A segmental hidden semi-Markov model (HSMM)-based diagnostics and prognostics framework and methodology

Diagnostics and prognostics are two important aspects in a condition-based maintenance (CBM) program. However, these two tasks are often separately performed. For example, data might be collected and analysed separately for diagnosis and prognosis. This practice increases the cost and reduces the efficiency of CBM and may affect the accuracy of the diagnostic and prognostic results.In this paper, a statistical modelling methodology for performing both diagnosis and prognosis in a unified framework is presented. The methodology is developed based on segmental hidden semi-Markov models (HSMMs). An HSMM is a hidden Markov model (HMM) with temporal structures. Unlike HMM, an HSMM does not follow the unrealistic Markov chain assumption and therefore provides more powerful modelling and analysis capability for real problems. In addition, an HSMM allows modelling the time duration of the hidden states and therefore is capable of prognosis. To facilitate the computation in the proposed HSMM-based diagnostics and prognostics, new forward–backward variables are defined and a modified forward–backward algorithm is developed. The existing state duration estimation methods are inefficient because they require a huge storage and computational load. Therefore, a new approach is proposed for training HSMMs in which state duration probabilities are estimated on the lattice (or trellis) of observations and states. The model parameters are estimated through the modified forward–backward training algorithm. The estimated state duration probability distributions combined with state-changing point detection can be used to predict the useful remaining life of a system.The evaluation of the proposed methodology was carried out through a real world application: health monitoring of hydraulic pumps. In the tests, the recognition rates for all states are greater than 96%. For each individual pump, the recognition rate is increased by 29.3% in comparison with HMMs. Because of the temporal structures, the same HSMMs can be used to predict the remaining-useful-life (RUL) of the pumps.     

基于隐半马尔可夫模型设备退化状态识别方法研究

机械设备从正常到故障往往经历一系列退化状态,正确识别与估计设备当前所处的退化状态,对预防设备进一步退化和故障的发生具有重要意义。隐半马尔可夫模型(Hidden Semi-MarkovModels,HSMM)是隐马尔可夫模型(hidden Markov models,HMM)的一种扩展模型,克服了因马尔可夫链的假设造成HMM建模所具有的局限性,比HMM具有更好的建模能力和分析能力。由状态识别和HMM本质上的相通性,将HSMM引入到机械设备的状态识别中,提出了一种基于HSMM状态识别方法,描述了该模型的拓扑结构和主要参数以及相应的训练和识别算法。最后通过滚动轴承试验系统验证了方法的有效性。     

A prognosis method using age-dependent hidden semi-Markov model for equipment health prediction

Health monitoring and prognostics of equipment is a basic requirement for condition-based maintenance (CBM) in many application domains. This paper presents an age-dependent hidden semi-Markov model (HSMM) based prognosis method to predict equipment health. By using hazard function (h.f.), CBM is based on a failure rate which is a function of both the equipment age and the equipment conditions. The state values of the equipment condition considered in CBM, however, are limited to those stochastically increasing over time and those having non-decreasing effect on the hazard rate. The previous HSMM based prognosis algorithm assumed that the transition probabilities are only state-dependent, which means that the probability of making transition to a less healthy state does not increase with the age. In the proposed method, in order to characterize the deterioration of equipment, three types of aging factors that discount the probabilities of staying at current state while increasing the probabilities of transitions to less healthy states are integrated into the HSMM. With an iteration algorithm, the original transition matrix obtained from the HSMM can be renewed with aging factors. To predict the remaining useful life (RUL) of the equipment, hazard rate is introduced to combine with the health-state transition matrix. With the classification information obtained from the HSMM, which provides the current health state of the equipment, the new RUL computation algorithm could be applied for the equipment prognostics. The performances of the HSMMs with aging factors are compared by using historical data colleted from hydraulic pumps through a case study.     

