Adaptive learning approach to landmine detection

We consider landmine detection using forward-looking ground penetrating radar (FLGPR). The two main challenging tasks include extracting intricate structures of target signals and adapting a classifier to the surrounding environment through learning. Through the time-frequency (TF) analysis, we find that the most discriminant information is TF localized. This observation motivates us to use the over-complete wavelet packet transform (WPT) to sparsely represent signals with the discriminant information encoded into several bases. Then the sequential floating forward selection method is used to extract these components and thereby a neural network (NNW) classifier is designed. To further improve the classification performance and deal with the problem of detecting mines in an unconstraint environment, the AdaBoost algorithm is used. We integrate the feature selection process into the original AdaBoost algorithm. In each iteration, AdaBoost identifies the hard-to-learn examples and a new set of features which provide the specific discriminant information for these hard samples is extracted adaptively and a new classifier is trained. Experimental results based on measured data are presented, showing that a significant improvement on the classification performance can be achieved.     

Classification of radar targets using synthetic neural networks

Radar target classification performance of neural networks is evaluated. Time-domain and frequency-domain target features are considered. The sensitivity of the neural network algorithm to changes in network topology and training noise level is examined. The problem of classifying radar targets at unknown aspect angles is considered. The performance of the neural network algorithms is compared with that of decision-theoretic classifiers. Neural networks can be effectively used as radar target classification algorithms with an expected performance within 10 dB (worst case) of the optimum classifier     

A Sequential Approach to Target Discrimination

The Wald sequential probability ratio test is applied to the discrimination of targets observed by a radar or other sensor and a form for the classifier involving linear predictive filtering is developed. In this sequential approach, a target is illuminated with consecutive pulses until a classification of the target can be made to within a prescribed probability of error. Because of the linear-predictive formulation, the computational and storage requirements for the classifier are related only to the number of returns necessary to predict the target signature and not to the length of signature observed; a classifier with modest storage and computational requirements can be employed to process signatures consisting of an arbitrarily large number of returns. The classifier is based on some well-known results in mean-square filtering theory and has a simple intuitive interpretation. The classifier structure can also be related to autoregressive time series analysis and innovations process concepts and has an interpretation in the frequency domain in terms of the maximum entropy and maximum likelihood spectral estimates for the target signatures.     

Identification of ground targets from sequential high-range-resolution radar signatures

An approach to identifying targets from sequential high-range-resolution (HRR) radar signatures is presented. In particular, a hidden Markov model (HMM) is employed to characterize the sequential information contained in multiaspect HRR target signatures. Features from each of the HRR waveforms are extracted via the RELAX algorithm. The statistical models used for the HMM states are formulated for application to RELAX features, and the expectation-maximization (EM) training algorithm is augmented appropriately. Example classification results are presented for the ten-target MSTAR data set.     

基于序列图像的自动目标识别算法

由于利用单幅二维图像进行三维目标识别存在识别的多义性,提出了一种基于二维序列图像的三维目标自动识别算法。首先以修正的Hu不变矩构造目标的图像识别特征,进而采用BP神经网络分类器构造关于目标融合识别的基本置信指派函数,以神经网络的训练误差构造证据理论不确定性度量,采用基于吸收法的DS证据理论实现高冲突证据的贯序式融合。对各姿态飞机图像识别的仿真表明,该算法对飞机的空间姿态变化具有很强的鲁棒性,能快速地准确识别飞机类型。此外,算法对先验性参数具有一定的鲁棒性。     

Softmax分类器深度学习图像分类方法应用综述

基于深度学习的人工智能图像分类方法研究是当前计算机视觉领域的研究热点。面向深度学习中的Softmax图像分类方法，首先回顾了图像分类技术的发展历程，接着介绍了图像识别技术中的分类器，并解释了Softmax回归函数的分类实现原理。基于Softmax回归分类器的应用，详细阐述了多种图像分类技术，具体包括浅层神经网络、深度置信网络、深度自编码器和卷积神经网络。同时，对比介绍了各种级联模型的具体结构、训练方法、实际应用、分类效果以及优缺点。最后，从Softmax回归分类器、深度学习网络模型和高维数据分类三个方面对基于Softmax回归分类器的深度学习模型在图像分类方面的发展与应用前景进行了展望。     

