HMM-SVM混合模型在模拟电路故障诊断中的应用

针对单一的隐马尔科夫模型(HMM)或支持向量机(SVM)在模拟电路早期的软故障中识别率不高的特点,将HMM-SVM混合模型应用到模拟电路早期的软故障识别中。首先通过主成分分析(PCA)将原始数据样本降维实现初步划分;接着利用HMM计算测试样本与各故障状态的匹配程度形成特征向量;最后由SVM做故障状态判别。实验结果表明,HMM-SVM混合模型的早期故障识别率优于单一的HMM或SVM模型,将平均故障识别率提高到95%以上。     

异常样本监督下的最优超球的研究及应用

针对模拟电路新故障类型状态检测中,因支持向量描述(SVDD)构建时忽略异常样本监督和特征空间存在混叠,导致诊断新故障类型准确率不高的缺点,提出了异常样本监督下的最优超球SVDD。该方法利用训练后超球SVDD各参数,将相似度高的已知类型样本融合形成新类,降低新类间样本相似度。并从原理上改进SVDD,实现异常样本监督,从而在最大程度上降低虚警和漏警率以适应模拟电路新故障类型状态检测。经过模拟电路新故障类型检测应用实例可以看出该方法在模拟电路新故障类型检测中,较常用方法诊断准确率有所提升。     

基于Takagi-Sugeno型自适应模糊神经网络的模拟电路故障诊断

该文提出了一种基于Takagi-Sugeno型自适应模糊神经网络故障诊断方法。首先通过电路仿真获得故障样本,其次利用主成分分析对故障样本进行降维处理,减少自适应模糊神经网络的输入,降低训练时间,然后采用BP算法与最小二乘法相结合的混合学习算法训练自适应模糊神经网络的连接权值和隶属度函数。仿真结果表明,此方法能够快速有效地对模拟电路的故障进行诊断和定位,表现出了很好的应用潜力,在容差模拟电路故障诊断领域具有较好的应用前景。     

子网络级模拟故障诊断的神经网络实现

利用容差模拟电路节点电压灵敏度序列守恒定理,得到了模拟电路元件的软、硬故障统一样本。然后利用统一样本集训练BP神经网络,并将神经网络用于子网络级模拟故障诊断。实例验证表明,软、硬故障统一样本集使得用于神经网络训练所需样本数目大大减少,但经过训练的神经网络可以诊断容差模拟电路的全部软、硬故障,而且诊断正确率较高。     

基于PCA-LVQ的模拟电路故障诊断

为了解决模拟电路故障难于识别的问题,提出一种基于主成分分析(PCA)和学习矢量量化神经网络(LVQ)的模拟电路故障诊断新方法。该方法用PCA提取模拟电路故障特征,然后将降维后的故障特征信息输入LVQ网络训练和故障模式的分类识别。通过对Sallen-Key带通滤波器电路的故障诊断实例表明,该方法是有效的,具有较高的故障诊断率。     

浅析模拟与混合信号电路测试与故障检测方法

模拟和混合信号电路本身具有相当高的复杂性及专业性,使得模拟和混合信号电路测试与故障检测无法在传统数字电路测试方法下得到满足。文章通过介绍模拟和混合信号电路测试与故障检测的研究现状,分析了模拟与混合信号电路的测试与故障检测方法,并在传统测试技术的基础上研究了新的诊断方法,具有参考价值。     

改进LDA在模拟电路故障预测中的应用

针对模拟电路早期故障识别难度大的问题,提出一种改进线性辨别分析法和隐马尔科夫相结合的故障预测方法。首先设置元件的参数,提取幅频特征;然后采用改进的线性辨别分析（LDA）对电压特征进行提取消除特征的冗余性和高维性;最后将提取的特征用于训练和测试HMM,以实现模拟电路的状态监测。通过实验验证了其具有良好的模拟电路早期故障监测能力。     

行阶梯观测矩阵下语音压缩感知观测序列的Volterra+Wiener模型研究

针对压缩感知理论下,语音信号经随机高斯矩阵投影后得到的观测序列随机性太强,难以建模的问题,提出了一种基于行阶梯观测矩阵的语音压缩感知观测序列的Volterra模型,利用该模型实现对语音压缩感知观测序列的预测,研究了Volterra滤波器输入维数与阶数对预测效果的影响,并利用维纳滤波器进一步降低预测误差。在相同的已知数据量下,基于部分压缩感知观测序列、Volterra模型、Wiener滤波器的重构,获得了优于高斯随机观测序列的重构性能。模型的研究为压缩感知与语音技术的结合提供一定的参考价值。      

一种模拟电路故障诊断方法研究

将LSSVM算法应用于模拟电路故障诊断模型,使用PSO算法对LSSVM算法的参数进行寻优。以带通滤波器电路和双二次高通滤波器电路的故障诊断实例对该文研究的模拟电路故障诊断方法进行验证。使用三层小波包分解输出电压信号,得到8个频带能量特征向量,通过Monte Carlo仿真得到数据样本,用于故障诊断模型的训练和测试。结果表明,该文使用的改进LSSVM算法构建的故障诊断模型针对8种故障的诊断准确率均高于95%,具有较好的故障诊断性能。     

混合因子分析的重新抽样方法

混合因子分析是一种对具有复杂结构的多维数据建立模型的方法.本文提出了一种进行混合因子分析的重新抽样方法.当给定一组数据样本时,我们首先建立样本概率分布的混合高斯模型,然后为每一个高斯混合项重新抽取新的数据样本,在新的样本上再对每一个高斯混合项进行因子分析.与已有的算法相比较,避免了计算各个高斯混合项在每个样本值之下的后验概率,又减少了进行因子分析时参与计算的数据样本的数量.     

