Dynamic Fault Detection Research Articles

Adaptive State Observer for Robot Manipulators Diagnostics and Health Degree Assessment

To avoid serious damages caused by the dynamic environment, fault detection and health assessment are essential for an integrated robotic system. In this paper, we propose a fault detection algorithm and a health degree assessment approach for a robot manipulator system. Both the internal disturbance and the output measurement disturbance are considered in the proposed method. In addition, an adaptive observer is utilized to reconstruct the real system of robot manipulators. Under the proposed observer, the real system is estimated to detect the fault and obtain the health degree of the robot manipulator. The feasibility and reliability of the proposed fault detection algorithm and health degree assessment index for robot manipulator systems are proved by simulation experiments.

面向云计算环境的动态故障检测与诊断方法

面向云计算环境的动态故障检测与诊断方法

Tensor sequence component analysis for fault detection in dynamic process

For dynamic processes, using sequence information to augment the data can improve fault detection performance. Traditional approaches transform raw data into augmented vectors, which leads to losses in structural information in the variables and increases the data dimension. This paper proposes a novel data dimension reduction algorithm called tensor sequence component analysis (TSCA) and applies it to dynamic process fault detection. The algorithm extends each sample into a matrix comprising current and past process data, and simultaneously reduces the dimensions of time delay and the variables for feature extraction, solving the problem of the curse of dimensionality. For the dimension reduction of time delay, in order to extract similar information from the samples, each sample is reconstructed with time neighbourhoods. For the dimension reduction of the variables, considering the information of different variables variance information of the latent variables is maximized for feature extraction. Finally, a numerical example and the Tennessee Eastman process are used to demonstrate the efficacy of the proposed method.

Dynamic process fault detection and diagnosis based on a combined approach of hidden Markov and Bayesian network model

The present study introduces a novel methodology for fault detection and diagnosis (FDD), based on a combined approach of data and process knowledge driven techniques. The Hidden Markov Model (HMM) detects the abnormalities based on process history while the Bayesian Network (BN) diagnoses the root causes of faults. An HMM is trained with standard operating condition data while the structure of BN is developed based on process knowledge. The log-likelihoods (LL) of process historical data strings are used to establish the conditional probability tables of the BN. HMM identifies the abnormal behaviour of the system. The time of detection of abnormality, respective log-likelihood value, and the probabilities of being in the process condition at the time of detection are used as evidence to BN. The updated BN is then used to diagnose the root cause by considering the respective changes in the probabilities. Performance of the new technique is tested and validated with published data of Tennessee Eastman Process. Eight of the ten selected faults were successfully detected and diagnosed.

Therapeutic drug level measurement in breast milk

In this paper, design issues of data-driven optimal dynamic fault detection systems for stochastic linear discrete-time processes are addressed without precise distribution knowledge of unknown inputs and faults. Concerning a family of faults with different distribution profiles in mean and covariance matrix, we introduce a bank of parameter vectors of parity space and construct the parity relation based residual generators using process input and output data. In the context of minimizing the missed detection rate for a prescribed false alarm rate, the design of fault detection system is formulated as a bank of distribution independent optimization problems without posing specific distribution assumption on unknown inputs and faults. It is proven that the optimal selection of individual parameter vector can be formulated as a generalized eigenvalue–eigenvector problem in terms of the means and covariance matrices of residuals in fault-free and each faulty cases, and is thus solved via singular value decomposition. The tight upper bounds of false alarm rate and missed detection rate are simultaneously achieved quantitatively. Besides, the existence condition of the optimal solutions is investigated analytically. Experimental study on a three-tank system illustrates the application of the proposed scheme.

Fault diagnosis and economic performance evaluation for a simulated base metal leaching operation

A dynamic process model based on a base metal refinery including critical control layers has been developed. The critical control layers include sensors, actuators, regulatory controllers, alarm systems, safety interlocks and supervisory control. With the help of expert knowledge, a fault (abnormal event) library was incorporated into the dynamic model.Fault diagnosis methods are used to detect and identify abnormal process conditions. With the use of the dynamic process model, fault detection and identification methods can be more rigorously and accurately evaluated for hydrometallurgical industry use.An economic performance function is developed from expert knowledge. Once the fault diagnosis is complete, an economic impact analysis based on the fault diagnosis results is completed. A possible economic case for fault diagnosis is concluded from the results.

