A signal coordination algorithm for two adjacent intersections based on approximate dynamic programming

To reduce vehicle emissions in road networks, a new signal coordination algorithm based on approximate dynamic programming （ADP） is developed for two intersections. Taking the Jetta car as an experimental vehicle, field tests are conducted in Changchun Street of Changchun city and vehicle emission factors in complete stop and uniform speed states are collected. Queue lengths and signal light colors of approach lanes are selected as state variables, and green switch plans are selected as decision variables of the system. Then the calculation model of the optimization index during the planning horizon is developed based on the basis function method of the ADP. The temporal-difference algorithm is employed to update the weighting factor vector of the approximate function. Simulations are conducted in Matlab and the results show that the established algorithm outperforms the conventional coordination algorithm in reducing vehicle emissions by 8.2%. Sensitive analysis of the planning horizon length on the evaluation index is also conducted and the statistical results show that the optimal length of the planning horizon is directly proportional to the traffic load.

Characterization of surface EMG signals using improved approximate entropy

An improved approximate entropy (ApEn) is presented and applied to characterize surface electromyography (sEMG) signals. In most previous experiments using nonlinear dynamic analysis, this certain processing was often confronted with the problem of insufficient data points and noisy circumstances, which led to unsatisfactory results. Compared with fractal dimension as well as the standard ApEn, the improved ApEn can extract information underlying sEMG signals more efficiently and accu- rately. The method introduced here can also be applied to other medium-sized and noisy physiological signals.

Truncated sparse approximation property and truncated q-norm minimization

This paper considers approximately sparse signal and low-rank matrix’s recovery via truncated norm minimization minx∥xT∥q and minX∥XT∥Sq from noisy measurements.We first introduce truncated sparse approximation property,a more general robust null space property,and establish the stable recovery of signals and matrices under the truncated sparse approximation property.We also explore the relationship between the restricted isometry property and truncated sparse approximation property.And we also prove that if a measurement matrix A or linear map A satisfies truncated sparse approximation property of order k,then the first inequality in restricted isometry property of order k and of order 2k can hold for certain different constantsδk andδ2k,respectively.Last,we show that ifδs(k+|T^c|)<√(s-1)/s for some s≥4/3,then measurement matrix A and linear map A satisfy truncated sparse approximation property of order k.It should be pointed out that when Tc=Ф,our conclusion implies that sparse approximation property of order k is weaker than restricted isometry property of order sk.

基于SO-PAA-GAF和AdaBoost集成学习的高压断路器故障诊断

针对在小样本和复杂工况下高压断路器故障诊断识别精度不高的问题,提出一种基于振动信号处理和Ada Boost集成学习的高压断路器故障诊断方法。首先,搭建高压断路器实验平台并采集8种工况下的分闸振动信号。其次,对振动信号进行绝对值处理后,使用分段聚合近似(piecewise aggregate approximation,PAA)进行分段平均,将输出的新序列采用格拉姆角场(Gramian angular field,GAF)转换成图片,并使用Relief F方法对提取的高维图片特征进行重要度排序。最后,将保留的重要特征输入到Ada Boost集成学习模型进行故障诊断,并用蛇优化算法确定最优PAA分段步长和输入分类器特征数量,以进一步提高故障诊断精度。通过分析多种信号处理方式及分类模型可知,图片信号和Ada Boost集成学习模型能够有效处理振动信号并准确判断故障类型,为准确、可靠地诊断高压断路器故障提供了新途径。

面向数字信号处理领域的近似计算技术应用与研究进展

在信号处理领域,近似计算技术备受关注。复杂算法和海量数据限制了应用的处理速度且增加了系统硬件消耗。由于信号具有冗余性,精确结果并非必需,满足用户可接受的结果已足够。因此,采用近似计算技术可以有效减少计算量,提高计算效率和系统性能。该文以近似计算技术的不同设计层次为切入,首先介绍了信号处理应用的特点,综述了近年来近似计算技术在算法和电路层面的研究进展,并调研了通信、视频图像以及雷达等信号处理方向的近似计算技术方案。最后,对该领域的发展方向进行了讨论和展望,为推动近似计算技术在信号处理领域的应用提供了思路。