A weighted hidden Markov model approach for continuous-state tool wear monitoring and tool life prediction

Tool wear is one of the important indicators to reflect the health status of a machining system. In order to obtain tool’s wear status, tool condition monitoring (TCM) utilizes advanced sensor techniques, hoping to find out the wear status through those sensor signals. In this paper, a novel weighted hidden Markov model (HMM)-based approach is proposed for tool wear monitoring and tool life prediction, using the signals provided by TCM techniques. To describe the dynamic nature of wear evolution, a weighted HMM is first developed, which takes wear rate as the hidden state and formulates multiple HMMs in a weighted manner to include sufficient historical information. Explicit formulas to estimate the model parameters are also provided. Then, a particular probabilistic approach using the weighted HMM is proposed to estimate tool wear and predict tool’s remaining useful life during tool operation. The proposed weighted HMM-based approach is tested on a real dataset of a high-speed CNC milling machine cutters. The experimental results show that this approach is effective in estimating tool wear and predicting tool life, and it outperforms the conventional HMM approach.     

基于FRWT的模拟电路早期故障诊断

针对模拟电路早期故障诊断的难题,基于分数阶小波转换(fractional wavelet trarsform,FRWT)并结合隐马尔科夫模型(hidden Markov model,HMM),提出了一种模拟电路故障特征分析的新方法。首先将无故障状态和各故障状态下模拟待测试电路(circuit under test,CUT)的响应序列进行分数阶小波分解得到子带响应序列,然后从子带响应序列提取出故障特征向量并构成观测序列训练出HMM,最后利用训练好的HMM对未知状态电路进行诊断。实验结果表明,该方法能有效提取模拟电路的故障特征,完成模拟电路早期故障检测和故障定位。     

基于无限隐Markov模型的旋转机械故障诊断方法研究

针对传统隐Markov模型(HMM)在机械故障诊断中存在的不足,即HMM过学习或溢出问题以及隐状态数需要事先假定,提出了基于无限隐马尔可夫模型(i HMM)的机械故障诊断方法。在提出的方法中,以谱峭度为特征提取,i HMM为识别器,并以最大似然估计来确定设备运转中出现的故障类型。同时,将提出的方法与传统的HMM故障识别方法进行了对比分析。实验结果表明,提出的方法是有效的,得到了非常满意的识别效果。提出的方法能够有效避免了HMM在建模初期遗留下的不足,可以自适应确定模型中隐藏状态数和模型数学结构,因此,提出的方法明显优于HMM故障识别方法。     

主分量分析和因子隐Markov模型在机械故障诊断中的应用

主分量分析(principal component analysis,PCA)是统计学中分析数据的一种有效方法,可以将高维数据空间变换到低维特征空间,因而可用于多通道冗余消除和特征提取.因子隐Markov模型是隐Markov模型的扩展,它比隐Markov模型更有优势,适用于动态过程时间序列的建模,并具有强大的时序模型分类能力,特别适合非平稳、信号特征重复再现性不佳的信号分析.文中结合主分量分析与因子隐Markov模型,提出一种新的故障识别方法,即以主分量分析方法进行冗余消除和故障特征提取,因子隐Markov模型作为分类器.并应用到机械故障诊断中,同时与基于主分量分析的隐Markov模型的识别方法相比较,实验结果表明基于PCA的因子隐Markov模型识别法和基于PCA的隐Markov模型识别法在故障识别上都是有效的,但对于相同的状态空间,前者的训练速度快于后者,尤其是状态空间越大,这种优势越明显.     

分层隐Markov模型在设备状态识别中的应用研究

与传统的隐Markov模型（HMM）相比较而言,应用分层隐Markov模型（HHMM）对设备进行状态识别有诸多优点,而且能以概率的形式更为精确地计算识别结果。针对模型参数随着设备状态的增加呈指数倍增这一问题,引入动态贝叶斯网络这一新的方法,由于该方法可以有效地降低模型的计算复杂度并缩短推理时间,所以将HHMM表达为动态贝叶斯网络,利用预处理的振动信号对设备的健康状态进行识别;针对现有状态分类方法的局限性,提出了基于K均值算法和交叉验证方法相结合的状态数优化方法;以齿轮箱全寿命实验为依据,对该模型实现状态识别的基本框架和计算过程进行了研究,研究结果为复杂设备的状态识别提供了新的思路。     

