基于Gabor滤波器的图像目标识别方法

为了给机器人视觉导航提供有效信息,提出一种基于图像匹配的目标识别方法.对CCD采集的目标图像,由 Gabor 变换生成的二维Gabor 滤波器有着优良的滤波器性能,不需要对图像进行分割,能适应一定的旋转、尺度、光照的变化,通过多个频率和角度的Gabor算子与图像的卷积,获取图像全局信息的特征描述.分类方法采用统计学习理论基础上发展起来的一种新的机器学习方法——支持向量机(SVM, Support Vector Machines),它可解决模型选择、过学习、维数灾难等问题.通过支持向量机进行多维特征向量的分类.该方法可达到较高的识别率,达到实时处理的要求,可以在人脸识别、机器人视觉定位等领域得到应用.      

基于TDICCD相机有源低通滤波器的优化研究

文章通过分析低通滤波器的特性,指出航天TDICCD相机中有源低通滤波器的重要性,提出优化实现的可行性,并给出了具体原理电路及仿真结果。     

带BP神经网络的卡尔曼滤波算法在GPS/SINS组合导航系统中的应用

针对带权值调整的Kalman滤波器,运用L-M的BP算法,将BP神经网络嵌入该滤波器中,与BP神经网络滤波器相比,减小了层数,提高了网络训练速度及精度。以GPS/SINS组合导航系统为例进行了仿真,结果既能抑制滤波发散,又能提高滤波精度。     

基于Wash-out滤波器技术的机翼颤振控制

研究应用wash-out滤波器技术对具有立方非线性俯仰刚度的二元机翼颤振的控制。首先,确定需要引入Hopf分岔的点,并在该点将原系统方程Jordan化;其次,对于引入的wash-out滤波控制器,先按Hopf分岔条件确定线性控制增益,再用规范型直接法得到受控系统的规范型,由分岔类型与规范型系数的关系确定非线性控制增益,从而将原系统的亚临界Hopf分岔变为超临界Hopf分岔;最后通过数值模拟验证了控制的有效性,并发现受控系统的颤振幅值（极限环大小）大大降低。     

基于测量数据积累的弱目标检测前跟踪算法

检测前跟踪(TBD)算法是一种有效的弱目标联合检测与跟踪方法。当目标强度参数未知时,基于点目标假设的主流TBD算法通常将其建模为增广目标状态变量,在贝叶斯框架下联合估计目标运动状态与强度信息。本文研究发现强度参数估计精度受限于点扩散函数(PSF)的设置,提出了一种基于测量数据积累的弱目标TBD算法。该方法利用目标的航迹,积累多帧间目标占据分辨单元的测量数据,在实现弱目标的联合检测与跟踪的同时,对目标强度参数完成准确估计,为后续目标分类或识别提供有用的目标强度信息。仿真结果表明:在低信噪比条件下,所提算法的强度参数估计精度和稳健性优于主流TBD算法,并最终收敛于参数完全已知时的最优联合检测与跟踪结果。     

Event-triggered cooperative target tracking in wireless sensor networks

Since the issues of low communication bandwidth supply and limited battery capacity are very crucial for wireless sensor networks, this paper focuses on the problem of event-triggered cooperative target tracking based on set-membership information filtering. We study some fundamental properties of the set-membership information filter with multiple sensor measure-ments. First, a sufficient condition is derived for the set-membership information filter, under which the boundedness of the outer ellipsoidal approximation set of the estimation means is guaranteed. Second, the equivalence property between the parallel and sequential versions of the set-membership information filter is presented. Finally, the results are applied to a 1D event-triggered target tracking scenario in which the negative information is exploited in the sense that the measurements that do not satisfy the triggering conditions are modelled as set-membership mea-surements. The tracking performance of the proposed method is validated with extensive Monte Carlo simulations.     

Near real-time PPP-based monitoring of the ionosphere using dual-frequency GPS/BDS/Galileo data

Ionosphere delay is very important to GNSS observations, since it is one of the main error sources which have to be mitigated even eliminated in order to determine reliable and precise positions. The ionosphere is a dispersive medium to radio signal, so the value of the group delay or phase advance of GNSS radio signal depends on the signal frequency. Ground-based GNSS stations have been used for ionosphere monitoring and modeling for a long time. In this paper we will introduce a novel approach suitable for single-receiver operation based on the precise point positioning (PPP) technique. One of the main characteristic is that only carrier-phase observations are used to avoid particular effects of pseudorange observations. The technique consists of introducing ionosphere ambiguity parameters obtained from PPP filter into the geometry-free combination of observations to estimate ionospheric delays. Observational data from stations that are capable of tracking the GPS/BDS/GALILEO from the International GNSS Service (IGS) Multi-GNSS Experiments (MGEX) network are processed. For the purpose of performance validation, ionospheric delays series derived from the novel approach are compared with the global ionospheric map (GIM) from Ionospheric Associate Analysis Centers (IAACs). The results are encouraging and offer potential solutions to the near real-time ionosphere monitoring.     

基于改进GPC的无人直升机动态路径规划

针对小型无人直升机航迹受到风场扰动和导航系统定位误差影响下的航迹最优规划问题,建立了风扰动和导航误差影响下的飞行航迹线性模型。基于混合整数规划思想,采用粒子滤波改进广义模型预测控制算法实时规划航迹,降低了导航定位误差造成的航迹估计偏差,提高了航迹估计精度。仿真结果表明,该算法能够生成三维可行航迹并躲避障碍物,且规划时间和无人机能耗较传统方法显著减少。     

基于样条拟合和双向滤波的助推段弹道估计

三次B样条拟合在助推段弹道估计中应用广泛,但存在弹道两端点邻域估计误差过大的问题.通过对B样条基函数矩阵的特性进行分析,发现其原因为端点邻域估计时能利用的有效观测数据最少.因此,提出一种基于样条拟合和双向滤波相结合的弹道估计策略：首先将双星观测数据利用无迹变换转换为等效位置观测;然后利用三次B样条拟合弹道;最后基于一种自适应卡尔曼滤波器,采用双向滤波策略改善样条拟合的端点估计精度.仿真结果表明：组合方法使样条拟合的端点估计误差减少40％,同时使估计弹道更平滑.此结果满足实时性要求.     

微小型无人直升机姿态信号的互补滤波融合算法

基于惯性测量技术获取飞行器姿态信号的方法已被广泛应用,然而对于微小型或低成本无人飞行器而言,通过MEMS传感器获取满足精度要求且稳定可靠的姿态信号仍然是具有一定难度的工程问题。文章针对民用微小型无人直升机的低成本要求和使用环境,利用三轴角速率陀螺仪和加速度计在姿态估计中频率互补的特性,采用互补滤波算法,改进了滤波参数的自适应调整方法,实现了俯仰角和滚转角信号的有效融合,得到的姿态信息可用于微小型无人直升机的内环增稳控制。仿真和试验数据表明,该方法测量精度较高,工程实现简单,算法稳定,能有效解决微小型无人直升机姿态信号的工程求解和使用问题。