A Novel Radius Adaptive Based on Center-Optimized Hybrid Detector Generation Algorithm

Negative selection algorithm(NSA)is one of the classic artificial immune algorithm widely used in anomaly detection.However,there are still unsolved shortcomings of NSA that limit its further applications.For example,the nonselfdetector generation efficiency is low;a large number of nonselfdetector is needed for precise detection;low detection rate with various application data sets.Aiming at those problems,a novel radius adaptive based on center-optimized hybrid detector generation algorithm(RACO-HDG)is put forward.To our best knowledge,radius adaptive based on center optimization is first time analyzed and proposed as an efficient mechanism to improve both detector generation and detection rate without significant computation complexity.RACO-HDG works efficiently in three phases.At first,a small number of self-detectors are generated,different from typical NSAs with a large number of self-sample are generated.Nonself-detectors will be generated from those initial small number of self-detectors to make hybrid detection of self-detectors and nonself-detectors possible.Secondly,without any prior knowledge of the data sets or manual setting,the nonself-detector radius threshold is self-adaptive by optimizing the nonself-detector center and the generation mechanism.In this way,the number of abnormal detectors is decreased sharply,while the coverage area of the nonself-detector is increased otherwise,leading to higher detection performances of RACOHDG.Finally,hybrid detection algorithm is proposed with both self-detectors and nonself-detectors work together to increase detection rate as expected.Abundant simulations and application results show that the proposed RACO-HDG has higher detection rate,lower false alarm rate and higher detection efficiency compared with other excellent algorithms.

A Novel Stochastic Blind Adaptive Multiuser Detector for CDMA Systems

Blind adaptive multiuser detector has become a research hotspot in recent years due to a number of advantages, but many blind adaptive algorithms involve low convergence rate. This paper presents a novel stochastic blind adaptive multiuser detector without requiring training sequences, which needs only two system parameters: the signature sequence of the desired user i, s i and the variance of the additive white Gaussian noise (AWGN),σ 2. Simulation results show that by reasonably choosing time varying step size, the proposed detector can not only improve the convergence rate, but also reduce the limiting NSE (Normalized Squared Error) values, so it can effectively increase the performance of the system.

A Neuro-inspired Adaptive Motion Detector

In this paper, a novel motion detector is proposed to perceive the weak changes in a image sequence. This is inspired by the mechanism of fixational eye movement and dynamics of vertebrate’s cortex. We realized respectively an artificial model of visual attention selection, called dual-probe adaptive model (DPAM), and an active tremor operation (ATO) approach. It is found that between them there exists a resonance phenomenon. The phenomenon is enhanced when the ATO and the DPAM are in-phase and is suppressed when they are anti-phase.?Based on this, we construct a novel motion detector combined by the ATO and the DPAM to resonate with the motion direction. This allows capturing moving edges even in the image sequences with lighting change and noisy background. Simulation and Experimental results demonstrate the effectiveness.

Robust adaptive radar beamforming based on iterative training sample selection using kurtosis of generalized inner product statistics

In engineering application,there is only one adaptive weights estimated by most of traditional early warning radars for adaptive interference suppression in a pulse reputation interval(PRI).Therefore,if the training samples used to calculate the weight vector does not contain the jamming,then the jamming cannot be removed by adaptive spatial filtering.If the weight vector is constantly updated in the range dimension,the training data may contain target echo signals,resulting in signal cancellation effect.To cope with the situation that the training samples are contaminated by target signal,an iterative training sample selection method based on non-homogeneous detector(NHD)is proposed in this paper for updating the weight vector in entire range dimension.The principle is presented,and the validity is proven by simulation results.