A Novel Method for the Diagnosis of the Incipient Faults in Analog Circuits Based on LDA and HMM

Diagnosis of incipient faults for electronic systems, especially for analog circuits, is very important, yet very difficult. The methods reported in the literature are only effective on hard faults, i.e., short-circuit or open-circuit of the components. For a soft fault, the fault can only be diagnosed under the occurrence of large variation of component parameters. In this paper, a novel method based on linear discriminant analysis (LDA) and hidden Markov model (HMM) is proposed for the diagnosis of incipient faults in analog circuits. Numerical simulations show that the proposed method can significantly improve the recognition performance. First, to include more fault information, three kinds of original feature vectors, i.e., voltage, autoregression-moving average (ARMA), and wavelet, are extracted from the analog circuits. Subsequently, LDA is used to reduce the dimensions of the original feature vectors and remove their redundancy, and thus, the processed feature vectors are obtained. The LDA is further used to project three kinds of the processed feature vectors together, to obtain the hybrid feature vectors. Finally, the hybrid feature vectors are used to form the observation sequences, which are sent to HMM to accomplish the diagnosis of the incipient faults. The performance of the proposed method is tested, and it indicates that the method has better recognition capability than the popularly used backpropagation (BP) network.     

Diagnosis of Incipient Faults in Weak Nonlinear Analog Circuits

Aiming at the problem to diagnose incipient faults in weak nonlinear analog circuits, an approach is presented in this paper. The approach calculates the fractional Volterra correlation functions beforehand. The next step is to use the fractional Volterra correlation functions and different angle parameters of the fractional wavelet packet transform (FRWPT) to extract the fault signatures. Meanwhile, the computational complexity is analyzed. Then the variables of the fault signatures are constructed, which are used to form the observation sequences of the hidden Markov model (HMM). HMM is used to accomplish the fault diagnosis. The simulations show that the presented method can significantly improve the incipient fault diagnosis capability.     

Research on fault diagnosis of airborne fuel pump based on EMD and probabilistic neural networks

Airborne fuel pump is a key component of the airborne fuel system, which once fails will bring a huge negative impact on aircraft safety. Therefore, accurate, reliable and effective fault diagnosis must be performed. However, the current airborne fuel pump has several difficulties: fault samples shortage, high maintenance costs and low diagnostic efficiency. In this paper, after Failure Mode, Effects and Criticality Analysis (FMECA) of airborne fuel pump, an experimental platform of airborne fuel transfusion system is developed and then a fault diagnosis model based on empirical mode decomposition (EMD) and probabilistic neural networks (PNN) is established. Meanwhile, the diagnosis model is verified by practical experiments, and the sensor layout of the experimental platform is optimized. Firstly, the vibration signals and pressure signals under normal state and six types of typical fuel pump faults are acquired on the experimental platform. Then EMD method is applied to decompose the original vibration signals into a finite Intrinsic Mode Functions (IMFs) and a residual. Secondly, the energy of first four IMFs is extracted as vibration signals fault feature, combined with the mean outlet pressure to construct fault feature vectors. Then feature vectors are divided into training samples and testing samples. Training samples are used to train PNN fault diagnosis model and testing samples are used to verify the model. Finally, the experimental results show that only one pressure sensor and one y-axis vibration sensor are needed to achieve 100% fault diagnosis. Furthermore, compared with SVM and GA-BP, the PNN fault diagnosis model has fast convergence, high efficiency and a higher performance and recognition for the typical faults of airborne fuel pump.     

Synergistic use of soft computing technologies for fault detection in gas turbine engines

In this paper, we present a synergistic approach to startup fault detection and diagnosis (FDD) in gas turbine engines. The method employs statistics, signal processing, and soft computing techniques in a complementary manner to address fault detection at transient conditions. Traditional turbine engine FDD methods are based on engine data collected at steady-state conditions. However, incipient faults are difficult to diagnose using steady-state engine data; only engine faults that are fairly developed can be detected using conventional methods. Because incipient engine component faults are often manifest in the engine startup characteristics, we present a method to characterize the engine transient startup. Engine sensor data during engine startup are recorded in time series format. The sensor profiles corresponding to "good" and "bad" engine startups are sampled using the bootstrap technique. A feature vector is extracted in two steps, and signal processing is followed by the feature vector selection. In the signal processing step, principal component analysis (PCA) is applied to reduce the samples consisting of sensor profiles into a smaller set. In the feature vector selection step, a cost function is defined, and important discriminating features for fault diagnosis are distilled from the PCA output vector. The features obtained from this step are then classified using neural-network-based methods. The "leave-one-out" approach to cross validation is applied to obtain an objective evaluation of the neural network training. The proposed FDD method is evaluated using actual engine startup data, and the results are presented.     

