Dynamic Fault Diagnosis Research Articles

Application of fault propagation intensity in fault diagnosis of CNC machine tool

ABSTRACTTo realize dynamic and real-time multi-fault decoupling and diagnostic CNC machine tools, this study proposes a dynamic fault diagnosis method that is based on fault propagation intensity. Integrated fault mechanism analysis, directed graph theory, and interpretative structure model are used to construct a fault propagation hierarchical model to visually depict complex fault causality. The influence degree of component nodes and the fault influence degree of edges are calculated using PageRank and a coupling degree function. The fault propagation probability of component nodes is determined by synthesizing node fault probability. Fault propagation intensity is defined by the probability of fault propagation and edge-betweenness to characterize the behavior of fault propagation dynamically. Combined with the hierarchical fault propagation model, the critical path and node are determined. A certain type of CNC machine tool is taken as an example to carry out a specific application. Results show that the hierarchical model of system fault propagation realizes multi-fault decoupling and clarifies the process of fault propagation. The critical path is identified according to the fault propagation intensity, the deviation caused by describing the behavior of fault propagation based on a single index is avoided, and the accuracy of fault diagnosis is improved.

Research on time-varying stiffness of bearing based on local defect and varying compliance coupling

The stiffness characteristics of bearings are influenced by the variation in their compliance and the contour characteristics of local defects. Local profile functions including the depth and varying circumferential extent of local defect are introduced into the analysis model based on the bearing quasi-static model. The bearing quasi-static analysis model with local defect is established, and the correctness of the analysis model is verified. Firstly, the circumferential extents of bearing raceway defects and the influence of axial preload on the number of loaded balls, load distribution and stiffness of static bearing are investigated. Secondly, the influence of varying compliance on the time-varying stiffness of defect-free ball bearing at different rotational speeds is explored. And finally, the time-varying stiffness of bearing with coupling of local defects and varying compliance is examined. Results show that the load distribution and the time-varying stiffness in the coupling state are not the simple superposition of the two states acting separately; they will interact and influence each other, and jointly change the supporting characteristics of bearing. This study can serve as a basis for the nonlinear dynamic analysis, reliability analysis and fault diagnosis of bearing–rotor systems and the design of bearing parameters.

Real-time combustion torque estimation and dynamic misfire fault diagnosis in gasoline engine

Abstract In this research, an innovative state observer of gasoline engine based on the combination of Luenberger and sliding mode technique is proposed. This state observer is designed to track crankshaft angular speed and estimate engine combustion torque based on the experimental crankshaft angular speed of a four-cylinder Spark Ignition (SI) engine. Then, a new advance in the application of Artificial Neural Networks (ANNs) based on the estimated results of automated dynamic misfire fault diagnosis both under steady state and non-stationary condition is discussed in detailed. In order to effectively obtain data for network training, the estimated engine combustion torque is segmentally preprocessed according to the crank angle displacement of automobile engine. Furthermore, a series of experiments are carried out under normal and a variety of misfire conditions. The ANN systems are trained and tested using prepared cases. Finally, the Back-Propagation Neural Network (BPNN), Elman Neural Network (ENN), and Support Vector Machine (SVM) are applied to diagnose misfire fault, the effectiveness of each is evaluated respectively. Based on the estimated engine combustion torque, the experimental results show that the designed ENN is able to correctly diagnose misfire fault with a running time of 0.6 s, including single misfire, intermittent double-cylinder misfire, and continuous double-cylinder misfire in transient working condition.

Effects of gear eccentricity on time-varying mesh stiffness and dynamic behavior of a two-stage gear system

Gear eccentricity, one of the most common defects of gear systems, affects not only dynamic behavior but also mesh stiffness. Accurate mesh stiffness under defect conditions is vital for the dynamic simulation and fault diagnosis of gear systems. Hence, the timevarying mesh stiffness affected by gear eccentricity, which is commonly neglected in most studies, is incorporated in the proposed twostage spur gear dynamic model and calculated by a novel gear mesh model with an improved potential energy method. The effects of mesh stiffness with and without gear eccentricity on the gear system are compared on the basis of the constructed dynamic model. The influences of gear eccentricity on dynamic transverse and torsional responses, as well as dynamic transmission errors, are also quantitatively studied. Some helpful analytical results are then presented.