A new detector in EBPSK communication system

In order to raise the detection precision of the extended binary phase shift keying （EBPSK） receiver, a detector based on the improved particle swarm optimization algorithm （IMPSO） and the BP neural network is designed. First, the characteristics of EBPSK modulated signals and the special filtering mechanism of the impacting filter are demonstrated. Secondly, an improved particle swarm optimization algorithm based on the logistic chaos disturbance operator and the Cauchy mutation operator is proposed, and the EBPSK detector is designed by utilizing the IMPSO-BP neural network. Finally, the simulation of the EBPSK detector based on the MPSO-BP neural network is conducted and the result is compared with that of the adaptive threshold-based decision, the BP neural network, and the PSO-BP detector, respectively. Simulation results show that the detection performance of the EBPSK detector based on the IMPSO-BP neural network is better than those of the other three detectors.

Adaptive detection in the presence of signal mismatch

The problem of adaptive detection in the situation of signal mismatch is considered; that is, the actual signal steering vector is not aligned with the nominal one. Two novel tunable detectors are proposed. They can control the degree to which the mismatched signals are rejected. Remarkably, it is found that they both cover existing famous detectors as their special cases. More importantly, they possess the constant false alarm rate（CFAR）property and achieve enhanced mismatched signal rejection or improved robustness than their natural competitors. Besides, they can provide slightly better matched signals detection performance than the existing detectors.

Research on monopulse forward-looking high-resolution imaging algorithm based on adaptive iteration

In this paper,we proposed a monopulse forward-looking high-resolution imaging algorithm based on adaptive iteration for missile-borne detector.Through iteration,the proposed algorithm automatically selects the echo signal of isolated strong-scattering points from the receiving echo signal data to accurately estimate the actual optimal monopulse response curve(MRC) of the same distance range,and we applied optimal MRC to realize the azimuth self-focusing in the process of imaging.We use real-time echo data to perform error correction for obtaining the optimal MRC,and the azimuth angulation accuracy may reach the optimum at a certain distance dimension.We experimentally demonstrate the validity,reliability and high performance of the proposed algorithm.The azimuth angulation accuracy may reach up to ten times of the detection beam-width.The simulation experiments have verified the feasibility of this strategy,with the average height measurement error being 7.8%.In the out-field unmanned aerial vehicle(UAV) tests,the height measurement error is less than 25 m,and the whole response time can satisfy the requirements of a missile-borne detector.

ANALYSIS OF CELL-AVERAGING BASED DETECTORS FOR χ^2 FLUCTUATING TARGETS IN MULTITARGET ENVIRONMENTS

The χ^2 family of signal fluctuation distributions represents the main fluctuation models which most radar targets follow it in their reflections. This family can be categorized as fluctuation distribution with two degrees of freedom and those with four degrees of freedom. The first category represents all important class of fluctuation models which when illuminated by a coherent pulse train, return a train of fully correlated pulses （Swerling Ⅰ model） or fully decorrelated pulses （Swerling Ⅱ model）. The detection of this type of fluctuating targets is therefore of great importance. This paper is devoted to the analysis of Cell-Averaging （CA） based detectors for the case where the radar receiver noncoherently integrates M square-law detected pulses and the signal fluctuation obeys 2 statistics with two degrees of freedom. These detectors include the Mean-Of （MO）, the Greatest-Of （GO） and the Smallest-Of（SO） schemes. In these processors, the estimation of the noise power levels from the leading and the trailing reference windows is based on the CA technique. Exact formulas for the detection probabilities are derived, in the absence as well as in the presence of spurious targets. The primary and the secondary interfering targets are assumed to be fluctuating in accordance with the χ^2 fluctuation model with two degrees of freedom （SWI ＆ SWII）. The numerical results show that the MO version has the best homogeneous performance, the SO scheme has the best multiple-target performance, while the GO procedure does not offer any merits, neither in the absence nor in the presence of outlying targets.

