Multi-scale traffic vehicle detection based on faster R-CNN with NAS optimization and feature enrichment

It well known that vehicle detection is an important component of the field of object detection. However, the environment of vehicle detection is particularly sophisticated in practical processes. It is compara-tively difficult to detect vehicles of various scales in traffic scene images, because the vehicles partially obscured by green belts, roadblocks or other vehicles, as well as influence of some low illumination weather. In this paper, we present a model based on Faster R-CNN with NAS optimization and feature enrichment to realize the effective detection of multi-scale vehicle targets in traffic scenes. First, we proposed a Retinex-based image adaptive correction algorithm (RIAC) to enhance the traffic images in the dataset to reduce the influence of shadow and illumination, and improve the image quality. Second, in order to improve the feature expression of the backbone network, we conducted Neural Architecture Search (NAS) on the backbone network used for feature extraction of Faster R-CNN to generate the optimal cross-layer connection to extract multi-layer features more effectively. Third, we used the object Feature Enrichment that combines the multi-layer feature information and the context information of the last layer after cross-layer connection to enrich the information of vehicle targets, and improve the robustness of the model for challenging targets such as small scale and severe occlusion. In the imple-mentation of the model, K-means clustering algorithm was used to select the suitable anchor size for our dataset to improve the convergence speed of the model. Our model has been trained and tested on the UN-DETRAC dataset, and the obtained results indicate that our method has art-of-state detection performance.     

基于二进小波变换的噪声图像自适应增强新算法

针对图像增强算法通常会放大原图像中噪声分量的问题,提出了一种基于二进小波变换的图像增强新算法。该算法充分利用了二进小波变换的平移不变性和各尺度上小波系数间的相关性,有效改善了增强过程中噪声放大和边缘失真问题。此外,算法具有高度的自适应能力,适用性更强。实验结果表明,与目前已有的各类多尺度图像增强算法相比,该算法在抑制噪声和凸显图像特征两方面均有明显改进。     

型号项目工期风险管理决策模型与自适应遗传算法

针对型号项目的过程特点,建立了型号项目工期风险的管理决策模型。模型考虑了型号项目中活动重叠、活动迭代、活动执行时间的不确定性和可更新资源总量等重要的工期风险影响因素。从模型的特点出发,给出了问题求解的基于自适应遗传算法的仿真优化方法。算例显示,该算法能较好地求解本文的工期风险管理决策问题。     

一种适用于非合作源雷达的波束形成技术

非合作源雷达要完成对运动目标的检测,必须能够有效地抑制强烈的直达波信号。常规的波束形成方法在存在阵元误差的情况下,其性能将显著下降。提出的数值式波束形成算法既可以实现误差校正,又能在已知的干扰方向上形成较深的零陷,且能保持主瓣的形状和限制副瓣的响应。     

基于期望系统噪声的自适应机动目标跟踪研究

状态估计是目标跟踪中的基本问题,也是目标跟踪的一个难点。首先对标准IMM算法的优缺点进行了论述,针对其缺点和不足,提出了基于期望系统噪声模型（MIMM）的自适应多模型算法,该算法能有效地对机动目标的状态进行自适应估计。仿真结果表明,该算法比标准的IMM算法有较好的改善。     

基于二阶锥规划的MIMO雷达稳健自适应波束形成

针对MIMO雷达自适应波束形成中期望目标导向矢量的失配问题,提出了一种基于二阶锥规划(SOCP)的稳健自适应波束形成算法。该算法首先将1个MN维(M,N分别为发射和接收阵元数)的权矢量分解成2个低维(1个M维和1个N维)权矢量的Kronecker积,然后分别限制实际的目标发射导向矢量和目标接收导向矢量与假定的导向矢量之间的误差范数的边界,通过优化最差性能,利用SOCP求得分解后的2个权矢量,最后再合成原权矢量。通过降维处理,算法在保证波束形成器性能的基础上,有效地降低了运算复杂度。仿真结果验证了算法的有效性。     

用于目标跟踪的自适应广义调频波形设计算法

提出了一种用于雷达目标跟踪的自适应广义调频波形设计算法。该算法根据跟踪器的动态需求,以广义调频信号为样板波形自适应设计下一时刻的发射波形,其目的是使预测的目标跟踪均方误差最小化,并假定高信噪比条件,且目标跟踪运动模型和观测模型均为线性。利用与某一波形相对应的克拉美-罗下限(CRLB)以及卡尔曼滤波器,通过最小化预测的跟踪均方误差来实现广义调频波形的自适应设计。仿真结果表明:在信噪比相同的情况下,与使用固定参数、自适应参数的线性调频波形设计算法相比,所提出的算法能够获得更低的目标跟踪均方误差。     

基于直接数据域自适应算法的相干信号DOA估计

利用直接数据域自适应算法的稳态权值代替噪声子空间构建空间谱,构造了一种超分辨波达方向(DOA)估计方法.为了解决谱峰搜索时的伪峰问题,采用参考阵元轮换得到多组线性无关稳态权值,逼近噪声子空间,能有效去除伪峰.针对相干信号的DOA估计,进一步提出了直接数据域取对称共轭向量的解相干方法.相比子空间分解类算法,本文算法不需估计信号源数目和协方差矩阵、不需特征分解,复杂度仅为O(MP),同时能有效完成解相干处理.     

Minimum deterrence and missile defenses: what's new,what's true,what's arguable

Minimum deterrence is a compromise, or halfway house, between nuclear abolition or nearly zero and assured destruction, the dominant paradigm for strategic nuclear arms control during and after the cold war. Minimum deterrence as applied to the current relationship between the United States and Russia would require downsizing the numbers of operationally deployed long-range nuclear weapons to 1000, or fewer, on each side. More drastic bilateral Russian–American reductions would require the cooperation of other nuclear weapons states in making proportional reductions in their own arsenals. In addition, US plans for European-based and global missile defenses cause considerable angst in Russia and threaten to derail the Obama “reset” in Russian–American relations, despite the uncertainties about current and plausible future performances of missile defense technologies.     

一类不确定时滞混沌系统的全局鲁棒自适应神经网络同步控制器设计

研究一类不确定时滞混沌系统的全局鲁棒自适应神经网络同步控制器设计,其系统中的不确定时滞项不是简单的线性有界条件,而是允许其存在高阶项,因此具有全局特性．在控制器的设计上;首先通过选取合适的径向基函数（RBF）神经网络的权向量去逼近时滞系统中的未知连续有界部分;然后在RBF神经网络输出的基础上,选用一个鲁棒自适应控制器来趋近时滞系统的不确定部分;同时,利用Lyapunov稳定性理论对混沌同步的条件给出了论证;最后,数据仿真的结果表明该方法的有效性．