基于神经网络液体横向晃动时罐体受力的预测

用Fluent软件,VOF(Volume of Fluid)模型,对基准罐体在不同充液比下受到横向加速度时的受力进行数值模拟;以充液比、前两时刻基准罐体受力、加速度及将要经历的加速度作为输入,以下一时刻受力作为目标输出,选用合理的计算结果作为训练样本,建立基于BP神经网络液体横向晃动时基准罐体受力的预测模型,用158个样本对完成训练的网络进行可靠性验证,横向力、垂向力和侧倾力矩最大预测误差分别为8.88%,0.36%,1.38%,符合精度要求.基准罐体的时间步长和受力进行修正后,与一般圆柱及椭圆形罐体受力的大小和规律基本一致;对于作横向运动的柱形罐体,受力大小与罐体长度成正比.通过修正基准罐体的时间步长和受力,对一般圆柱及椭圆形罐体的受力也可实现BP神经网络的预测,为罐车动力学分析快速有效地提供所需数据.      

神经网络在无人机电控活塞发动机试验中应用

分析了无人机用电控活塞发动机试验特点以及试验中存在的难点,针对电控发动机高海拔标定试验中进气歧管压力(manifold air pressure,简称MAP)传感器数据的传统线性插值方法不能完全表述电控发动机非线性特性的缺陷,提出采用BP(back propagation)神经网络模型的解决方案.为避免目前应用神经网络方法中所存在的不足,通过采用原始数据分组方法进行网络训练误差的实时反馈和控制,较好地解决了神经网络训练过程中容易陷入"局部最优"和"过拟合"状态,并对BP神经网络预测结果给予了详细研究,训练误差和预测误差分析结果表明了该方法的可行性和计算结果的可信性.      

航空电子高速交换式网络的服务质量保证方法

以保证航空电子高速交换式网络的实时性和高带宽为目的,提出光纤通道(FC,Fibre Channel)通信协议承载于波分复用(WDM,Wavelength Division Multiplexing)传输机制之上的架构.采用适用于实时交换的周期性数据流模型,提出了基于负载匹配的输出轮询(LOR,Load-matching Output Round)调度算法.该算法仅使用一次仲裁即可达到输入/输出的最大匹配(100%).给出LOR算法的核心代码,时间复杂度仅为O(1),易于在硬件中实现.最后通过将LOR算法嵌入到网络处理器中,注入时间敏感的负载流量,以服务质量特性(延迟、吞吐和超时消息数)为性能衡量指标进行实验,结果表明LOR算法在轻负载(30%)下,服务质量特性与经典算法(iSLIP, DRR, EDRR)基本持平;在重负载(70%)下,仅LOR算法能满足延迟的实时性要求,且LOR吞吐率高于经典算法近10个百分点,超时消息数比经典算法少了100个以上.      

基于深度学习的无人机航拍目标检测研究综述

目标检测是提高无人机（UAV）感知能力的关键技术之一,其研究对于无人机的应用有着重要意义。与基于手工特征的传统方法相比,基于卷积神经网络的深度学习方法具有强大的特征学习和表达能力,成为目前目标检测任务的主流算法。近年来,目标检测技术已经在自然场景图像上取得了一系列突破性进展,在无人机领域的研究也逐渐成为热点。首先系统阐述了基于深度学习的目标检测算法的研究进展,并总结了相关算法的优缺点。对常见的航空影像数据集进行了梳理并介绍了迁移学习的方法;从无人机影像背景复杂、目标较小、视场大、目标具有旋转性的特点出发,对无人机目标检测在近期的研究进行了归纳和分析。最后讨论了存在的问题和未来可能的发展方向。     

一种移动卫星网络最小费用切换算法(英文)

For mobile satellite networks, an appropriate handover scheme should be devised to shorten handover delay with optimized application of network resources. By introducing the handover cost model of service, this article proposes a rerouting triggering scheme for path optimization after handover and a new minimum cost handover algorithm for mobile satellite networks. This algorithm ensures the quality of service （QoS） parameters, such as delay, during the handover and minimizes the handover costs. Simulation indicates that this algorithm is superior to other current algorithms in guaranteeing the QoS and decreasing handover costs.     

