Artificial neural-network-based diagnosis of CVD barrel reactor

This paper presents an artificial neural network (ANN) based diagnostic strategy applied to a chemical vapor deposition (CVD) barrel reactor of the type commonly used in silicon epitaxy. The strategy is based on the spatial variation of the rate of deposition of silicon on a facet of the reactor. Our hypothesis is that this spatial variation, quantified as a vector of variously measured standard deviations, encodes a pattern reflecting the state of the reactor. Therefore, a process fault (event) can be diagnosed by decoding the pattern by an ANN. We implemented this simple scheme by simulating different events by means of a regression model relating the rate of deposition to the process settings. Three different events were simulated and various ANNs were trained to detect and classify these events. It is shown that a single ANN or a combination of ANNs does an excellent job. We also demonstrate that the threshold rule for setting the threshold of a binary output neuron performing a classification task enhances the diagnostic performance. A novel multiple expert scheme that refers to several ANNs trained in the same classification task for decision-making in order to resolve ambiguities and improve the reliability of the final decision is presented and shown to be effective     

Weight-elimination neural networks applied to coronary surgery mortality prediction

The objective was to assess the effectiveness of the weight-elimination cost function in improving classification performance of artificial neural networks (ANNs) and to observe how changing the a priori distribution of the training set affects network performance. Backpropagation feedforward ANNs with and without weight-elimination estimated mortality for coronary artery surgery patients. The ANNs were trained and tested on cases with 32 input variables describing the patient's medical history; the output variable was in-hospital mortality (mortality rates: training 3.7%, test 3.8%). Artificial training sets with mortality rates of 20%, 50%, and 80% were created to observe the impact of training with a higher-than-normal prevalence. When the results were averaged, weight-elimination networks achieved higher sensitivity rates than those without weight-elimination. Networks trained on higher-than-normal prevalence achieved higher sensitivity rates at the cost of lower specificity and correct classification. The weight-elimination cost function can improve the classification performance when the network is trained with a higher-than-normal prevalence. A network trained with a moderately high artificial mortality rate (artificial mortality rate of 20%) can improve the sensitivity of the model without significantly affecting other aspects of the model's performance. The ANN mortality model achieved comparable performance as additive and statistical models for coronary surgery mortality estimation in the literature.     

Predicting high-risk preterm birth using artificial neural networks.

A reengineered approach to the early prediction of preterm birth is presented as a complimentary technique to the current procedure of using costly and invasive clinical testing on high-risk maternal populations. Artificial neural networks (ANNs) are employed as a screening tool for preterm birth on a heterogeneous maternal population; risk estimations use obstetrical variables available to physicians before 23 weeks gestation. The objective was to assess if ANNs have a potential use in obstetrical outcome estimations in low-risk maternal populations. The back-propagation feedforward ANN was trained and tested on cases with eight input variables describing the patient's obstetrical history; the output variables were: 1) preterm birth; 2) high-risk preterm birth; and 3) a refined high-risk preterm birth outcome excluding all cases where resuscitation was delivered in the form of free flow oxygen. Artificial training sets were created to increase the distribution of the underrepresented class to 20%. Training on the refined high-risk preterm birth model increased the network's sensitivity to 54.8%, compared to just over 20% for the nonartificially distributed preterm birth model.     

Classification of seismic signals by integrating ensembles ofneural networks

We examine a classification problem in which seismic waveforms of natural earthquakes are to be distinguished from waveforms of man-made explosions. We present an integrated classification machine (ICM), which is a hierarchy of artificial neural networks (ANNs) that are trained to classify the seismic waveforms. In order to maximize the gain of combining the multiple ANNs, we suggest construction of a redundant classification environment (RCE) that consists of several “experts” whose expertise depends on the different input representations to which they are exposed. In the proposed scheme, the experts are ensembles of ANN, trained on different bootstrap replicas. We use various network architectures, different time-frequency decompositions of the seismic waveforms, and various smoothing levels in order to achieve an RCE. A confidence measure for the ensemble's classification is defined based on the agreement (variance) within the ensembles, and an algorithm for a nonlinear integration of the ensembles using this measure is presented. An implementation on a data set of 380 seismic events is described, where the proposed ICM had classified correctly 92% of the testing signals. The comparison we made with classical methods indicates that combining a collection of ensembles of ANNs can be used to handle complex high dimensional classification problems     

Loss estimation and control mechanism in bufferless optical packet-switched networks based on multilayer perceptron