基于GA-MP的低复杂度OTFS检测算法

针对正交时频空(Orthogonal Time Frequency Space,OTFS)通信系统信号检测复杂度高的问题,提出一种改进的高斯近似消息传递(Gaussian Approximate Message Passing,GA-MP)检测算法。依据最大后验概率检测准则,对发送信号及隐变量进行逐符号高斯近似,基于置信传播算法与联合因子图进行消息传递,用边缘后验概率替代GA-MP中的外部信息以减少运算量,结合阻尼因子提升收敛速度,同时引入概率阈值减少后续更新的节点数,从而使运算复杂度得到有效降低。实验结果表明,改进后的GA-MP算法在保证误码率性能的前提下具有更低的复杂度。

A wavelet-approximate entropy method for epileptic activity detection from EEG and its sub-bands

Epilepsy is a common brain disorder that about 1% of world's population suffers from this disorder. EEG signal is summation of brain electrical activities and has a lot of information about brain states and also used in several epilepsy detection methods. In this study, a wavelet-approximate entropy method is ap-plied for epilepsy detection from EEG signal. First wavelet analysis is applied for decomposing the EEG signal to delta, theta, alpha, beta and gamma sub- ands. Then approximate entropy that is a chaotic measure and can be used in estimation complexity of time series applied to EEG and its sub-bands. We used this method for separating 5 group EEG signals (healthy with opened eye, healthy with closed eye, interictal in none focal zone, interictal in focal zone and seizure onset signals). For evaluating separation ability of this method we used t-student statistical analysis. For all pair of groups we have 99.99% separation probability in at least 2 bands of these 6 bands (EEG and its 5 sub-bands). In comparing some groups we have over 99.98% for EEG and all its sub-bands.

A search-free near-field source localization method with exact signal model

Most of the near-field source localization methods are developed with the approximated signal model,because the phases of the received near-field signal are highly non-linear.Nevertheless,the approximated signal model based methods suffer from model mismatch and performance degradation while the exact signal model based estimation methods usually involve parameter searching or multiple decomposition procedures.In this paper,a search-free near-field source localization method is proposed with the exact signal model.Firstly,the approximative estimates of the direction of arrival(DOA)and range are obtained by using the approximated signal model based method through parameter separation and polynomial rooting operations.Then,the approximative estimates are corrected with the exact signal model according to the exact expressions of phase difference in near-field observations.The proposed method avoids spectral searching and parameter pairing and has enhanced estimation performance.Numerical simulations are provided to demonstrate the effectiveness of the proposed method.

Effect on Intensity Correlation Time by Input Signal in a Single-Mode Laser with Bias Signal Modulation

The effect on intensity correlation time T by input signal is studied for gain-noise model of a single-mode laser driven by colored pump noise and colored quantum noise with colored cross-correlation with a bias signal modulation in this paper. By using the linear approximation method, we detect that there exists maximum （i.e., resonance） in the curve of the intensity correlation time T upon bias current io when the noise correlation coefficient λ is positive; and there exists minimum （i.e., suppression） in the T-io curve when λ is negative. And whenλ is zero, T increases monotonously with increasing io. Furthermore, the curve of T upon the signal frequency Ω is also studied. Our study shows that no matter what the value of λ is, there exists minimum （i.e., suppression） in the T-Ω curve.

APPROXIMATE SAMPLING THEOREM FOR BIVARIATE CONTINUOUS FUNCTION

An approximate solution of the refinement equation was given by its mask, and the approximate sampling theorem for bivariate continuous function was proved by applying the approximate solution . The approximate sampling function defined uniquely by the mask of the refinement equation is the approximate solution of the equation , a piece-wise linear function , and posseses an explicit computation formula . Therefore the mask of the refinement equation is selected according to one' s requirement, so that one may controll the decay speed of the approximate sampling function .

