USB多路数据采集器

针对目前大多数采集系统无法实现多通道高分辨率采样及携带不方便等情况,提出了一种基于USB的多路数据采集系统的设计方法。系统利用ARM+FPGA+AD7656的系统组合实现16路通道同步信号采样,FPGA用于实现对A/D转换的逻辑控制,通过ARM7处理器对A/D转换数据进行处理,再由USB接口与计算机进行数据通信。测试结果表明,基于FPGA与ARM的多通道数据采集系统结构简单、控制方便、设计成本较低,能够准确快速地对16路信号同时进行采集。     

自适应光学波前处理机D/A电路的优化

为了适应当前的自适应光学系统的要求，介绍一种多路D/A转换电路的优化及实现方法。它将输出的多路数字信号依次地通过一路D/A，进行数模转换，再通过采样保持器，将各路信号进行采样保持，分成了多路信号来控制变形镜，从而不但简化了系统，而且综合小信号误差，减少故障率。     

数字波形匹配的导航信号载波精密同步方法

导航信号质量评估中,为了避免中频采样下变频处理带来的非线性失真,采用微波直接采样信号.但是由于数据量庞大,传统的载波同步方法实现困难.针对这个问题,提出一种采用匹配滤波技术的载波精密同步方法.首先,用大步进量产生一系列初相不同的伪随机码调制本地波形,伪随机码与导航信号相同,对采样的导航信号进行匹配滤波,根据匹配滤波输出结果确定导航信号载波相位的粗估计值,进行载波粗剥离.然后进行导航信号的伪随机码同步,以载波相位粗估计值为中心,在一定范围内以更精细的步进量进行载波波形匹配,实现载波精密同步.最后对算法进行仿真,仿真结果表明提出算法实际可行,载波同步精度高,载波频率为L波段时,选择合适步长,同步误差可小于0.01弧度.     

机械振动故障诊断系统的设计与应用

针对机械设备振动监测的高速、实时和同步处理的要求,以Visual C++为平台开发了基于频谱分析和轴心轨迹分析的机械振动故障诊断系统。采用USB 2850采集卡搭建振动信号采集系统,通过分组采集方式实现多通道信号的同步采集,利用键相信号对振动信号整周期重采样,将原始连续信号整周期化。根据信号采样和数据处理流程,结合各模块算法执行效率,设计多线程流水处理架构,保证系统的多流程同步执行。最后,通过转子不平衡故障和不对中故障模拟实验,验证了故障诊断方法正确无误,信号分析系统稳定、高效、准确。     

一种窄带水声信号数字处理系统

针对窄带水声接收机参数估计精度低、目标识别能力差的弱点,提出了数字鉴宽的概念及方法,利用复杂可编程逻辑器件(Complex Programmable Logic Device, CPLD)实现对窄带水声信号宽度的精确识别.该方法可广泛适用于数字滤波和机械滤波窄带水声接收电路,鉴宽结果性能稳定、可靠,可满足大多数窄带水声系统对信号及其宽度识别的精度要求.同时,根据实际水声系统对水声接收机增益调节的需要,提出了利用MSP430单片机及 D/A器件实现水声接收机手动及自动增益控制的方法,它可应用于各种水声接收电路中.所提出的窄带水声信号处理方法有算法简便、实用的优点,使用的理想效果在实际应用中得到了验证.     

正弦波形同步采样条件的识别与判定

针对正弦信号同步采样的定量判定,以及不同步采样时同步误差的定量评价,提出了一种基于四参数拟合的评价判定方法,可避免同步判定时影响分辨力的 ±1的计点误差.在此基础上,定义并实现了同步误差的定量评价.在实际正弦曲线、方波曲线和三角波曲线的采集测量实验中,验证了所述方法的正确性及可行性.使用单频数字滤波处理后,将其推广应用...     