Artificial neural networks (ANNs) are well-known estimators for the output of broad range of complex systems and functions. In this paper, a common ANN architecture called multilayer perceptron (MLP) is used as a fast optical packet loss rate (OPLR) estimator for bufferless optical packet-switched (OPS) networks. Considering average loads at the ingress switches of an OPS network, the proposed estimator estimates total OPLR as well as ingress OPLRs (the OPLR of optical packets sent from individual ingress switches). Moreover, a traffic policing algorithm called OPLRC is proposed to control ingress OPLRs in bufferless slotted OPS networks with asymmetric loads. OPLRC is a centralized greedy algorithm which uses estimated ingress OPLRs of a trained MLP to tag some optical packets at the ingress switches as eligible for drop at the core switches in case of contention. This will control ingress OPLRs of un-tagged optical packets within the specified limits while giving some chance for tagged optical packets to reach their destinations. Eventually, the accuracy of the proposed estimator along with the performance of the proposed algorithm is evaluated by extensive simulations. In terms of the algorithm, the results show that OPLRC is capable of controlling ingress OPLRs of un-tagged optical packets with an acceptable accuracy.     

噪声环境下遗传算法的性能评价

 为了评价遗传算法在噪声环境下的优化性能,提出"平均最优解"和"最优解分布标准差"两个指标,实验结果表明新指标可以有效地评价噪声环境下遗传算法的优化性能.研究了实数编码遗传算法在噪声强度递增环境下的性能.结果表明小生境策略和多种群策略可以改善遗传算法在噪声环境下的性能,单点交叉在噪声环境下的性能要优于混合交叉.     

基于人工神经网络的时间调制阵列谐波波束形成

针对时间调制阵列(time-modulated array, TMA)提出了一种基于人工神经网络(artificial neural network, ANN)的谐波波束形成技术.该技术通过一个由编码器和解码器串联组成的ANN实现时序信息的优化,其中,编码器以目标角度的方向图约束值作为输入,通过在线训练输出对应的激励值;而解码器经过预训练可以实时输出辐射方向图.然后利用训练完成优化后的激励可以获得不同阵元的开关导通持续时间和起始时刻. 8元/16元不同指向TMA谐波波束形成算例仿真结果表明,所提方法可以有效抑制副瓣电平(sidelobe level, SLL),快速精确控制方向图,在目标角度实现高方向性波束形成,同时该方法具有耗时短、鲁棒性好和易调节的特点,有望应用于快速目标搜索和跟踪.     

A novel hybrid model for inversion problem of atmospheric refractivity estimation

Atmospheric refractivity estimation is an important issue for performance evaluation of communication systems and air surveillance radars. A novel hybrid model based on artificial neural networks (ANNs) and genetic algorithms (GAs) for inversion problem of atmospheric refractivity estimation is introduced. In this paper, inversion problem and clutter model problem of refractivity from clutter (RFC) method are separated and only inversion problem is studied. A problem specific ANN structure is designed and an original GA is developed to fulfill atmospheric refractivity estimations. In hybrid method, ANNs make pre-estimation and GAs use these results as a starting population for post-estimation. When the results obtained from the single solutions of ANNs and GAs are compared to the results obtained from hybrid model, a significant improvement in the accuracy of estimated results is observed.     

Analog CMOS deterministic Boltzmann circuits

CMOS circuits implementing an analog neural network (ANN) with on-chip deterministic Boltzmann learning (DBL) and capacitive synaptic weight storage have been designed, fabricated, and tested. Weights are refreshed by periodic repetition of the training data. The circuits were used to build a 12-neuron, 132-synapse ANN that performed well in a variety of learning experiments, including a 36-input to 4-output mapping problem. Adaptive systems such as those described here can compensate for imperfections in the components from which they are constructed and therefore can be built using simple silicon area-efficient analog circuits. The test results indicate that deterministic Boltzmann ANNs can be implemented efficiently using analog CMOS circuitry     

Analogue electronic circuit diagnosis based on ANNs

Feed-forward artificial neural networks (ANNs) have been applied to the diagnosis of nonlinear dynamic analogue electronic circuits. Using the simulation-before-test (SBT) approach, a fault dictionary was first created containing responses observed at all inputs and outputs of the circuit. The ANN was considered as an approximation algorithm to capture mapping enclosed within the fault dictionary and, in addition, as an algorithm for searching the fault dictionary in the diagnostic phase. In the example given DC and small signal frequency domain measurements were taken as these data are usually given in device’s data-sheets. A reduced set of data per fault (DC output values, the nominal gain and the 3 dB cut-off frequency, measured at one output terminal) was recorded. Soft (parametric) and catastrophic (shorts and opens) defects were introduced and diagnosed simultaneously and successfully. Large representative set of faults was considered, i.e., all possible catastrophic transistor faults and qualified representatives of soft transistor faults were diagnosed in an integrated circuit. The generalization property of the ANNs was exploited to handle noisy measurement signals.     

