Practical compensation for nonlinear dynamic thrust measurement system

The real dynamic thrust measurement system usually tends to be nonlinear due to the complex characteristics of the rig, pipes connection, etc. For a real dynamic measuring system,the nonlinearity must be eliminated by some adequate methods. In this paper, a nonlinear model of dynamic thrust measurement system is established by using radial basis function neural network(RBF-NN), where a novel multi-step force generator is designed to stimulate the nonlinearity of the system, and a practical compensation method for the measurement system using left inverse model is proposed. Left inverse model can be considered as a perfect dynamic compensation of the dynamic thrust measurement system, and in practice, it can be approximated by RBF-NN based on least mean square(LMS) algorithms. Different weights are set for producing the multi-step force, which is the ideal input signal of the nonlinear dynamic thrust measurement system. The validity of the compensation method depends on the engine’s performance and the tolerance error0.5%, which is commonly demanded in engineering. Results from simulations and experiments show that the practical compensation using left inverse model based on RBF-NN in dynamic thrust measuring system can yield high tracking accuracy than the conventional methods.     

基于区域的卷积神经网络在空对地车辆检测中的应用

针对传统空对地车辆检测算法在光照变换、场景变化时检测效果不佳的问题,提出了基于区域卷积神经网络Faster RCNN模型的空对地车辆检测方法,介绍了Faster RCNN模型以及模型训练过程.实验结果表明,基于Faster RCNN的空对地车辆检测方法是可行的,对不同光照和场景下的车辆检测可以取得较好的效果.     

地基深空网发射机现状及未来发展

我国未来深空探测工程的发展对地基深空站发射机提出了更高的要求,不仅仅需要具有更高的输出功率,同时需要解决发射机高频谱纯度、高效率以及上行功率合成等。针对这些要求,对国内外地基深空网的组成和深空站高功率发射机现状以及未来发展趋势进行了研究,分析了速调管发射机与固态发射机在高功率输出条件下的优势,在此基础上,提出了我国深空站速调管发射机需要开展高功率连续波速调管、高功率低损耗滤波器、低纹波高功率开关电源、上行功率合成等关键技术,并为相应关键技术提供了理论建议及相应的解决方法,这可以为我国未来深空探测工程高功率发射机研制提供技术支撑,为后续地基深空网建设和发展提供参考。     

基于PCA-TN的直升机主减速器通用热分析计算

针对当前热分析方法无法满足直升机主减速器结构复杂多变的问题,提出了一种快速获取主减速器温度场分布的通用热分析方法。以典型直升机主减速器为样本,进行了主成分分析（principal component analysis,PCA）,并结合热网络法（thermal network,TN）,分析获得3种通用热分析单元体模型。对于不同结构的主减速器,通过单元体及其衍生体的组合可以迅速且有效地建立其系统热分析模型,求解获得温度场分布。以某直升机主减速器为例,应用该方法进行了热分析,并与试验进行了对比,结果表明:该方法能高效计算主减速器的温度场分布,计算值与试验值最大误差为5.07%,满足主减速器热分析工程计算需求。      

基于IPSO Elman神经网络的航空发动机故障诊断

为提高航空发动机故障诊断的精度,提出改进粒子群优化的Elman神经网络对航空发动机故障诊断的方法。利用MIV(平均影响值)对神经网络的输入端自变量进行筛选,降低输入维度;采用改进粒子群优化算法对Elman神经网络的权值和阀值进行优化,并对优化的神经网络进行训练;用训练好的神经网络对航空发动机故障进行诊断并与常规的BP（back propagation）、Elman神经网络、GM（1,n）、SVM (support vector machines)进行对比。仿真结果表明:IPSO Elman(improved particle swarm optimization Elman neural network)神经网络的诊断误差在不同数量训练样本时都小于其他方法,并且在参选故障诊断的性能参数不同时,其诊断误差相近,展现出较强的适应能力。      