A novel dynamic PCA algorithm for dynamic data modeling and process monitoring

Principal component analysis (PCA) has been widely applied for data modeling and process monitoring. However, it is not appropriate to directly apply PCA to data from a dynamic process, since PCA focuses on variance maximization only and pays no attention to whether the components contain dynamics or not. In this paper, a novel dynamic PCA (DiPCA) algorithm is proposed to extract explicitly a set of dynamic latent variables with which to capture the most dynamic variations in the data. After the dynamic variations are extracted, the residuals are essentially uncorrelated in time and static PCA can be applied. The new models generate a subspace of principal time series that are most predictable from their past data. Geometric properties are explored to give insight into the new dynamic model structure. For the purpose of process monitoring, fault detection indices based on DiPCA are developed based on the proposed model. Case studies on simulation data, data from an industrial boiler process, and the Tennessee Eastman process are presented to illustrate the effectiveness of the proposed dynamic models and fault detection methods.

A comparison of static and dynamic fault detection techniques for transcritical refrigeration

Repairing faults in heating, ventilation, air conditioning, and refrigeration (HVAC&R) systems can improve overall system energy efficiency and prevent leaked refrigerant, thereby generating substantial economic and environmental benefits. In this paper, static and dynamic fault detection and diagnosis (FDD) metrics are examined for both sub-critical and transcritical refrigeration cycles, with particular emphasis on transcritical cycles. Virtual sensors based on measurements from low-cost sensors are used to identify faulty behaviour. This paper demonstrates in simulation and experiment that current low-cost, static FDD techniques are capable of detecting multiple soft faults, even in the presence of secondary system faults. These results are shown for both subcritical and transcritical refrigeration cycles. The potential benefits of dynamic FDD over currently-available static methods are assessed through both simulation and experiment. This analysis illustrates that although there may be minor benefit in using dynamic FDD algorithms, there is limited increase in signal sensitivity with a large cost in increased complexity of the FDD algorithms.

Speculative Lookahead for Energy-Efficient Microprocessors

In addition to being the in situ performance monitor for adaptive voltage scaling (AVS), timing speculation mechanisms (e.g., razor) featuring dynamic timing fault detection and correction help to relax timing constraints for simple logic structures and low-power cells. Conventional timing fault detection mechanisms require substantial buffers to prevent race conditions on short paths for double sampling, which can overwhelm energy savings from timing relaxation and voltage scaling. This paper proposes a novel timing speculation scheme, speculative lookahead (SL), comprising duplicate timing-relaxed datapaths, the short paths of which do not introduce race conditions and thus require no additional buffer insertion. In experiments using a 40-nm CMOS technology, SL consumed a 54.89% area of a razor-based 32-bit multiplier, and conserved 59.77% energy per operation at nominal 1.1 V and 53.49% when AVS was applied. An ARM Cortex M0-like microprocessor unit (MPU) was designed using an SL-based datapath, the timing fault detection and correction mechanism of which can be dynamically deactivated for latency-tolerant instructions [i.e., on-demand timing speculation (ODTS)] to further conserve up to 31.08% energy in the execution unit. In addition, an field-programmable gate array prototype of the SL/ODTS MPU was constructed to demonstrate the effectiveness of delay variation tolerance and implementation flexibility.

Dynamic process fault detection and diagnosis based on dynamic principal component analysis, dynamic independent component analysis and Bayesian inference

Dynamic principal component analysis (DPCA) and dynamic independent component analysis (DICA), as the frequently-used dimensional reduction methods, have been widely applied to monitor dynamic process. Considering the respective advantages of DPCA and DICA in different data distribution characteristics, this paper proposes a novel process monitoring algorithm named DPCA, DICA and Bayesian Inference (DPCA–DICA–BI). The main idea of DPCA–DICA–BI is to put the process variables with same distribution characteristic (Gaussian or non-Gaussian) into a block on the basis of the variable normality by Jarque–Bera test and then to respectively apply DPCA and DICA in Gaussian and non-Gaussian blocks. Finally, to combine the monitoring performance of both blocks, Bayesian inference is employed to make an integrated decision. The DPCA–DICA–BI as well as PCA, ICA, DPCA, and DICA has been used to a numerical example and Tennessee Eastman process. The simulation results show the superiority of DPCA–DICA–BI.