High-resolution forward-looking imaging algorithm for missile-borne detectors

Aiming at a novel missile-borne detector in the optional burst height proximity fuze, a self-adaptive high-resolution forward-looking imaging algorithm (SAHRFL-IA) is presented. The echo data are captured by the missile-borne detector in the target regions;thereby the azimuth angulation accuracy at the same distance dimension is improved dynamically. Thus, azimuth information of the targets in the detection area may be obtained accurately. The proposed imaging algorithm breaks through the conventional misconception of merely using azimuth discrimination curves under ideal conditions during monopulse angulation. The real-time echo data from the target region are used to perform error correction for this discrimination curve, and finally the accuracy of the azimuth angulation may reach the optimum at the same distance dimension. A series of experiments demonstrate the validity, reliability and high performance of the proposed imaging algorithm. Azimuth angulation accuracy may reach ten times that of the detection beam width. Meanwhile, the running time of this algorithm satisfies the requirements of missile-borne platforms.

New blind adaptive approach to multiuser detection in CDMA systems

Due to a number of advantages, blind adaptive multiuser detector has become a research hotspot in recent years. But low convergence rate problem occurs to many blind adaptive algorithms. A new blind adaptive approach to multiuser detection is presented. The simulation results show that by reasonably choosing time varying step size, the proposed detector can not only improve the convergence rate, but also reduce the bit error rate (BER) of the system, and so it can effectively improve the system performance with less computational cost.

多尺度增强特征融合的钢表面缺陷目标检测

针对轻量级目标检测算法在钢表面缺陷检测任务中识别精度低的问题,提出一种多尺度增强特征融合的钢表面缺陷目标检测算法。该算法采用提出的自适应加权融合模块为不同层级特征自适应计算融合权重,将深层语义与浅层细节进行加权融合,使得浅层特征在不丢失细节信息的同时获得丰富的深层语义。利用提出的空间特征增强模块从3个独立方向强化融合特征,通过引出残差旁路增强网络结构的稳定性,使卷积过程能够挖掘到更多的关键信息。根据先验框与真实框的整体交并程度为模型选择更为合适的训练样本。实验结果表明,该算法的检测精度达到80.47%,相比原始算法提升6.81%。该算法的参数量为2.36 M,计算量为952.67 MFLOPs,能快速且高精度检测钢材表面的缺陷信息,具有较高的应用价值。

基于P-U-net的角锥波前探测器的波前校正方法

自适应光学技术被广泛应用于像差的校正,但在实际应用中,闭环校正需要较长的时间,并且由于变形镜的非线性响应,传统的开环校正方法的校正精度会受到影响。本文提出一种基于P-U-net网络的角锥波前探测器波前校正方法。P-U-net通过数据训练建立了整个自适应光学系统的模型,在角锥波前探测器的光瞳图像与变形镜的控制电压之间构建了直接的非线性映射关系。可以通过角锥波前探测器的光瞳图像直接计算变形镜的控制电压。通过使用仿真自适应系统理论验证了该方法的可行性,可以在50 ms内完成像差校正,校正精度达到0.1μm。与传统的闭环自适应校正算法相比,本文方法可以在更短的时间内完成像差校正,达到更高的精度,具有很好的应用前景。

CMOS探测器自适应信号读出技术研究

设计了一种根据光照强度自动调节增益的CMOS探测器,它能够在一帧时间内实现光照强度与积分增益自动适应,从而实现无论在弱光还是强光条件下CMOS探测器都能有适应的灵敏度和动态范围。相较于传统CTIA电路,自适应信号读出技术新增了比较器电路来控制CTIA积分电容大小,通过短曝光输出电压与参考阈值进行比较,输出信号结果用来调整长曝光积分增益,最终得到每个像素的输出电压和增益挡位。基于0.5μm 5 V-CMOS工艺进行了128×1线阵CMOS探测器设计仿真与流片,仿真结果表明,光电流在2 pA~100 nA六个数量级内分别自适应四个积分增益,都能有良好的信号读出。