基于DSAS算法的直升机贴地飞行在线航迹规划

通过改进优化传统A*算法,利用动态稀疏A*搜索(DSAS)算法在线设计了直升机贴地飞行轨迹.该算法采用逆向搜索方式规划航迹,在遇到新生探测威胁时,只需局部调整受到影响的航迹,减少了重新规划的范围,提高了搜索效率.此外,在生成节点时考虑了直升机性能及飞行约束,优化了搜索范围;并且根据影响贴地飞行航迹性能的各种因素,设计了航迹代价计算方法;利用层次分析法结合专家分析计算得到最优代价权值.仿真结果表明,该方法在线设计的飞行轨迹能够较全面地满足避障、贴地功能,相比稀疏A*搜索(SAS)算法节省了计算时间.     

基于RBF带学习能力高精度非线性FDD设计

在轨航天器故障检测与诊断问题需要面对模型的非线性,而且要求尽量提高其检测的精度,为此设计了基于径向基函数（RBF）神经网络的动量轮非线性故障检测与诊断（FDD）方案。首先应用RBF补偿建模误差,提高检测精度,并选择李雅普诺夫函数证明其收敛性;然后应用非线性观测器来产生故障残差,给出了阈值以及故障检测的时间;应用RBF网络对故障信号进行重构,并据此设计了带学习能力的FDD策略。再次建立了详细的动量轮模型,通过不同条件下的仿真研究分别验证残差的阈值特性、时间特性以及RBF的重构能力,仿真结果表明了算法的有效性。     

基于相关分析和组合神经网络的退化预测

产品剩余寿命预测是加速退化试验和故障预测与健康管理两大热点领域中的关键技术之一.为了解决复杂退化的预测问题,提出了一种新型预测方法,对退化轨迹能够实现较长距离的预测.此方法首先对复杂退化数据进行小波变换,通过Durbin-Watson方法和偏相关图分析各级分解序列的自相关性,最后根据序列的特点,组合BP(Back Propagation)和小波神经网络对退化轨迹进行预测.为了验证所提组合神经网络方法的有效性,采用小波神经网络的预测结果进行对比分析.实际退化数据的预测结果表明,所提方法比单独采用小波神经网络,具有更小的均方差(MSE,Mean Square Error),对剩余寿命(RUL,Remaining Useful Life)也具有更高的预测精度.     

空间相机热平衡试验温度预测方法改进及应用

为实现对航天器热平衡试验平衡温度的预测,建立了热平衡温度预测的方程式,对影响预测结果的各参数进行了分析。首先,由节点网络法推导出平衡温度表达式,介绍了其基于非线性最小二乘法的求解过程,建立了平衡温度的迭代方程;其次,结合大量的试验数据,对影响迭代结果的3个因素进行了统计分析,并最终确定了各因素的合理取值范围;最后,利用温度预测程序对某空间相机热平衡试验中低温工况的平衡温度进行了预测,并与试验结束后的结果进行了对比。结果表明,预测温度与试验数据的误差为±1℃（在3%之内）,满足精度要求,对热平衡温度准确的预测有效的缩短了试验时间,节省试验费用。     

Optimization of communication network topology for navigation sharing among distributed satellites

Navigation sharing among distributed satellites is quite important for coordinated motion and collision avoidance. This paper proposes optimization methods of the communication network topology to achieve navigation sharing. The whole communication network constructing by inter-satellite links are considered as a topology graph. The aim of this paper is to find the communication network topology with minimum communication connections’ number (MCCN) in different conditions. It has found that the communication capacity and the number of channels are two key parameters affecting the results. The model of MCCN topology for navigation sharing is established and corresponding method is designed. Two main scenarios, viz., homogeneous case and heterogeneous case, are considered. For the homogeneous case where each member has the same communication capacity, it designs a construction method (Algorithm 1) to find the MCCN topology. For the heterogeneous case, it introduces a modified genetic algorithm (Algorithm 2) to find the MCCN topology. When considering the fact that the number of channels is limited, the Algorithm 2 is further modified by adding a penalized term in the fitness function. The effectiveness of these algorithms is all proved in theoretical. Three examples are further tested to illustrate the methods developed in this paper.