A variational Bayesian robust linear dynamic system approach for dynamic process modelling and fault detection

In this work, a variational Bayesian robust linear dynamic system (VBRLDS) approach is proposed for dynamic process modeling and monitoring. Traditional linear dynamic system (LDS) constructed with Kalman filter is designed by Gaussian assumption which can be easily violated in outlier contaminated modeling situations. To deal with this issue, the conventional Gaussian-based Kalman filter is modified with heavy tailed Student's t-distribution so as to deal with modeling outliers. Then, a variational Bayesian expectation maximization (VBEM) algorithm is developed for learning parameters of the robust linear dynamic system. For process monitoring, traditional residual space modeling method of SPE statistics are modified. To explore the feasibility and effectiveness, our proposed method is applied into fault detection and is comparatively studied with several other methods on the Tennessee Eastman benchmark.

A Real-Time Fault Prewarning Approach to Generator Sets Based on Dynamic Threshold

Fault prewarning is important to guarantee the safe and stable operation of Generator Sets (GS). In order to generate prewarnings quickly and accurately before the failures or faults occur in GS, a real-time fault prewarning approach to GS based on dynamic threshold was put forward. This approach was consisted of five steps, that is operating condition (or abnormal event) synchronous, dynamic threshold selection, threshold analysis, fault detection and fault prewarning. The dynamic threshold (closely related to operating condition or abnormal event of GS) was the key of this approach, which can be obtained by means of expertise knowledge discovery. This approach can effectively reduce false positives and false negatives for the faults of GS, whose effectiveness is validated by the applications and practices of Gezhouba power plant.

Time Neighborhood Preserving Embedding Model and Its Application for Fault Detection

By incorporating the serial correlations of the process data, a new dimensionality reduction method, named time neighborhood preserving embedding (TNPE) is proposed and applied for dynamic process fault detection. TNPE aims at preserving the local neighborhood structure of the data to learn the underlying geometry manifold. To describe the dynamic characteristic while constructing the neighborhood graph, the search for nearest neighbors is performed with respect to the time sequence adjacent points of each data point. Furthermore, the local dynamic variations of the high-dimensional data are captured by reconstructing each data point from its nearest neighbors. On the basis of the locally linear property of the selected neighbors, the estimated local structure information is robust to the transformations of linear projection. Consequently, by considering the serial correlations, TNPE is able to explore the meaningful dynamic information hidden in high-dimensional data. For fault detection, Hotelling’s T2 ...

Advances in Data-Driven Monitoring Methods for Complex Process

In modern industrial processes, effective performance monitoring and quality prediction are the key to ensure plant safety and enhance product quality. The research significance and background of process monitoring and fault diagnosis technologies are described and the current advances in data-based process monitoring methods are summed up in this paper. Then the multivariate statistical process control (MSPC) methods for process with single constraint, especially for single non-Gaussian process or nonlinear process are elaborated. As real industrial process data often show strong non-Gaussian and dynamic behaviors, study on monitoring technologies for dynamic non-Gaussian process is of great importance. Finally, some challenges such as non-Gaussian and dynamic process, fault detection and diagnosis as well as new MSPC methods are indicated.

Dynamic risk assessment and fault detection using a multivariate technique

In the context of process safety, significant improvements are needed in fault detection methods, especially, in the areas of early detection and warning. In this article, a multivariate risk‐based fault detection and diagnosis technique is proposed. The key elements of this technique are to eliminate faults that are not serious and to provide a dynamic process risk indication at each sampling instant. A multivariable residual generation process based on the Kalman filter has been combined with a risk assessment procedure. The use of the Kalman filter makes the method more robust to false alarms, which is an important aspect of any fault detection algorithm that targets the safety of a process. In addition, we consider significant differences in the severity of the faults associated with different process variables. We also take into account the varying intensity of damage caused by the increasing and decreasing rates of fault and the need to treat those cases differently. © 2013 American Institute of Chemical Engineers Process Saf Prog 32: 365–375, 2013

Dynamic Risk Assessment and Fault Detection Using Principal Component Analysis

A methodology to calculate process risk in combination with a data based fault detection method is proposed in this paper. The proposed approach aims to identify and screen the faults which are not...

