A high speed 1000 fps CMOS image sensor with low noise global shutter pixels

A low read noise 8T global shutter pixel for high speed CMOS image sensor is proposed in this paper.The pixel has a pixel level sample-and-hold circuit and an in-pixel amplifier whose gain is larger than one.Using pixel level sample-and-hold circuit,the KTC noise on FD node can be effectively cancelled by correlated double sampling operation.The in-pixel amplifier with a gain larger than one is employed for reducing the pixel level sample-and-hold capacitors thermal noise and their geometric size.A high speed 1000 fps 256×256 CMOS image sensor based on the pixel is implemented in 0.18μm CMOS process.The chip active area is 5 mm×7 mm with a pixel size of 14μm×14μm.The developed sensor achieves a read noise level as low as 14.8e-while attaining a high fill factor of 40%.The full well capacity can contain 30840e-and the resulting signal dynamic range is 66 dB.     

Sensor fault diagnosis for a class of time delay uncertain nonlinear systems using neural network

In this paper, a sliding mode observer scheme of sensor fault diagnosis is proposed for a class of time delay nonlinear systems with input uncertainty based on neural network. The sensor fault and the system input uncertainty are assumed to be unknown but bounded. The radial basis function (RBF) neural network is used to approximate the sensor fault. Based on the output of the RBF neural network, the sliding mode observer is presented. Using the Lyapunov method, a criterion for stability is given in terms of matrix inequality. Finally, an example is given for illustrating the availability of the fault diagnosis based on the proposed sliding mode observer.     

Optimal State Estimation and Fault Diagnosis for a Class of Nonlinear Systems

This study proposes a scheme for state estimation and,consequently,fault diagnosis in nonlinear systems.Initially,an optimal nonlinear observer is designed for nonlinear systems subject to an actuator or plant fault.By utilizing Lyapunov's direct method,the observer is proved to be optimal with respect to a performance function,including the magnitude of the observer gain and the convergence time.The observer gain is obtained by using approximation of Hamilton-Jacobi-Bellman(HJB)equation.The approximation is determined via an online trained neural network(NN).Next a class of affine nonlinear systems is considered which is subject to unknown disturbances in addition to fault signals.In this case,for each fault the original system is transformed to a new form in which the proposed optimal observer can be applied for state estimation and fault detection and isolation(FDI).Simulation results of a singlelink flexible joint robot(SLFJR)electric drive system show the effectiveness of the proposed methodology.     

基于时延自相关降噪与局部特征尺度分解的齿轮故障诊断

In order to detect the gear fault accurately, a mode based on autocorrelation denoising and local characteristic-scale decomposition（LCD） was proposed. The autocorrelation function of the vibration signal of gear box was calculated,and the low delay time of [-20,20] and high delay time of last 20 points of the autocorrelation function were set to zeros. The delayed autocorrelation denoising signal was decomposed into some intrinsic mode components（ISC） by LCD, and the ISC included the mesh frequency were selected as the effective component. The fault was detected by the envelope aptitude spectrum. The analysis of broken tooth of gear fault data shows that the increase of the signal to noise ratio（SNR）can be 8.0963dB,the method can realize the occurrence of the fault effectively, can also detect the type of the gear fault, the method can effectively support fault diagnosis.     

Neural network technology in the plant fault diagnosis software application

In this paper, the conception and the development of the fault diagnosis technology are discussed, and the problems of fault diagnosis technology is solved in power plants by analyzing the actual and existing problems in the field of power plants fault diagnosis technology. Then we reveal the reliable technique to diagnose software by using BPNN in power plants fault diagnosis process. The experiment shows that complex model can be constructed by using this method and parameter estimation is done easily. This method is also fit for different datum sets, and it has less error. It is an efficient method in power plants fault diagnosis.     

Fault-screen diagnosis approach for linear circuits with tolerance

A fault location approach for linear circuits with tolerance is proposed based on fault screen theory and optimization technology. By searching two sets of optimum excitation in a tolerance circuit, the voltage differences between faulty branches and normal ones can be exposed under every set of optimum excitation. A fault attaching function is established base on the voltage differences. Fault location is performed by means of the fault attaching function. Thus, fault screen strategy is applied to linear circuits with tolerance and fuzziness between fault and tolerance effort is greatly reduced. The experimental results show that the proposed approach can provide accurate and effective diagnosis.     