改进人工神经网络算法及其在E面分支波导耦合器优化设计中的应用

将混合遗传算法用于人工神经网络,训练出全局最优的权值和偏差,解决了反向传播网络收敛于局部极值的问题.运用该方法训练出E面分支波导耦合器的输入输出人工神经网络模型,并以此仿真并优化其他结构的耦合器.相对于精确电磁场数值计算,前者在保证有较高仿真精度的前提下,大大提高了仿真速度.     

基于神经网络的光纤温度和应变快速解调方法

为了提高基于布里渊散射的分布式光纤传感系统实时性,在分析经典基于洛伦兹和伪Voigt模型拟合法优缺点的基础上,将多层前馈神经网络方法用于布里渊频移的估算.确定了神经网络的结构、输入及输出量、激活函数和训练算法,采用不同信噪比(5dB～40dB)和布里渊频移(10.62GHz～10.82GHz)的布里渊谱训练该网络,训练...     

Computer-aided diagnosis of solid breast nodules: use of an artificial neural network based on multiple sonographic features

A computer-aided diagnosis (CAD) algorithm identifying breast nodule malignancy using multiple ultrasonography (US) features and artificial neural network (ANN) classifier was developed from a database of 584 histologically confirmed cases containing 300 benign and 284 malignant breast nodules. The features determining whether a breast nodule is benign or malignant were extracted from US images through digital image processing with a relatively simple segmentation algorithm applied to the manually preselected region of interest. An ANN then distinguished malignant nodules in US images based on five morphological features representing the shape, edge characteristics, and darkness of a nodule. The structure of ANN was selected using k-fold cross-validation method with k = 10. The ANN trained with randomly selected half of breast nodule images showed the normalized area under the receiver operating characteristic curve of 0.95. With the trained ANN, 53.3% of biopsies on benign nodules can be avoided with 99.3% sensitivity. Performance of the developed classifier was reexamined with new US mass images in the generalized patient population of total 266 (167 benign and 99 malignant) cases. The developed CAD algorithm has the potential to increase the specificity of US for characterization of breast lesions.     

Linear space-time multiuser detection for multipath CDMA channels

We consider the problem of detecting synchronous code division multiple access (CDMA) signals in multipath channels that result in multiple access interference (MAI). It is well known that such challenging conditions may create severe near-far situations in which the standard techniques of combined power control and temporal single-user RAKE receivers provide poor performance. To address the shortcomings of the RAKE receiver, multiple antenna receivers combining space-time processing with multiuser detection have been proposed in the literature. Specifically, a space-time detector based on minimizing the mean-squared output between the data stream and the linear combiner output has shown great potential in achieving good near-far performance with much less complexity than the optimum space-time multiuser detector. Moreover, this space-time minimum mean-squared error (ST-MMSE) multiuser detector has the additional advantage of being well suited for adaptive implementation. We propose novel trained and blind adaptive algorithms based on stochastic gradient techniques, which are shown to approximate the ST-MMSE solution without requiring knowledge of the channel. We show that these linear space-time detectors can potentially provide significant capacity enhancements (up to one order of magnitude) over the conventional temporal single-user RAKE receiver     

基于Saber的反激式开关电源的仿真研究

反激式开关电源因成本低、外围元器件少、可宽电压范围输入能耗小、支持多组输出而备受欢迎,但因输出电压纹波大而严重影响其工作性能。从反激式开关电源的工作原理出发,采用反激式开关电源输出端增加输出滤波电路的方法,解决反激式开关电源输出电压纹波大的问题。运用Saber仿真软件分别对普通反激式开关电源和增加 输出滤波电路的反激式开关电源进行建模和仿真。试验仿真对比表明,通过该方法可改善反激式开关电源的输出电压纹波,提高了反激式开关电源的工作性能。     

BER-Optimal Linear Parallel Interference Cancellation for Multicarrier DS-CDMA in Rayleigh Fading

In this paper, we consider the design and bit-error performance analysis of linear parallel interference cancellers (LPIC) for multicarrier (MC) direct-sequence code division multiple access (DS-CDMA) systems. We propose an LPIC scheme where we estimate and cancel the multiple access interference (MAI) based on the soft decision outputs on individual subcarriers, and the interference cancelled outputs on different subcarriers are combined to form the final decision statistic. We scale the MAI estimate on individual subcarriers by a weight before cancellation. In order to choose these weights optimally, we derive exact closed-form expressions for the bit-error rate (BER) at the output of different stages of the LPIC, which we minimize to obtain the optimum weights for the different stages. In addition, using an alternate approach involving the characteristic function of the decision variable, we derive BER expressions for the weighted LPIC scheme, matched filter (MF) detector, decorrelating detector, and minimum mean square error (MMSE) detector for the considered multicarrier DS-CDMA system. We show that the proposed BER-optimized weighted LPIC scheme performs better than the MF detector and the conventional LPIC scheme (where the weights are taken to be unity), and close to the decorrelating and MMSE detectors.     