模糊极小极大神经网络在火箭发动机故障分类中的应用

研究了一种用模糊集表示火箭发动机故障模式的神经网络分类器。模糊集是由模糊超立方体聚集形成的集合体，模糊超立方体是一个极小点和极大点构成的用隶属函数表示的ｎ维方盒。极小点和极大点的确定用包含有扩张与收缩阶段的模糊极小极大学习算法实现，这种算法能在一次循环学习中形成非线性模式边界，无需对已知故障模式重新训练就可融合新样本和精炼已存在的故障模式。模糊集用于故障模式分类自然地提供了故障更刘水平分类的有用信     

基于自适应遗传BP算法神经网络数据融合的空中目标识别

空中目标识别是现代防空作战的重要研究内容。本文利用不同类型目标产生的多类型传感器的数据信息对目标进行识别。为了训练神经网络目标识别分类器，将遗传算法和BP算法相结合，提出了一种新的自适应遗传BP算法，利用这种神经网络来确定指标的权值。仿真试验结果表明，基于自适应遗传BP算法神经网络的识别是一种简单、可靠的目标识别方法，具有很好的目标识别效果。     

基于一维卷积神经网络和SoftMax分类器的风电机组行星齿轮箱故障检测

将卷积神经网络引入风机故障检测领域,设计了一种一维卷积神经网络的结构,并和SoftMax分类器相结合构造了一种双层智能诊断架构。一维卷积神经网络用于行星齿轮箱数据的特征提取,SoftMax分类器对提取的特征进行分类。与传统智能算法相比,该方法具有训练样本少,可直接使用原始数据训练网络;计算效率高,可以适应实时诊断的需要。试验结果证明,该方法可以有效地诊断出不同工况下的行星齿轮箱中的齿轮故障。     

Mutual and conditional mutual informations for optimizingdistributed Bayes detectors

This paper considers optimization of distributed detectors under the Bayes criterion. A distributed detector consists of multiple local detectors and a fusion center that combines the local decision results to obtain a final decision. Introduced first are distributional distance measures, the mutual information (MI) and the conditional mutual information (CMI), that are obtained by applying information theoretic concepts to detection problems. Error bound analyses show that these distance measures approximate the Bayesian probability of error better than the conventional ones regardless of the operational environments. Then, a new optimization technique is proposed for distributed Bayes detectors. The method uses the distributional distances instead of the original Bayes criterion to avoid the complexity barrier of the optimization problem. Numerical examples show that the proposed distance measures yield solutions far better than the existing ones     

Multiaspect classification of airborne targets via physics-basedHMMs and matching pursuits

Wideband electromagnetic fields scattered from N distinct target-sensor orientations are employed for classification of airborne targets. Each of the scattered waveforms is parsed via physics-based matching pursuits, yielding N feature vectors. The feature vectors are submitted to a hidden Markov model (HMM), each state of which is characterized by a set of target-sensor orientations over which the associated feature vectors are relatively stationary. The N feature vectors extracted from the multiaspect scattering data implicitly sample N states of the target (some states may be sampled more than once), with the state sequence modeled statistically as a Markov process, resulting in an HMM due to the “hidden” or unknown target orientation. In the work presented here, the state-dependent probability of observing a given feature vector is modeled via physics-motivated linear distributions, in lieu of the traditional Gaussian mixtures applied in classical HMMs. Further, we develop a scheme that yields autonomous definitions for the aspect-dependent HMM states. The paradigm is applied to synthetic scattering data for two simple targets     

基于动态变结构BP神经网络的目标威胁评估

针对传统目标威胁估计方法和BP神经网络的不足,在BP神经网络的基础上,建立了基于动态变结构BP神经网络的目标威胁估计模型.该模型通过在权值向量更新公式中引入冲量函数,加快了网络的搜索速度和精度,保证了网络获得全局最优值;通过实时调整隐含层节点数目,可以将网络结构优化,极大地提升了网络的灵活性.仿真结果表明,与传统目标威胁估计方法和BP神经网络相比,动态变结构BP神经网络具有更好的预测能力和收敛速度,可以快速、准确地完成目标威胁估计.     