Stochastic resonance in a gain-noise model of a single-mode laser driven by pump noise and quantum noise with cross-correlation between real and imaginary parts under direct signal modulation

Stochastic resonance （SR） is studied in a gain-noise model of a single-mode laser driven by a coloured pump noise and a quantum noise with cross-correlation between real and imaginary parts under a direct signal modulation. By using a linear approximation method, we find that the SR appears during the variation of signal-to-noise ratio （SNR） separately with the pump noise self-correlation time τ, the noise correlation coefficient between the real part and the imaginary part of the quantum noise λq, the attenuation coefficient γ＇ and the deterministic steady-state intensity I0. In addition, it is found that the SR can be characterized not only by the dependence of SNR on the noise variables of and λq, but also by the dependence of SNR on the laser system variables of γ and I0. Thus our investigation extends the characteristic quantity of SR proposed before.

Mathematical theory of signal analysis vs. complex analysis method of harmonic analysis

We present recent work of harmonic and signal analysis based on the complex Hardy space approach.

Intensity Correlation Function of a Single-Mode Laser Driven by Colored Cross-Correlation Noises in Modulation of a Bias Signal

By using the linear approximation method, the intensity correlation function is calculated for a single-mode laser modulated by a bias signal and driven by colored pump and quantum noises with colored cross-correlation. We found that, when the correlation time between the two noises is very short, the behavior of the intensity correlation function versus the time, in addition to decreasing monotonously, also exhibits several cases, such as one maximum, one minimum, and two extrema. When the correlation time between the two noises is very long, the behavior of the intensity correlation function exhibits oscillation and the envelope is similar to the case of short cross-correlation time.

Measuring System Regularity Using Fuzzy Similarity-based Approximate Entropy

Approximate entropy (ApEn), a measure quantifying regularity and complexity, is believed to be an effective analyzing method of diverse settings that include both deterministic chaotic and stochastic processes, particularly operative in the analysis of physiological signals that involve relatively small amount of data. However, the similarity definition of vectors based on Heaviside function, of which the boundary is discontinuous and hard, may cause some problems in the validity and accuracy of ApEn. To overcome these problems, a modified ApEn based on fuzzy similarity (mApEn) was proposed. The performance on the MIX stochastic model, as well as those on the Logistic map and the Hennon map with noise, shows that the fuzzy similarity-based ApEn gets more satisfying results than the standard ApEn when characterizing systems with different regularities.

压缩感知重构算法的两步深度展开策略研究

针对压缩感知中重构算法的深度展开问题,提出了一种两步深度展开策略(two-step deep unfolding,TwDU)。已有深度展开重构算法通常依赖前一步估计值估计当前值,TwDU对已有深度展开重构算法增加估计深度,依赖于前两步估计值估计当前展开值。TwDU对已有深度展开算法前两步估计值增加了两个训练权重。训练权重优化利用了信号估计值之间的相关特性,可以随着数据的特性自我学习和调整,所提TwDU策略应用于可学习迭代软阈值算法(learned iterative soft thresholding algorithm,LISTA)、可训练迭代软阈值算法(trainable iterative soft thresholding algorithm,TISTA)、可学习近似消息传递算法(learned approximate message passing,LAMP)等已有深度展开算法。通过在一维和二维稀疏信号的仿真验证,TwDU策略在重构精度和收敛速度上都更具有明显优势。

基于全特征值轨迹多项式逼近的双馈风机并网宽频振荡分析

高比例新能源接入下的电力系统振荡呈现宽频特性,其阻尼和振荡频率受系统参数影响显著。为避免运行参数变化和控制器参数设置不当造成系统失稳,亟须量化系统振荡模态与参数之间的关系并分析其背后隐藏的机理。文中提出了基于多项式逼近的全特征值轨迹计算方法,通过一系列多项式基函数的线性组合来准确描述系统所有特征值与多个参数之间复杂的隐式函数关系,得到特征值变化轨迹的显式表达式。针对双馈风机并网系统进行了详细电磁暂态建模,以单机无穷大系统和风火打捆外送系统为例,验证了所提方法的准确性,并利用特征值轨迹显式表达式分析了线路串补度、转子侧控制器比例和积分系数等重要参数对双馈风机并网系统次同步振荡、低频振荡等关键模态的影响。