Dynamic observer-based robust control and fault detection for linear systems

This study is concerned with the design of dynamic observer-based robust controller that also facilitates the acquisition of information used for fault detection (FD) purpose in feedback control systems. Through introducing a weighting matrix, the combination of observer states is utilised to generate a residual signal to detect faults. The first technical contribution is to construct a new linearising change-of-variables that is able to convert the dynamic observer-based controller design problem into linear matrix inequality-based optimisation problem. The second one is to show that the proposed dynamic observer-based controller can achieve a better H∞ performance compared with the existing static (Luenberger) observer-based controller design approaches. Finally, via the simple residual structure, a convex fault detector design condition with some parameter matrices fixed is developed for guaranteeing the H− performance used to measure the fault sensitivity. An F-18 aircraft model is given to show the satisfactory FD performance and control performance.

Dynamic fault detection and isolation for automotive engine air path by independent neural network model

Fault detection and isolation have become one of the most important aspects of automobile design. A new fault detection and isolation scheme is developed for automotive engines in this paper. The method uses an independent radial basis function neural network model to model engine dynamics, and the modelling errors are used to form the basis for residual generation. Furthermore, another radial basis function network is used as a fault classifier to isolate occurred fault from other possible faults in the system by classifying fault characteristics embedded in the modelling errors. The performance of the developed scheme is assessed using an engine benchmark, the mean value engine model with Matlab/Simulink. Five faults have been simulated on the mean value engine model, including three sensor faults, one component fault and one actuator fault. The three sensor faults considered are 10–20% changes superimposed on the measured outputs of manifold pressure, manifold temperature and crankshaft speed sensors; the component fault considered is air leakage in the intake manifold; the actuator fault considered is the malfunction of fuel injector. The simulation results show that all the simulated faults can be clearly detected and isolated in dynamic conditions throughout the engine operating range.

Pattern identification for eccentricity fault diagnosis in permanent magnet synchronous motors using stator current monitoring

A novel theoretical scrutiny is presented here, which extracts eccentricity fault signatures by monitoring the stator current in permanent magnet synchronous motors (PMSMs). In this analysis, effects of the stator slots and saturation are taken into account for static and dynamic eccentricity fault detection. Eccentricity signatures are utilised to introduce a particular frequency pattern, and amplitude of side-band components at proposed frequencies is employed as a proper index for eccentricity fault recognition. Competence of the nominated index to detect eccentricity, its type and its degree is investigated in faulty PMSM with different load levels. Hence, time-stepping finite-element method is used to model faulty PMSM and calculate stator currents for processing and obtaining aforementioned proposed index. The relation between the nominated index, static and dynamic eccentricity degrees is determined by a mutual information criterion. So, a white Gaussian noise is added to the simulated current and robustness of the proposed index is analysed with respect to the noise variance. Finally, the type and the degree of eccentricity are predicted using support vector machine as a classifier. The classification results indicate that the proposed features can estimate the eccentricity degree and its type. The simulation results are verified by the experimental results.

Robust target tracking using distributed unmanned aerial vehicle networks

In this article, a target tracking scheme for operation in distributed unmanned aerial vehicle (UAV) networks in which sensors may read the target position incorrectly is proposed. The proposed scheme operates two algorithms concurrently: semi-decentralized dynamic data fusion and fault detection. The semi-decentralized dynamic data fusion algorithm employs a median-consensus algorithm using extended non-faulty neighbours (whose sensor readings for the target position are within a prescribed tolerance level) and subsequently makes local estimates of the target position converge to nearly the actual target position. Meanwhile, the fault detection algorithm first asks each UAV to find the global median of the local target position through extended neighbours, and then diffuses the determined global median to all the UAVs in the network. As a result, the fault detection algorithm allows UAVs to detect and isolate faulty sensors quickly and to carry on target tracking in the semi-decentralized dynamic data fusion mode. As opposed to existing target tracking schemes, the proposed scheme is deterministic and guarantees the complete detection and isolation of faulty sensors on UAVs and thus successful target tracking. Numerical examples as well as experimental results are provided to support the developed ideas.