基于自适应加权增强总变差的CT偏置扫描重建算法

基于降低设备制造成本或辐射剂量等目的,计算机断层成像(Computer Tomography,CT)中的一个实际需求是以有限的探测器尺寸来获得更大的视野(Field of View,FOV),通过将探测器放置在横向偏移位置可以有效的扩大FOV。然而,常规的重建算法无法精确重建偏置投影数据,针对这一问题,本文提出了一种基于自适应加权增强总变差最小化的偏置重建模型及CP(Chambolle-Pock)求解算法。具体来说,构建自适应加权增强总变差范数作为正则项,其中自适应权重根据局部增强梯度自适应调整权值,进而设计了一种基于自适应加权增强总变差最小化的偏置重建模型(Weighted Adaptive-weight reinforced Total Variation,WAwrTV),并推导出了相应的CP算法。实验结果表明,所提算法能有效的重建偏置投影数据并提高重建精度,且具有良好的抗噪性能。

一种基于QoS的自适应网格失效检测器

失效检测器是构建可靠的网格计算环境所必需的基础组件之一.由于网格中存在大量对失效检测有着不同QoS需求的分布式应用,对于一个网格失效检测器来说,为保持其有效性和可扩展性,应该既能够准确提供应用程序所需的失效检测QoS,又能够避免为满足不同QoS而设计多套失效检测器所产生的多余负载.基于QoS基本评价指标(TDU,TMLR,TMU),采用PULL模式主动检测策略实现了一种新的失效检测器——GA-FD(adaptive failure detector for grid),可以同时支持多个应用程序定量描述的QoS需求,不需要关于消息行为和时钟同步的任何假设.同时,证明了GA-FD在部分同步模型下可实现一个◇P类的失效检测器,并给出了相应的实验及数据.

一种改进的阴性选择免疫算法

针对阴性选择免疫算法中字符串空间收敛效果差以及运行成本高的缺点,依据Forrest阴性选择算法提出了一种新的改进算法。该算法基于模式概念建立检测器库,检测器库存储了所有与自体不匹配的字符串个数,使得所占用的空间比所有与自体不匹配的字符串所占用的空间显著减少。采用了一种新的检测失败率算法公式,使得检测器的个数能够自适应自体规模的变化。理论分析与仿真结果表明,该算法与原算法相比,具有更快的运行速度和更低的检测失败率。

一种带功率检测和自适应偏置的CDMA功率放大器

设计了一款包含功率检测和自适应线性化偏置电路的CDMA功率放大器,功率检测器能根据输入信号的大小来调整功率管的偏置点,大幅提升低功率输出时的效率,从而提升系统整体效率;自适应线性化偏置能有效抑制功率放大器的增益压缩和相位失真,改善其线性度。采用2μm InGaP/GaAs HBT晶体管工艺成功流片,测试结果表明,与普通功率放大器相比较,当输出功率小于10 dBm时,新型功率放大器的集电极电流大幅降低,系统整体效率提高了2.2倍;1 dB增益压缩点由30 dBm增加到32 dBm;当输出功率为28 dBm时,ACPR和ALTR分别改善了5 dB和12 dB,IM3改善了11 dB。

基于自适应超环检测器的设备异常度检测方法

针对在无故障样本情况下如何快速检测设备异常度问题,在引入自己空间边界样本概念的基础上,提出一种自适应超环检测器。在描述自适应超环检测器生成算法的基础上,以Iris数据集为例进行分析,发现与已有的异常检测方法相比,自适应超环检测器异常检测方法在区分有较清晰类边界数据时,具有更好检测性能。利用自适应超环检测器异常检测方法分析轴承状态数据,不仅能反映出轴承的各种状态,而且能通过设备的异常程度反映出同类故障的轻重程度。基于自适应超环检测器的设备异常度检测方法,是在学习设备正常运行数据的基础上,寻找自己空间的边界样本,结合其方位信息与自己样本半径,建立能完全覆盖状态空间的自适应超环检测器,不需要设备运行的故障数据,它适合对故障数据缺乏的设备进行有效的异常检测。

红外焦平面阵列非均匀性校正技术的最新进展

面列阵红外焦平面探测器的非均匀性校正技术是正在探索的一项关键技术。介绍了近期国内外发展的红外焦平面阵列非均匀性校正技术:基于场景的代数算法、基于干扰抵消原理的自适应校正法和基于低次插值的多点校正法。