OTFS调制系统的低复杂度GAMP算法实现

针对正交时频空(Orthogonal Time Frequealy Space,OTFS)调制技术信号检测算法复杂度高的问题,结合OTFS信道矩阵特性构建匹配滤波器,在广义近似消息传递(Generalized Approximate Message Passing,GAMP)算法的基础上,提出了基于期望最大化(Empectation Maxinization,EM)的阻尼广义消息传递算法。相较于传统的消息传递(Message Passing,MP)、近似消息传递(Approximate Message Passing,AMP)和GAMP等检测方式,所提算法有更低复杂度。在500 km/h的高动态场景中对该算法进行仿真,其误码率性能优于最小均方差(Minimum Mean Squared Error,MMSE)和传统的GAMP检测算法,并且具有更好的收敛性。

基于多模态特征组合的真实驾驶疲劳状态识别

对驾驶员在驾驶过程中的疲劳状态进行实时准确的预判,可减少由于驾驶疲劳引发的交通事故。首先,通过无线体域网采集了12名驾驶员在真实驾驶过程中的多模态特征,提取了脑电、肌电、呼吸等3种生理信号的特征参数近似熵,其中基于畸变能密度理论(DED)确定肌电信号的采集位置为颈6棘突旁开2 cm处的上斜方肌;然后,通过模糊C聚类方法分析了3种特征参数组合对疲劳状态的反映效果;最后,建立基于马氏距离理论的真实驾驶疲劳判别模型。研究结果表明,驾驶员颈6部位比颈7部位肌电信号的ApEn值显著下降(P<0.05),表明颈6处肌肉比颈7处肌肉对驾驶员的疲劳状态反映更为敏感,实际检测结果与畸变能密度理论计算结果一致,证明了该位置提取肌电信号的正确性和准确性;脑、肌、呼吸这3种生理信号的ApEn值均随驾驶时间的延长呈递减变化,驾驶约90 min时递减趋势变缓,表明驾驶员进入疲劳状态;通过模糊C聚类分析可知,当脑电与肌电ApEn组合时,清醒与疲劳的概率分布界限清晰,可有效反映驾驶疲劳状态;以脑电和肌电近似熵为自变量,基于马氏距离理论建立真实驾驶过程的疲劳判别模型,其测试集准确率达90.92%,表明该模型能够比较准确的判别出驾驶员的疲劳状态。

可穿戴脑电监测系统的低功耗逐次逼近型模数转换器设计

基于TSMC 180 nm芯片工艺,设计了一个用于可穿戴脑电监测的10位低功耗逐次逼近型模数转换器(SAR ADC)。为了有效降低DAC电容阵列的切换功耗,采用新型差分DAC电容阵列,并使用单调开关切换方式。比较器采用两级动态比较器,在提高效率的同时也降低功耗及噪声。针对脑电信号的特点,采用栅压自举开关实现高线性的采样。后仿真结果表明:在1.8 V供电电压、40 kS/s采样率下,这款SAR ADC的功耗仅为2.4μW,其有效位数(ENOB)为9.68 bit,无杂散动态范围(SFDR)为70.6 dB,优值为73.1 fJ/(conv-step),设计的SAR ADC适用于可穿戴脑电监测设备。

高精度SAR ADC电容阵列设计及校准算法

在高精度逐次逼近寄存器模数转换器(SAR ADC)中,电容阵列是SAR ADC的核心之一。电容阵列中的电容失配问题是导致转换精度降低的一个重要原因。为了尽可能改善这一问题,设计了一种6+6+6分段电容阵列,并且基于这种阵列设计了权重迭代算法的前台数字校准。该方法不需要额外的电容阵列,利用自身的电容阵列与比较器量化出电容失配,计算出每一位输出码的权重校准系数,用来对正常量化出的输出码进行编码,实现校准功能。仿真结果表明,引入电容失配的18 bit SAR ADC经过该算法校准后,信噪比(SNR)从77.6 dB提升到107.6 dB,无杂散动态范围(SFDR)从89.8 dB提升到125.6 dB,有效位数(ENOB)从12.54 bit提升到17.54 bit。在SMIC 0.18μm工艺下,该校准算法对高精度SAR ADC的动态性能具有较大提升。