变周期信号的时域同步平均新算法

齿轮箱的转速受载荷和电压等的影响将会出现波动.无外时标采样时,用时域同步平均的传统算法处理转速波动的齿轮箱振动信号的降噪问题非常困难,提出变周期信号的时域同步平均新算法来解决这个问题.即利用转速波动的频域估计法确定同步平均的信号周期,再利用谱分析法计算数据长度,将变周期信号转换成周期信号进行时域同步平均.变周期信号时域同步平均新算法,不仅可省略外时标触发装置,而且可以改善由于转速波动引起的相位累积误差效应.模拟仿真和齿轮箱实验结果表明,该方法提高了信噪比,改善了相位累积误差.     

基于梅花阵列的复合材料全方位冲击定位方法

针对一维线性传感器阵列的盲区问题,研究了基于梅花阵列的二维多重信号分类(2D-MUSIC)冲击定位方法。梅花阵列的布置方式是在圆心处布置1个PZT传感器,另外在半径为 R 的圆周上均匀布置 M 个PZT传感器,首先联合冲击信号传播到参考阵元PZT0和第1个接收阵元PZT1的时间信息估计冲击信号的Lamb波传播波速;然后利用近场2D-MUSIC算法实现冲击源的距离和角度的同时定位;最后,在玻璃纤维复合材料层合板中进行了冲击定位方法的验证。实验结果表明: 基于梅花阵列的方法能够对0°~360°的冲击源进行精确地定位,定位结果精度较高,其中最大距离误差为2.2 cm,最大角度误差为5°。     

基于Hilbert变换科氏流量计相位差测量方法改进

离散Hilbert变换因其计算相位差准确度高,在科氏流量计中得到广泛应用。针对非整周期采样下信号变换失真问题,研究基于采样数据的减小信号变换失真算法,提高流量测量精度。分析离散Hilbert变换四步算法,根据离散变换特征,提出非整周期信号周期化插补算法和信号变换误差计算的区间循移算法。通过实验验证,在非整周期采样时,所研究算法可减小采样信号变换失真,相位差计算精度提高一位,质量流量计算误差小于原来误差1/5。     

小型化多道地震信号采集系统设计与实现

针对工程地震信号采集设备的便携性问题,采用MSP430F149驱动单片24位ADS1256以分时切换方式控制四路差分信号采集,实现系统的低功耗、小型化设计。重点介绍有限存储空间内大数据量的暂存、A/D通道切换与数据存储时间差的平衡问题,包括A/D分辨率和精度的提高方法。系统最终实现最大1kS/s等效采样率时四通道2048点的并行采集存储,分辨率在19位以上,从其在自主开发的无线遥测式地震仪应用效果来看,其信号频谱均满足有效地震波勘探需求。     

OFDM系统的一种精确的符号定时估计——纪念汪德昭院士诞辰100周年

作为一种高效的数据传输技术,正交频分复用(OFDM)技术具有良好的抗衰落能力,实现数据的高速传输。但是OFDM对同步误差特别是符号同步误差十分敏感。本文提出了一种低复杂度的符号定时估计,利用一个训练帧来实现同步。文中分析了该算法的性能,并在噪声信道中进行了性能分析,给出的仿真结果表明,该算法具有较高的精度和较低的计算复杂度。     

Optimal Synchronization of Local Clocks by GPS 1PPS Signals Using Predictive FIR Filters

In this paper, we discuss optimal synchronization of local clocks using Global Positioning System (GPS) one-pulse-per-second (1PPS) timing signals. To eliminate sawtooth errors that are peculiar to the 1PPS signals and optimally steer the clock errors each $M$ seconds, we propose the use of a ramp-predictive finite-impulse-response (FIR) filter that is known to be optimal for clock models on large averaging horizons. A low-pass filter is used to smooth the hold filter output between the optimally predicted points. A GPS-locked crystal clock has been investigated in detail in terms of the time interval error, Allan deviation, and precision time protocol (PTP) variance. A high-efficiency implementation of the proposed synchronization algorithm is experimentally demonstrated.      