基于粗糙集-神经网络的机载设备故障诊断方法研究

提出了一种基于粗糙集-神经网络的故障诊断方法。将粗糙集理论同神经网络结合起来可以用于机载设备的故障诊断。采用粗糙集理论对原始故障诊断样本进行处理,并根据条件属性对决策属性的正域的大小来选择条件属性,提取出对诊断故障贡献最大的最小故障特征集,从而确定神经网络的拓扑结构;通过训练神经网络建立故障特征与故障之间的映射关系,实现故障的诊断。通过A320飞机燃油系统的故障诊断仿真实例,表明这种故障诊断方法的有效性。     

基于组合不变矩与神经网络的卫星目标识别

首先介绍了Hu提出的不变矩和Jan Flusser提出的仿射不变矩,在不增加特征向量维数的前提下,将不变矩的部分分量与仿射不变矩组合成新的特征向量——组合不变矩。利用组合不变矩提取图像中的不变特征向量,并结合基于人工神经网络的分类器,较好的实现了对卫星目标的分类识别。     

Neural network prediction of the topside electron content over the Euro-African sector derived from Swarm-A measurements

This study presents the first prediction results of a neural network model for the vertical total electron content of the topside ionosphere based on Swarm-A measurements. The model was trained on 5 years of Swarm-A data over the Euro-African sector spanning the period 1 January 2014 to 31 December 2018. The Swarm-A data was combined with solar and geomagnetic indices to train the NN model. The Swarm-A data of 1 January to 30 September 2019 was used to test the performance of the neural network. The data was divided into two main categories: most quiet and most disturbed days of each month. Each category was subdivided into two sub-categories according to the Swarm-A trajectory i.e. whether it was ascending or descending in order to accommodate the change in local time when the satellite traverses the poles. Four pairs of neural network models were implemented, the first of each pair having one hidden layer, and the second of each pair having two hidden layers, for the following cases: 1) quiet day-ascending, 2) quiet day-descending, 3) disturbed day-ascending, and 4) disturbed day-descending. The topside vertical total electron content predicted by the neural network models compared well with the measurements by Swarm-A. The model that performed best was the one hidden layer model in the case of quiet days for descending trajectories, with RMSE = 1.20 TECU, R = 0.76. The worst performance occurred during the disturbed descending trajectories where the one hidden layer model had the worst RMSE = 2.12 TECU, (R = 0.54), and the two hidden layer model had the worst correlation coefficient R = 0.47 (RMSE = 1.57).In all cases, the neural network models performed better than the IRI2016 model in predicting the topside total electron content. The NN models presented here is the first such attempt at comparing NN models for the topside VTEC based on Swarm-A measurements.     

Ascent guidance law for a horizontal take-off vehicle with a multi-combined cycle engine

This paper researches the ascent guidance law for the vehicle with a multi-combined cycle propulsion. The guidance law comprises two parts, namely, the off-line optimal trajectories generation and online guidance. With respect to the off-line part, disturbances are discretized and incorporated into the trajectory optimization problem; subsequently, a set of trajectories is calculated to constitute a database. To quickly obtain a database that comprises a large number of trajectories, a novel ascent profile is proposed with respect to height and velocity. Based on this profile, only inequity constraints exist in the optimization model, and the original optimization problem is converted to a parameter searching problem. The optimal trajectories are calculated using a hybrid optimization method that comprises a particle swarm optimization (PSO) method and the Hooke-Jeeves (HJ) method. With respect to online guidance, the profile is updated using a radial basis function neural network (RBFNN) based on the current flight states and the database. Simulation validates the efficiency of the proposed optimization method by comparing the method with the pseudospectral method; the robustness of the guidance law is also validated using Monte Carlo simulation.     

无人机编队支撑网络的协同通信中继策略

针对无人机编队支撑网络优化选择中继成员进而提高编队协同通信能力问题，基于端到端技术（Device-to-Device，D2D），提出了一种同时考虑无人机编队成员的邻近域及编队成员间的社会性依存关系来优化选择中继成员的中继策略，旨在既能保证无人机编队的基本通信能力，又能同时优化提高编队的吞吐量，从而进一步提升编队支撑网络的整体性能。研究结果表明，所提的无人机编队支撑网络中继成员优化选择策略在系统吞吐量方面较已有方法具有明显优势。