Incipient Fault Detection Using an Associated Adaptive and Sliding-Mode Observer for Quadrotor Helicopter Attitude Control Systems

An associated adaptive and sliding-mode observer (AASMO) design is proposed to detect and estimate the incipient actuator faults of a quadrotor. The incipient faults considered are physical structure aging and quadrotor leakage. First, disturbances and nonlinear parameters are considered in system formulation for a realistic mathematical model of the quadrotor. Its fault model is also introduced. Second, the decomposed subsystems are obtained through coordinate transformations to separate the incipient faults from the disturbances. For the subsystem with no disturbance, the adaptive observer can estimate the incipient faults. For the subsystem with disturbances, the sliding-mode observer has strong robustness against the disturbances. Dynamic error convergence and system stability can also be guaranteed by Lyapunov stability theory. Finally, the simulation results of quadrotor helicopter attitude systems validate the efficiency of the proposed AASMO-based incipient fault detection algorithm.     

Fault clustering: modeling and observation on experimental LSI chips

Occurrence of fault clustering on large-scale integrated (LSI) MOS product was verified with optical microscopes on experimental chips that failed electrical testing. Two methods were used for determining clustering: analysis of the fault density derived from collected fault data, and separation of faults into two populations, one representing solitary faults, the other clusters. A model for the first method is presented and its effectiveness examined on a simulated fault set. The method is then applied to fault data representing two samples of MOS LSI experimental product. Population separation is finally carried out on one of the data samples, and the clustering data developed from this process are expressed by two factors. One factor can be used for refined yield estimates, the other was applied to quality measure calculations.     

Functional fault models for non-scan sequential circuits

The paper presents two functional fault models that are applied for functional delay test generation for non-scan synchronous sequential circuits: the pin pair state (PPS) fault model and the pin pair full state (PPFS) fault model. The PPS fault model deals with the pairs of stuck-at faults on the primary inputs and the primary outputs, as well as, with the pairs of stuck-at faults on the previous state bits and the primary outputs. The PPFS fault model encompasses the PPS model, and additionally deals with the pairs of stuck-at faults on the primary inputs and the next state bits, as well as, with the pairs of stuck-at faults on the previous state bits and the next state bits. The main factor in assessing the quality of obtained test sequences was the transition fault coverage at the gate level of the selected according to the appropriate fault model test sequences from the generated randomly ones. The experimental results demonstrate that the implementation using presented functional fault models allow selecting the test sequences from the initial test set without the loss of transition fault coverage in many cases, and the number of the selected test sequences is much lesser than that of the initial test set. This result demonstrates that the functional delay test can be generated using the presented functional delay fault models before structural synthesis of the circuit.     

免疫粒子群算法及其在矿井提升机故障诊断中的应用

基于人工免疫系统的故障诊断方法是人工智能领域发展起来的一个十分活跃的分支.为了提高免疫算法在矿井提升机故障诊断系统中的执行效率,通过对诊断问题进行更精确的建模和分析,提出了将免疫模型和离散粒子群进化算法相结合的提升机系统的故障诊断方法.该方法在免疫形态空间中采用核主元形式的相似性度量,解决了传统距离判别函数法在故障诊断中存在误差较大等问题.仿真结果表明,该方法能够适应诊断过程中出现的不确定性,并实现多故障诊断.     

Sensorless fault diagnosis of induction motors

Early detection and diagnosis of incipient faults is desirable for online condition assessment, product quality assurance, and improved operational efficiency of induction motors. In this paper, a speed-sensorless fault diagnosis system is developed for induction motors, using recurrent dynamic neural networks and multiresolution or Fourier-based signal processing for transient or quasi-steady-state operation, respectively. In addition to nameplate information required for the initial system setup, the proposed fault diagnosis system uses only motor terminal voltages and currents. The effectiveness of the proposed diagnosis system in detecting the most widely encountered motor electrical and mechanical faults is demonstrated through extensive staged faults. The developed system is scalable to different power ratings and it has been successfully demonstrated with data from 2.2, 373 and 597 kW induction motors.     

Practical Analog Circuit Diagnosis Based on Fault Features with Minimum Ambiguities

As numerous faults exist in practical analog circuits, new challenges arise in the field of diagnosis with large-scale target faults as well as fault features. To address this issue, firstly, an ambiguity model is built to measure the distinguishability between two faults. Then, the optimal fault features are obtained by analyzing the response curves of the circuit under test (CUT) to minimize the ambiguities among the faults. Finally, comparisons are made among three classification methods, including the maximum likelihood classifier (MLC), artificial neural networks (ANNs) and support vector machine (SVM), to demonstrate their own diagnostic abilities for practical use. Two examples are illustrated, and taking advantage of an automated implementation framework, 92 faults in total are examined in the second example. The experimental results show that good diagnostic performances can be obtained with the proposed method. However, when a practical case is encountered, the ANNs method may fail due to its high time and space complexity, while the MLC and SVM methods are still applicable.