Comparing the efficiency of artificial neural networks in sEMG-based simultaneous and continuous estimation of hand kinematics

Surface Electromyography (sEMG) plays a key role in many applications such as control of Human-Machine Interfaces (HMI) and neuromusculoskeletal modeling. It has strongly nonlinear relations to joint kinematics and reflects the subjects’ intention in moving their limbs. Such relations have been traditionally examined by either integrated biomechanics and multi-body dynamics or gesture-based classification approaches. However, these methods have drawbacks that limit their usability. Different from them, joint kinematics can be continuously reconstructed from sEMG via estimation approaches, for instance, the Artificial Neural Networks (ANNs). The Comparison of different ANNs used in different studies is difficult, and in many cases, impossible. The current study focuses on fairly evaluating four types of ANN over the same dataset and conditions in proportional and simultaneous estimation of 15 hand joint angles from 10 sEMG signals. The presented ANNs are Feedforward, Cascade-Forward, Radial Basis Function (RBFNN), and Generalized Regression (GRNN). Each ANN is applied to its special parametric study. All the methods efficiently solved the regression problem of the complex multi-input multi-output bio-system. The RBFNN has the best performance over the others with a 79.80% mean correlation coefficient over all joints, and its accuracy reaches as high as 92.67% in some joints. Interestingly, the highest accuracy over individual joints is 93.46%, which is achieved via the GRNN. The good accuracy suggests that the proposed approaches can be used as alternatives to the previously adopted ones and can be employed effectively to synchronously control multi-degrees of freedom HMI and for general multi-joint kinematics estimation purposes.     

基于神经网络的分振幅光偏振仪的数据处理

分振幅光偏振仪(DOAP)是一种高速测量光波偏振态的传感器。提出了一种基于人工神经网络(ANN)的分振幅光偏振仪的数据处理方法,将分振幅光偏振仪中电路系统输出的电信号作为神经网络的输入,入射光的斯托克斯参数作为神经网络的输出,建立一个前向多层神经网络模型。通过网络训练,使该网络确立了电路系统输出电信号与入射光斯托克斯参数之间的映射关系。由测量时得到的电信号,利用训练后的神经网络可以计算出待测的入射光的斯托克斯参数。测试结果表明,在测量精度方面,该方法获得的斯托克斯参数的总均方根偏差为1.9%,略优于基于矩阵运算的数据处理方法。     

ANN Synthesis Models Trained with Modified GA‐LM Algorithm for ACPWs with Conductor Backing and Substrate Overlaying

Accurate synthesis models based on artificial neural networks (ANNs) are proposed to directly obtain the physical dimensions of an asymmetric coplanar waveguide with conductor backing and substrate overlaying (ACPWCBSO). First, the ACPWCBSO is analyzed with the conformal mapping technique (CMT) to obtain the training data. Then, a modified genetic‐algorithm‐Levenberg‐Marquardt (GA‐LM) algorithm is adopted to train ANNs. In the algorithm, the maximal relative error (MRE) is used as the fitness function of the chromosomes to guarantee that the MRE is small, while the mean square error is used as the error function in LM training to ensure that the average relative error is small. The MRE of ANNs trained with the modified GA‐LM algorithm is less than 8.1%, which is smaller than those trained with the existing GA‐LM algorithm and the LM algorithm (greater than 15%). Lastly, the ANN synthesis models are validated by the CMT analysis, electromagnetic simulation, and measurements.     

基于GSA-SVR算法的MEMS温度漂移补偿方法

针对微机电系统(MEMS)仪表零偏受温度变化影响较大的问题,该文提出了一种基于引力搜索算法-支持向量回归(GSA-SVR)的MEMS零偏温度漂移补偿方法。先通过小波变换对MEMS陀螺和MEMS加速度计输出信号进行预处理,再采用GSA-SVR算法对MEMS在不同工作状态下进行温度建模并补偿。实验结果表明,在稳定工作阶段,与补偿前相比,补偿后加速度计和陀螺的输出标准差分别降低了90%和85%。与传统SVR相比,该文方法准确性较高,实用性较好,GSA-SVR算法将加速度计和陀螺输出的标准差分别降低了6%和10%。