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.

Dynamic characteristic analysis and fault diagnosis of load‐sensitive drive system for loaders

A certain type of crawler loader often appears flameout in unilateral steering, which makes the performance of the whole machine greatly affected. By analysing the dynamic characteristics of the load-sensitive drive system, it is found that the hydraulic fluctuation and the power matching between the hydraulic pump and the engine are the fundamental reasons for the failure. In order to solve this problem, a simple and easy improvement scheme is designed on the premise of not affecting the engine efficiency: a suitable size damping is connected in the middle of the two load feedback lines of the drive system to solve the power matching between the engine and the hydraulic pump and reduce the hydraulic fluctuation of the system. Simulation and experimental results show that this method can effectively solve the problem of stalling in the steering process of loaders and greatly improve the dynamic performance of the system.

Nonlinear dynamic behavior research on high-speed turbo-expander refrigerator rotor

Abstract Low-frequency coupled vibrations significantly affect the stability and safety of the high-speed turbo-machinery. This study focused on the nonlinear dynamic behaviors of high-speed turbo-expander refrigerator supported on aerostatic bearings. The critical speed characteristics changing with bearing gas supply pressure are given by numerical calculation and modal experiment methods. The results show that the restrained modal tests are closer to the results of operational tests, which provides the basic data for the dynamic analysis and fault diagnosis of the bearing-rotor system. The coupled vibration phenomenon are presented and discussed in the paper, and the adjustment measures for eliminating the coupled vibration are used to improve the stability of gas bearing-rotor system.

Induction Motors Dynamic Eccentricity Fault Diagnosis Based on the Combined Use of WPD and EMD-Simulation Study

The frequency-domain analysis using the fast Fourier transform (FFT) for diagnosis of eccentricity fault has been widely used in squirrel-cage induction motor (IM). However, with the restriction of sampling frequency and time acquisition, FFT analysis could not provide ideal results under low levels of dynamic eccentricity (DE). In this paper, a combined use of the wavelet packet decomposition (WPD) and empirical mode decomposition (EMD) method is presented to diagnose the IM fault under low degrees of purely DE. The proposed method is based on the decomposition of apparent power signal and extracts the characteristic component. The fault severity factor (FSF) has been defined to evaluate the eccentricity severity. Simulation results using the finite element method (FEM) are tested to verify the effectiveness of the presented method under different load conditions.

Real Time Fault Diagnosis with Tests of Uncertain Quality for Multimode Systems and its Application in a Satellite Power System

Dynamic fault diagnosis must consider complex fault situations such as fault evolution, coupling, unreliable tests and so on. Previous dynamic fault diagnostic models and inference algorithms are mainly designed for the steady state systems, which are not suitable for the multimode systems. In this paper, a time varying dynamic model to solve the multimode fault diagnosis problem is proposed. Its structure and formulation are presented. Fault diagnosis based on this model is realized by means of inference calculation given the test result, which is formulated as an optimization problem. A new algorithm to solve this problem is proposed. Simulation experiments on different scenarios are carried out to validate the model and the algorithm. As an example, the case of a satellite electrical power system is studied in detail. Both the simulation result and the application result show that the method proposed in this paper can be used to solve the dynamic fault diagnosis problem for multimode systems considering the complex circumstances such as uncertain tests and system delay.

RETRACTED: Dynamic process fault diagnosis using improved Fisher discriminant analysis – An approach towards IoT

RETRACTED: Dynamic process fault diagnosis using improved Fisher discriminant analysis – An approach towards IoT

Cubic causality modeling and uncertain inference method for dynamic fault diagnosis

Cubic causality modeling and uncertain inference method for dynamic fault diagnosis

Condition Based Railway Pantograph Dynamic Behavior Measurement and Fault Diagnosis

Railway pantographs are used for collecting power from the overhead line system. Faults or failures of a pantograph will degrade the interaction between the pantograph and the overhead line, reducing the reliability of railway operation. The current pantograph maintenance techniques are highly dependent on visual and manual checks, which are not efficient. To improve pantograph maintenance efficiency and reduce the maintenance budget, a laboratory based pantograph test rig is developed to measure pantograph dynamic behaviors through different tests and develop a quick pantograph fault diagnosis method. Based on the test rig and the measurement result, pantograph 3D multibody and lumped mass models are established. Three tests are carried out: hysteresis measurement using different excitation speeds, frequency response function with various excitation levels, and a novel changing gradient test. The measurement data collected from a new pantograph is considered as the reference. Differences between the dynamic behaviors measured on a test pantograph and the reference data indicate faults or component degradation. The results indicate that the hysteresis test is effective in detecting pneumatic actuator and elbow joint issues; the frequency response function can monitor all faults, but it is time consuming; the changing gradient test is a quick way to identify and locate the fault.

