI方法得到了一种实现系统全局渐近稳定性的LMI形式条件,通过实例仿真验证了结论的正确性。     

贝叶斯网络在飞控系统可靠性评估中的应用

把一种基于贝叶斯网络的故障树分析法应用于飞控系统俯仰控制功能的可靠性评估中,突破传统故障树分析法的局限性——各部件独立、状态为二值,不但可以计算系统的可靠性指标,而且可以定量给出某个事件或某几个事件在系统可靠性中所占的地位,找出系统的薄弱环节。     

Neural network development for the forecasting of upper atmosphere parameter distributions

This paper presents a neural network modeling approach to forecast electron concentration distributions in the 150–600 km altitude range above Arecibo, Puerto Rico. The neural network was trained using incoherent scatter radar data collected at the Arecibo Observatory during the past two decades, as well as the Kp geomagnetic index provided by the National Space Science Data Center. The data set covered nearly two solar cycles, allowing the neural network to model daily, seasonal, and solar cycle variations of upper atmospheric parameter distributions. Two types of neural network architectures, feedforward and Elman recurrent, are used in this study. Topics discussed include the network design, training strategy, data analysis, as well as preliminary testing results of the networks on electron concentration distributions.     

A New Exposed-terminal-free MAC Protocol for Multi-hop Wireless Networks

This article presents a new multichannel medium access control (MAC) protocol to solve the exposed-terminal (ET) problem for efficient channel sharing in multi-hop wireless networks. It uses request-to-send and clear-to-send (RTS/CTS) dialogue on a common channel and flexibly opts for conflict-free traffic channels to carry out the data packet transmission on the basis of a new channel selection scheme. The acknowledgment (ACK) packet for the data packet transmission is sent back to the sender over another common channel thus completely eliminating the exposed-terminal effects. Any adjacent communication pair can take full advantage of multiple traffic channels without collision and the spatial reuse of the same channel is extended to other communication pairs which are even within 2 hops from them. In addition, the hidden-terminal effect is also considerably reduced because most of possible packet collisions on a single channel are avoided due to traffic load balance on multichannels. Finally, a performance comparison is made between the proposed protocol and other typical MAC protocols. Simulation results evidence its obvious superiority to the MAC protocols associated with other channel selection schemes and traditional ACK transmission scheme as well as cooperative asynchronous multichannel MAC (CAM-MAC) protocol in terms of four performance indices: total channel utilization, average channel utilization, average packet delay, and packet dropping rate.     

时滞细胞神经网络平衡点的全局指数稳定性研究

利用常数变易法和变量替换的方法，研究了具有时滞的细胞神经网络平衡点的奎局指数稳定性，给出了网络平衡点的全局指数稳定性的充分判据。     

空中交通网络管理的概念与策略研究

提出以空中交通网络为主要研究对象的解决空中交通管理的思路,定义了空中交通网络的概念,并对空中交通网络管理的策略和机制进行了探讨。     

采用BP网络辨识航空发动机数学模型

运用ＢＰ网络和实测数据作为学习样本，对某型航空发动机的数学模型进行了辨识研究。辨识模型输出的结果与实测数据比较误差较小。这种方法收敛速度快、精度高，结果表明用ＢＰ网络辨识方法能够得到比较精确的发动机数学模型。