用于相异性设计的模糊重用库

容错设计是提高软件可靠性的有效途径，但是由于设计多样性实现很困难且开销大，导致该项技术实现很困难。为此，提出了一个基于重用的应用框架。该框架的提出将相异性设计过程转变为根据可重用参数从大量具有相同功能软件组件中选择合适的组件来构成容错系统，从而简化了软件开发过程。同时为了更准确地描述组件，采用了模糊逻辑来表达其与上层概念的关联关系。本文介绍了重用库系统结构的设计及其对容错设计的支持。     

基于意图推断的高超声速滑翔目标贝叶斯轨迹预测

为解决高超声速滑翔目标机动给探测防御造成的困难,研究基于意图推断的贝叶斯轨迹预测方法。首先对目标进行动力学建模,利用气动参数设计机动模式集。假定高超声速滑翔目标必定攻击某目标,结合飞行意图合理构造意图代价函数,借鉴贝叶斯理论迭代推导机动模式和运动状态递推公式。通过蒙特卡洛采样实现轨迹预测算法。仿真结果表明,算法能有效提高目标机动不确定条件下轨迹预测的精度,当多目标可能被打击时,通过对预测轨迹进行在线重新规划,降低目标误判给轨迹预测造成的不利影响。     

基于广义回归神经网络集成的宽带波束形成算法

首先构造宽带波束成形所需要的协方差矩阵,利用基于粒子群优化算法对核主成分分析方法和广义回归神经网络进行了优化。在对神经网络的输入变量进行降维处理后,生成多个复杂度低的泛回归神经网络模型。利用提出的基于聚类启发式集成算法求出波束成形时的权系数,既考虑了网络的差异性,又考虑了网络的正确性。仿真结果表明,提出的基于聚类启发式神经网络集成的波束形成算法在网络结构十分简单的情况下,仍然具有较好的性能。     

Topology tracking of dynamic UAV wireless networks

Wireless network is the communication foundation that supports the intelligentization of Unmanned Aerial Vehicle(UAV) swarm. The topology of UAV communication network is the key to understanding and analyzing the behavior of UAV swarm, thus supporting the further prediction of UAV operations. However, the UAV swarm network topology varies over time due to the high mobility and diversified mission requirements of UAVs. Therefore, it is important but challenging to research dynamic topology infere...     

基于专家系统的机场停机位自动分配系统研究

民航机场停机位分配是民航机场运营组织的关键环节.根据专家系统的基本原理,设计出了基于规则的机场停机位自动分配系统,并对该系统的知识库和推理机的工作机理进行了较深入探讨,以期为民航机场停机位分配提供先进的技术手段.     

Takagi-Sugeno fuzzy model identification for turbofan aero-engines with guaranteed stability

This paper is concerned with identifying a Takagi-Sugeno (TS) fuzzy model for turbofan aero-engines working under the maximum power status (non-afterburning). To establish the fuzzy system, theoretical contributions are made as follows. First, by fixing antecedent parameters, the estimation of consequent parameters in state-space representations is formulated as minimizing a quadratic cost function. Second, to avoid obtaining unstable identified models, a new theorem is proposed to transform the prior-knowledge of stability into constraints. Then based on the aforementioned work, the identification problem is synthesized as a constrained quadratic optimization. By solving the constrained optimization, a TS fuzzy system is identified with guaranteed stability. Finally, the proposed method is applied to the turbofan aero-engine using simulation data generated from an aerothermodynamics component-level model. Results show the identified fuzzy model achieves a high fitting accuracy while stabilities of the overall fuzzy system and all its local models are also guaranteed.     

基于模糊聚类非负矩阵分解的软件缺陷预测

为了进一步提高对不平衡软件缺陷数据的预测精度,提出了一种基于模糊聚类非负矩阵分解的缺陷预测方法。该方法首先选取最近邻算法作为不平衡缺陷数据预测的基分类器,然后将非负矩阵分解算法引入软件缺陷预测领域,并提出利用模糊聚类初始化非负矩阵分解算法的思路。该方法一方面解决了最近邻算法处理多属性缺陷数据时计算量大导致性能降低的问题,另一方面克服了非负矩阵分解的随机初始化导致结果陷入局部最优的不足,提高了不平衡软件缺陷数据的预测精度。最后通过两组算例验证了方法的有效性。     

飞机机电系统PHM的综合诊断推理机设计

在分析区域级PHM功能及区域管理器结构的基础上,开展了综合诊断推理机设计基础准备工作,包括机电系统功能建模、aFMECA分析、机电系统关联分析。最终,设计出飞机机电系统PHM综合诊断推理机的结构及其工作流程。所设计的综合诊断推理机综合了以往多种推理机的推理机制,更加系统全面地实现了对飞机机电系统的综合诊断功能,具有一定的工程应用价值。     

Attitude tracking control for variable structure near space vehicles based on switched nonlinear systems

An adaptive robust attitude tracking control law based on switched nonlinear systems is presented for a variable structure near space vehicle (VSNSV) in the presence of uncertainties and disturbances. The adaptive fuzzy systems are employed for approximating unknown functions in the flight dynamic model and their parameters are updated online. To improve the flight robust performance, robust controllers with adaptive gains are designed to compensate for the approximation errors and thus they have less design conservation. Moreover, a systematic procedure is developed for the synthesis of adaptive fuzzy dynamic surface control (DSC) approach. According to the common Lyapunov function theory, it is proved that all signals of the closed-loop system are uniformly ultimately bounded by the continuous controller. The simulation results demonstrate the effectiveness and robustness of the proposed control scheme.     

模糊集与故障树复合分析法研究及其在飞行器模拟训练系统中的应用

在传统的故障树分析中 ,类似人为错误的模糊、不精确事件不能得到有效地处理。为了克服这一不足 ,本文运用模糊集评价和概率估计提出了一种故障复合分析方法来评价非常事件。将此方法运用到飞行器模拟训练系统中可有效地对该系统进行故障分析。