变转速变载荷下滚动轴承故障的小波-HMM诊断

常规的谱分析等方法难以对滚动轴承变转速变载荷工况进行故障诊断，为此，采用具有时频局域特征的小波分析法。对变转速变载荷工况下滚动轴承的振动信号。用小波包分解法提取各频带的能量作为特征参数。再采用连续隐Markov模型（HMM）对滚动轴承的状态进行识别。试验证明，小波-HMM可以在变转速变载荷工况下以及未知转速情况下对滚动轴承的各种故障有效地进行诊断。     

基于改进HMM的潜在电子故障状态识别模型

针对复杂电子装备隐性故障难以诊断的难题,在深入分析隐马尔可夫模型的核心问题及基本算法的基础上,探讨了其在故障诊断应用中存在的主要问题,建立了多状态电子装备可靠性评估模型,利用系统可靠性评估结果作为隐马尔可夫模型的初始模型特征量,改进了传统的隐马尔可夫模型,并对Baum-Welch训练算法进行了优化,形成了一套适于复杂电子装备潜在故障状态跟踪识别的数学模型.实验结果显示,理论方法及模型能够更好地识别潜在故障状态,加快了模型训练速度,提高了故障状态识别率.     

FAULT DIAGNOSIS APPROACH BASED ON HIDDEN MARKOV MODEL AND SUPPORT VECTOR MACHINE

Aiming at solving the problems of machine-learning in fault diagnosis, a diagnosis approach is proposed based on hidden Markov model (HMM) and support vector machine (SVM). HMM usually describes intra-class measure well and is good at dealing with continuous dynamic signals. SVM expresses inter-class difference effectively and has perfect classify ability. This approach is built on the merit of HMM and SVM. Then, the experiment is made in the transmission system of a helicopter. With the features extracted from vibration signals in gearbox, this HMM-SVM based diagnostic approach is trained and used to monitor and diagnose the gearbox's faults. The result shows that this method is better than HMM-based and SVM-based diagnosing methods in higher diagnostic accuracy with small training samples.     

Condition monitoring and classification of rotating machinery using wavelets and hidden Markov models

Condition monitoring and classification of machinery state is of great practical significance in manufacturing industry, because it provides updated information regarding machine status on-line, thus avoiding the production loss and minimising the chances of catastrophic machine failure. In this paper, the condition classification is based on hidden Markov models (HMMs) processing information obtained from vibration signals. We present an on-line fault classification system with an adaptive model re-estimation algorithm. The machinery condition is identified by selecting the HMM which maximises the probability of a given observation sequence. The proper selection of the observation sequence is a key step in the development of an HMM-based classification system. In this paper, the classification system is validated using observation sequences based on the wavelet modulus maxima distribution obtained from real vibration signals, which has been proved to be effective in fault detection in previous research.     

基于量子粒子群优化Volterra时域核辨识的隐Markov模型识别方法

将量子粒子群优化算法引入Voherra级数模型的非线性辨识中,并结合隐Markov模型(hidden Markov model,HMM),提出了一种基于量子粒子群优化的Voherra时域核特征提取的HMM识别方法,在提出的方法中,利用量子粒子群优化算法辨识得到的前三阶Volterra时域核作为故障特征,输入到各种状态的HMM中,其中,输出概率最大的HMM对应的状态即为设备的当前运行状态.提出的方法克服了传统的基于Volterra模型系统的机械故障诊断要求目标函数连续可导、容易陷入局部最小以及抗干扰能力差等缺陷.最后,将提出的方法应用到旋转机械故障诊断中.实验结果验证了该方法的有效性.     

Detection and diagnosis of bearing and cutting tool faults using hidden Markov models

Over the last few decades, the research for new fault detection and diagnosis techniques in machining processes and rotating machinery has attracted increasing interest worldwide. This development was mainly stimulated by the rapid advance in industrial technologies and the increase in complexity of machining and machinery systems. In this study, the discrete hidden Markov model (HMM) is applied to detect and diagnose mechanical faults. The technique is tested and validated successfully using two scenarios: tool wear/fracture and bearing faults. In the first case the model correctly detected the state of the tool (i.e., sharp, worn, or broken) whereas in the second application, the model classified the severity of the fault seeded in two different engine bearings. The success rate obtained in our tests for fault severity classification was above 95%. In addition to the fault severity, a location index was developed to determine the fault location. This index has been applied to determine the location (inner race, ball, or outer race) of a bearing fault with an average success rate of 96%. The training time required to develop the HMMs was less than 5 s in both the monitoring cases.     