Research and Design of a Fuzzy Neural Expert System

We have developed a fuzzy neural expert system that has the precision and learning ability of a neural network.Knowledge is acquired from domain experts as fuzzy rules and membership functions.Then,they are converted into a neural network which implements fuzzy inference without rule matching.The neural network is applied to problem-solving and learns from the data obtained during operation to enhance the accuracy.The learning ability of the neural network makes it easy to modify the membership functions defined by domain experts.Also,by modifying the weights of neural networks adaptively,the problem of belief propagation in conventional expert systems can be solved easily.Converting the neural network back into fuzzy rules and membership functions helps explain the inner representation and operation of the neural network.     

Improved empirical DC Ⅰ-Ⅴ model for 4H-SiC MESFETs

A novel empirical large signal direct current （DC）Ⅰ-Ⅴ model is presented considering the high saturation voltage, high pinch-off voltage, and wide operational range of drain voltage for 4H-SiC MESFETs. A comparison of the presented model with Statz, Materka, Curtice-Cubic, and recently reported 4H-SiC MESFET large signal Ⅰ-Ⅴ models is made through the Levenberg-Marquardt method for fitting in nonlinear regression. The results show that the new model has the advantages of high accuracy, easily making initial value and robustness over other models. The more accurate results are obtained by the improved channel modulation and saturation voltage coefficient when the device is operated in the sub-threshold and near pinch-off region. In addition the new model can be implemented to CAD tools directly, using for design of 4H-SiC MESFET based RF＆MW circuit, particularly MMIC （microwave monolithic integrate circuit）.     

On fault tolerance of 3-dimensional mesh networks

In this paper, the concept of k-submesh and k-submesh connectivity fault tolerance model is proposed. And the fault tolerance of 3-D mesh networks is studied under a more realistic model in which each network node has an independent failure probability. It is first observed that if the node failure probability is fixed, then the connectivity probability of 3-D mesh networks can be arbitrarily small when the network size is sufficiently large. Thus, it is practically important for multicomputer system manufacturer to determine the upper bound for node failure probability when the probability of network connectivity and the network size are given. A novel technique is developed to formally derive lower bounds on the connectivity probability for 3-D mesh networks. The study shows that 3-D mesh networks of practical size can tolerate a large number of faulty nodes thus are reliable enough for multicomputer systems. A number of advantages of 3-D mesh networks over other popular network topologies are given.     

Bagging Neural Networks for Predicting Water Consumption

Neural network ensembles can improve the generalization and stabilization ability of learning systems effectively through aggregating numbers of neural networks. It is considered as a promising kind of engineering neural computing technology. In this paper, we construct a learning model based on bagged neural network ensembles by bagging BP（Back Propagation） networks. Through analyzing the historical data of civic water consumption, the model can predict water consumptions efficiently for various kinds of users and several of periods of time. Our experiments and practical applications show that the model has high predicting precision, good stabilization as well as feasibility.     

脉频调制神经网络VLSI的设计及应用

本文提出了一种用于故障诊断识别的改进脉冲频率调制(PFM)VLSI神经网络电路,改进了传统的基于软件的机械故障诊断模式,发挥了神经网络超大规模集成电路(VLSI)的优势.利用单层感知器网络、场效应管电路实现了一种新的数字模拟混合突触乘法/加法器电路,而且该神经网络电路的突触权值不需要学习调整,降低了电路的复杂性.以此电路为基础,设计了进行主轴承噪声故障诊断的神经网络故障识别系统.将含有故障信息的原始噪声信号,经过前置信号处理分析、故障特征值提取和神经网络运算,得出VLSI电路输出端电容的电压——代表待识别信号与模板故障信号的“欧氏距离”,进而判断出故障的类别.经过仿真测试,基于硬件的诊断系统的识别性能接近于基于软件的系统.     

模拟神经网络VLSI脉冲流技术在故障诊断中的应用分析

介绍了模拟神经网络VLSI脉冲流技术实现神经网络模式识别硬件电路的方法，并且直接将故障分类。提出利用包含有故障信息的原始模拟噪声信号，经过前置信号处理和神经网络运算，得出VLSI电路输出端电容的电压值－代表待识别信号与模板故障信号的“欧氏距离”，以实现噪声故障信号的实时硬件在线识别。     

