ID 原文 译文
55647 最后,通过空地战场威胁态势评估模型,证明在参数学习过程中考虑知识的不确定性可有效改善先验知识错误时的BN参数学习效果。 Finally, by clearing the battlefield threat situation assessment model, and prove that considering the uncertainty of knowledge in the process of parameter learning can effectively improve the prior knowledge mistake of BN parameter learning.
55648 针对大包线下变体飞行器在翼型快速变化过程中的姿态控制问题,提出了一种基于速度线性化的鲁棒线性变参数(linear parameter varying,LPV)控制方法。 For parcel off-line variant aircraft in airfoil attitude control problem in the process of rapid change, this paper proposes a speed based on the linear robust linear parameter varying (linear parameter varying, LPV) control method.
55649 所提方法将状态量的微分增广至状态方程,得到系统任意状态下的线性化模型,解除了传统小扰动线性化方法要求系统处于平衡点邻域的限制。 Proposed method will state the amount of differential augmented to state equation, get system under arbitrary linear model, lifted the small disturbance linearization approach requires that traditional system in limit equilibrium neighborhood.
55650 建立了具有凸多胞形式的变体飞行器LPV模型,给出了参数依赖鲁棒H∞状态反馈控制器存在的充分条件,依据顶点控制器和变参数进行凸插值得到鲁棒LPV控制器。 Convex set up many variations in the form of aircraft LPV model, gives the parameter dependent robust h-infinity sufficient conditions for the existence of state feedback controller, according to the vertex convex interpolation, robust controller and variable parameters of LPV controller.
55651 仿真结果表明,设计的控制器在飞行器快速变形和参数大范围快速变化的情况下仍具有良好的控制性能和鲁棒性。 Simulation results show that the design of the controller in vehicle fast deformation and parameters under the condition of a wide range of rapidly changing still has good control performance and robustness.
55652 为了减少边缘轮廓上微小变化和边缘噪声对检测结果的影响,提高检测准确率,提出了一种利用边缘轮廓上多尺度自相关矩阵的角点检测算法。 In order to reduce the edge profile on small changes and the impact of noise on edge detection results, improve the detection accuracy, an edge contour on multi-scale autocorrelation matrix of corner detection algorithm.
55653 该算法首先利用边缘检测器提取图像的轮廓,同时利用不同尺度的高斯核函数平滑图像,对于轮廓曲线上的每一个像素,利用微分算子获得不同尺度下灰度变化信息来构建自相关矩阵; The algorithm firstly edge detector is utilized to extract the contours of the image at the same time by using the different scales of gaussian kernel function is a smooth image, for each pixel of the contour curve, using the differential operator for the gray level change information under different scale to build the autocorrelation matrix;Finally to different scales of normalized autocorrelation matrix eigenvalue of the product as a new angular point measurement.
55654 最后将不同尺度下自相关矩阵的归一化特征值的乘积作为新的角点测度。 If the Angle point measure value is greater than the preset threshold, and the maximum value of local window, is to determine angular point.
55655 如果该角点测度值既大于预设阈值,又是局部窗口内的极大值,就判定是角点。仿真结果证实所提算法可以更准确地检测图像中的角点且具有更低的错检概率。 Simulation results prove the proposed algorithm can more accurately detect the image of the angular point and has lower fault detection probability.
55656 D-S(Dempster-Shafer)证据理论能够有效地处理不确定信息,在信息融合领域得到了广泛的应用。 (Dempster - Shafer) D - S evidence theory can efficiently deal with uncertain information, has been widely used in the field of information fusion.