通道间延迟时间差的测量不确定度

介绍了用正弦波拟合法评价测量系统通道间延迟时间差不确定度的分析和评定过程。讨论了几个主要的不确定度来源,包括信号源误差、采集序列的谐波失真、噪声及非谐波失真、抖动、软件拟合运算误差等。结合一个实例,给出了通道间延迟时间差测量不确定度的评定结果。该过程可用于相应计量标准的不确定度分析与评定。     

高精度数据采集在压力测控系统中的应用

基于Window98平台，利用PC-7423A/D转换卡对压力测控系统中的压力信号进行采集，并通过几种定时方法的比较，选用多媒体定时器实现了对压力数据的定时获取。同时，采用数字滤波、自适应校正算法进一步提高了采集精度。最后，给出了用VC 开发的程序实例。     

Iterative algorithm for phase extraction from interferograms with random and spatially nonuniform phase shifts

An advanced iterative algorithm is presented to extract phase distribution from randomly and spatially nonuniform phase-shifted interferograms. The proposed algorithm divides the interferograms into small blocks and retrieves local phase shifts accurately by iterations. Therefore, the phase distribution can be calculated with high precision by eliminating the effect of tilts occurring during phase shifting. Simulated results and experiments demonstrate that the proposed algorithm exhibits high precision and converges faster than previous algorithms even when the tilt errors are up to 27.6% of the normal phase step.     

An FFT-based method to evaluate and compensate gain and offset errors of interleaved ADC systems

Interleaved analog-digital converter (ADC) systems can be used to increase the sampling rate for a given ADC implementation technique. In theory, the maximum sampling rate that can be achieved is limited only by the bandwidth and the practical limits related to the power and space of integrated circuits. In this paper, a solution to increase the sampling rate of a digitizing system based on interleaved ADCs is presented. An error analysis, which takes into consideration offset and gain errors of the different ADC channels, is performed in order to quantify the effect of such errors in the system's performance. A software method based on the fast Fourier transform is presented for offset and gain error compensation of interleaved ADC associations. Numerical simulations and experimental results are used to validate the theory and the proposed compensation algorithm.     

Phase correction of interferograms using digital all-pass filters

A method for the phase correction of interferograms in Fourier transform infrared spectroscopy is presented. It is shown that phase error can be canceled to within an arbitrary angular precision by a low-order digital all-pass filter. Such a filter only modifies the phase of the Fourier transform of the interferogram and keeps the magnitude unchanged, like the Mertz method, for example. However, our method minimizes the asymmetric apodization that results in photometric errors when using the Mertz method alone. A practical example is provided in which phase correction over a frequency range of 800 cm(-1) to 4000 cm(-1) using a 9-pole all-pass filter resulted in a photometric error of <0.01%, much less than the 0.3% error of the Mertz method. An alternative and faster (approximately 100 ms) approach is to use an all-pass filter with lower angular precision followed by the Mertz method. Removing most of the phase error with the filter brings the interferogram to an optimal state so that the residual phase error can be completely removed with the Mertz procedure without introducing photometric error. The method can be used in most experiments, including emission spectroscopy, where conventional techniques are inadequate. A simple all-pass filter design algorithm is given.     

Quasi-synchronous sampling algorithm and its applications

Synchronous sampling is a commonly used method with good performance. However, when perfect synchronization is impossible, or when the signal contains interharmonics, undesirable asynchronous deviations between the sampling and some components of the signal may exist. Consequently, possibly large leakage or truncation errors may occur. Consisting of a quasi-synchronous window to reduce the long-range leakage, and a compensation algorithm to reduce the short-range leakage after normal FFT, the method described in this paper will give much more accurate measurement of many electrical quantities, e.g., spectral content of signal and phase difference between two periodic signals in power networks, than the traditional one