一种面向噪声环境中旋转机械故障诊断的多模态耦合输入神经网络

旋转机械设备是工业生产中的关键性设备,对其进行高效故障诊断,对于保障工业安全生产具有重要意义.传统的旋转机械设备智能故障诊断方法采取人工特征提取策略,存在依赖专家经验知识、特征泛化性差、特征完备性不足等局限性,导致故障诊断模型精度差,特别是在噪声环境下性能下降明显.对此,提出一种用于旋转机械故障诊断的多模态耦合输入神经网络模型.首先,利用信号分解方法将原始输入信号分解为多个子信号,并将子信号与原始信号成对组成二维矩阵并输入到神经网络中,使得网络能够提取其间重要的相关特征;然后,利用双通道并行的卷积神经网络和长短期记忆网络分别提取信号中的时空间特征并融合,大大提高网络模型的特征表达完备性,实现对旋转机械设备的高精度故障分类.通过实验验证了所提出模型相较于传统故障模型具有更高的准确率,并且对于噪声干扰也有较好的适应性.     

基于集成多神经网络在模拟电路故障诊断中的研究

以进行模拟电路故障诊断为主要目的,针对单神经网络故障字典法在进行复杂电路系统故障诊断时,对多故障和多任务诊断的不足之处,讨论了基于多故障的神经网络集成技术,采用集成多神经网络来提高诊断速度和精度,提出了集成多神经网络故障字典法来解决多故障任务,对基于层次分类模型的多重结构神经网络进行了研究,给出了两种对故障定位的统一融合算法,克服了采用单神经网络多故障时学习速度慢,出现新故障的网络要重新进行学习等缺点.并给出了应用实例.     

基于脉宽调制的传感器读取电路设计与实现

设计了一种新型的传感器信号读取电路,该电路将传统的脉宽调制PWM(PulseWidthModulation)电路进行改进,对PWM信号采用占空比和频率同时调制而不是单一的占空比调制,在信号传输过程中,该电路可将两路电压输入信号调制到一路PWM信号上,通过对输出PWM信号进行解调可还原两路输入信号的电压值.实验结果表明,该电路输出PWM信号占空比和频率分别与两路输入电压信号呈良好的线性关系,电压转换精度分别达到0.34%、0.26%.此外,该电路具有抗干扰能力强、转换精度高和成本低的优点,非常适合传感器信号的调理和读取.     

基于PCA和神经网络的故障诊断技术

提出一种基于PCA和神经网络的故障诊断/识别方法,利用主元分析法提取故障样本集的主元,实现故障样本的最优压缩,简化故障诊断中神经网络分类器的结构,提高神经网络的分类速度和测试精度。仿真结果表明,该方法可以有效减少输入层神经元个数,提高神经网络模型的学习效率和诊断的准确性,具有良好的故障识别能力。     

基于BP小波神经网络的模拟电路故障诊断

针对现有BP网络在模拟电路故障诊断中存在的问题,提出了一种基于BP小波神经网络的故障诊断方法.该法将小波函数与BP网络结合构成BP小波网络,这种网络具有小波变换的时频局域化性质和BP网络的自学习能力.分别用BP小波网络和传统BP网络对实例电路进行故障诊断,结果表明本方法是有效的,而且比传统BP网络方法的学习收敛速度快得多.     

基于神经网络的大规模模拟电路故障检测系统

设计了一个基于小波和神经网络的信号处理系统，该系统主要针对大规模模拟电路故障检测。针对传统诊断技术的局限性，讨论了利用神经网络方法分级诊断大规模模拟电路软故障的方案，通过小波变换提取故障特征，并利用神经网络的非线性映射特性逼近故障诊断模型。诊断结果表明基于人工神经网络的电路故障诊断方法是行之有效的。此方法具有广阔的应用前景，为大规模模拟电路故障诊断提供了新的理论依据和检测方法，并有希望研制成一套高效的检测设备。     

火电厂生产过程故障诊断神经网络专家系统

将人工神经网络和专家系统结合地火电厂生产过程的故障诊断，提出并设计了热工过程故障诊断系统。同时对常规ＢＰ算法做了改进，神经网络则采用分层化与并联运用相结合的结构，从而使系统诊断速度和准确率得到提高。仿真结果表明了该系统设计的合理性和可行性。     

基于模糊BP神经网络的模拟电路实时故障诊断

以进行模拟电路实时故障诊断为主要目的,对BP神经网络故障字典法进行了深入研究,针对传统BP算法收敛速度慢、易产生局部最优等不足,采用神经网络与模糊理论相结合的方法,根据模式识别原理实现模拟电路故障的实时诊断。实验结果表明:该算法在网络收敛速度和识别精度上较传统的BP算法均有明显的改善。