基于HMM-SVM的故障诊断模型及应用

针对直升机减速器故障诊断中机器学习方法存在的问题，根据隐马尔可夫模型（HMM）适合于处理连续动态信号与支持向量机（SVM）适合于模式分类的长处，提出了基于HMMSVM串联结构的故障诊断模型。通过从减速箱振动信号中有效提取AR特征，利用HMM汁算未知信号与减速器各状态的匹配程度，形成特征向量提供给SVM最后判别，实验结果表明该方法优于单纯的HMM或SVM诊断方法，能利用少量训练样本有效地完成直升机减速器的故障诊断。     

Fault detection in roller bearing operating at low speed and varying loads using Bayesian robust new hidden Markov model

This paper uses Bayesian robust new hidden Markov modeling (BRNHMM) for bearing fault detection and diagnosis based on its acoustic emission signal. A variational Bayesian approach is used that simultaneously approximates the distribution over the hidden states and parameters with simpler distribution hence using Bayesian inference for the estimation of the posterior HMM hyperparameters. This allows for online detection as small data sets can be used. Also, the Kullback-Leibler (KL) divergence is effectively used to access the divergence of the probability function of the BRNHMM, to find its lower bound approximation and by applying a linear transform to the maximum output probability parameter generation (MOPPG). The training set result obtained from BRNHMM is then compared to the result from artificial neural network (ANN) fault detection for same complex system of low speed and varying load conditions which are difficult from a diagnostic perspective, as found in rolling mills.     

机电系统BIT的非永久故障分析建模与诊断

非永久故障是导致机电系统BIT（built-in test）虚警的一个主要原因,诊断非永久故障既可保证BIT的高故障检测率,同时又可有效地抑制虚警。但是目前缺乏对非永久故障的机理分析与建模,相应的诊断研究也很少。在分析研究非永久故障的表现形式、产生原因与机理的基础上,对被测系统的状态进行马尔可夫建模,再根据被测系统和隐马尔可夫模（hidden Markov model,HMM）的状态都是通过表现来感知的特点,利用HMM对BIT被测系统建模,并提出基于HMM的BIT非永久故障诊断方法,最后通过实验验证表明,此方法能有效地诊断非永久故障,降低BIT虚警。     

基于连续高斯密度混合HMM的滚动轴承故障诊断研究

滚动轴承在直升机的传动系统中占有十分重要的地位，对其进行快速有效的状态监测与故障诊断具有重大意义。由故障诊断和隐马尔可夫模型(Hidden Markov Model，HMM)本质上的相通性，利用连续高斯密度混合隐马尔可夫模型分析滚动轴承的振动信号，先以基于短时傅里叶变换的倒谱系数为特征训练模型，再利用模型进行状态监测和故障诊断，实验结果表明该方法能利用少量样本进行训练和有效诊断，且具有训练时间短、诊断速度快的优点。     

航天发射场供气系统健康管理技术

针对新一代航天发射场采用全新的在线供气模式,难以有效评估单样本设备健康状态的问题,提出一种基于隐马尔可夫的设备健康状态管理与预测方法。首先,利用设备监测数据构建隐马尔可夫健康状态评估模型,通过对不同观测序列与不同观测次数下的预测准确率进行仿真,确定出最优的模型参数;其次,把实时数据代入模型,根据模型的计算结果取最小值,从而判断出设备的健康状态;最后,将当前数据与历史数据进行拟合,预测出系统的安全可靠寿命。经实际检验,该方法有效解决了单样本多状态设备的健康评估。     

基于AR-连续HMM的故障诊断模型及应用

在状态监测与故障诊断中,被测设备的状态一般不能直接观察到,要通过测量被测设备的表现来感知,这和隐马尔可夫模型(HMM)在本质是相通的。因此可以利用连续高斯密度混合HMM分析被测设备的振动信号,首先以AR模型系数为特征,研究不同状态数与不同混合高斯数对HMM模型分类的影响,再利用较优的状态数与混合高斯数HMM模型进行状态监测和故障诊断,诊断与对比实验结果表明该方法能利用少量样本进行训练和有效诊断。