成败型产品基于验后概率的Bayes序贯检验技术

分析了基于损失的Bayes检验方法和基于风险的SPOT(sequential posterior odd test)方法存在的不足.对装备试验与评价中成败型产品的Bayes序贯检验,在给定的两类风险要求下,建立了基于验后概率的Bayes序贯检验模型(PBSTM),给出了模型的求解算法和实际验后风险的计算公式. 最后结合装备可靠性分析案例,进行了示例分析,从数据上证明了PBSTM模型优越性和有效性.研究结果表明:PBSTM模型从理论上保证了序贯检验的结论同时满足小概率事件原理和两类风险要求,避免了现有假设检验方法存在的不足.      

基于LSTMAttention网络的短期风电功率预测

提出一种基于LSTMAttention网络的短期风电功率预测方法。首先,使用LSTM网络对数值天气预测(NWP)数据的特征信息进行提取,同时采用注意力机制有效分析了模型输入与输出的相关性,从而获取了更多重要时间的整体特征；其次,使用卷积神经网络(CNN)提取NWP数据的局部特征,并引入压缩和奖惩网络(SE)模块学习特征权重,利用特征重新标定方式提高网络表示能力；最后,将局部特征和整体特征进行特征融合,通过分类器输出分类结果。利用NOAA提供的美国加利福尼亚州某风电场的数据进行案例分析,证明了所提方法的有效性。试验结果表明,与BP神经网络、自回归积分滑动平均模型(ARIMA)模型和LSTM模型相比,LSTMAttention模型具有更高的预测精度,证明了该方法的有效性。     

自适应遗传神经网络算法在推力估计器设计中的应用

为了在全包线内能够准确方便估计出航空发动机推力,提出了一种自适应遗传神经网络算法:将遗传算法和神经网络技术相结合充分发挥遗传算法和神经网络各自的全局收敛性和局部搜索快速性的优点,其中通过自适应概率遗传操作及局部寻优算子直接优化出神经网络拓扑结构及权值(包括阈值),克服了神经网络隐层节点需凭经验尝试的缺点和神经网络对初始权值(包括阈值)敏感的缺点,再应用神经网络对上述优化的权值(包括阈值)进行"精调",最后设计出全包线推力估计器.经验证,此推力估计器具有较高估计精度和良好泛化能力.      

基于Bayes分类器的DoS／DDoS攻击防御技术

提出了一种基于贝叶斯分类算法的分布式拒绝服务攻击防御技术。利用贝叶斯分类算法来计算数据包的分值特征,按照分值表对数据包进行评分并映射成危险等级。然后对危险等级进行评估,根据不同危险等级对网络流量进行过滤。基于该防御技术,设计并实现了防御系统。通过实验分析,该系统在DoS／DDoS攻击发生时能有效区分正常流量与异常流量,从而实现对DoS／DDoS攻击进行实时防御。     

神经网络集成在发动机故障诊断中的应用研究

提出一种基于A daBoost的集成神经网络故障诊断方法,利用多层前向神经网络作为故障弱分类器,通过简单地训练若干个单一神经网络并将其预测结果进行合成,实现了对航空发动机多类故障的诊断。针对一个涡轮喷气发动机气路部件的仿真实验表明,这种方法提高了最终故障分类器的泛化能力,便于工程应用。      

一种提高SAR目标识别率的有效方法

在合成孔径雷达自动目标识别SAR ATR中，SAR像的预处理是提高识别率的关键技术之一。给出了一种简单有效的SAR图像预处理方法，该方法首先对SAR目标像进行对数变换后，再做傅立叶变换。经预处理后的SAR像用支持矢量机SVM分类器进行目标识别。实验结果表明：本方法不但有效地提高了目标识别率，而且保证了目标的平移不变性并具有良好的推广能力。     

基于DSP的三自由度肌电假手实时控制方法

 利用安置在拇长屈肌，指深屈肌和指伸肌上的3个电极所测得的肌电信号，采用所提出的新的模式分类器，可以实现基于DSP的三自由度假手手指运动的实时控制。该分类器采用自回归(AR)参数模型和样本熵的方法构造特征矢量，经过由弹性反向传播(RP)算法构建的3层前馈神经网络的分类，能够成功地分辨出拇指、食指和中指的弯曲与伸展运动，平均识别率可以达到91%以上。实验结果表明，该分类器具有很高的辨识能力，同时由于其较小的计算量，也为嵌入式的多自由度肌电假手控制提供了一种新的控制方法。