Optimization of a Dynamic Fault Diagnosis Model Based on Machine Learning

A delay dynamic coupled fault diagnosis (DDCFD) model is a kind of effective fault diagnosis model based on probability graph for complex systems. Previous research assumes that its structure and parameters are all known before application. Since the initial model is always constructed by experts, it is possible that it is inconsistent with practical systems, which may impact the diagnostic accuracy. In order to improve the model by applying the practical data, this paper develops a machine learning method for the DDCFD model. A parameter learning method and three structure learning methods are studied. Using this approach, prior knowledge and test data can be combined together to get a better diagnostic model. The methods are tested and compared on simulated examples and an application case. The result indicates that diagnostic accuracy can be improved by means of parameter learning compared to the initial model, and it can be further improved by structure learning. The score-based algorithm is the best choice for learning the structure of the DDCFD model.

Predictive maintenance strategy of running fault based on ELM algorithm for power transformer

Transformer is the most important core equipment of power system, and once the fault happened, the economic losses and adverse social impacts resulted in faults by the twinkling of eye are difficult to estimate. In order to prevent transformer fault happened, the paper explored a sort of predictive maintenance strategy of transformer fault power transformer supported by the extreme learning machine (ELM) algorithm. In this paper, it made the anatomy of the drawbacks in breakdown maintenance (BM) and preventive maintenance (PM) maintenance system, pointed out the advantage of predictive reliability maintenance (PRM) maintenance system, studied on prediction of the fault diagnosis and predictive algorithm based on ELM, discussed the dynamic fault diagnosis model of power transformer in smart grid environment. A comparative study of different fault pattern prediction algorithms confirmed the rationality and feasibility of the fault prediction algorithm based on ELM. Transformer operation experience shows that the ELM algorithm can provide powerful technical support for the maintenance strategy of transformer fault prediction.

The NARX Model-Based System Identification on Nonlinear, Rotor-Bearing Systems

In practice, it is usually difficult to obtain the physical model of nonlinear, rotor-bearing systems due to uncertain nonlinearities. In order to solve this issue to conduct the analysis and design of nonlinear, rotor-bearing systems, in this study, a data driven NARX (Nonlinear Auto-Regressive with exogenous inputs) model is identified. Due to the lack of the random input signal which is required in the identification of a system′s NARX model, for nonlinear, rotor-bearing systems, a new multi-harmonic input based model identification approach is introduced. Moreover, the identification results of NARX models on the nonlinear, rotor-bearing systems are validated under different conditions (such as: low speed, critical speed, and over critical speed), illustrating the applicability of the proposed approach. Finally, an experimental test is conducted to identify the NARX model of the nonlinear rotor test rig, showing that the NARX model can be used to reproduce the characteristics of the underlying system accurately, which provides a reliable model for dynamic analysis, control, and fault diagnosis of the nonlinear, rotor-bearing system.

Research on fault detection method for heat pump air conditioning system under cold weather

Abstract Building energy consumption accounts for nearly 40% of global energy consumption, HVAC (Heating, Ventilating, and Air Conditioning) systems are the major building energy consumers, and as one type of HVAC systems, the heat pump air conditioning system, which is more energy-efficient compared to the traditional air conditioning system, is being more widely used to save energy. However, in northern China, extreme climatic conditions increase the cooling and heating load of the heat pump air conditioning system and accelerate the aging of the equipment, and the sensor may detect drifted parameters owing to climate change. This non-linear drifted parameter increases the false alarm rate of the fault detection and the need for unnecessary troubleshooting. In order to overcome the impact of the device aging and the drifted parameter, a Kalman filter and SPC (statistical process control) fault detection method are introduced in this paper. In this method, the model parameter and its standard variance can be estimated by Kalman filter based on the gray model and the real-time data of the air conditioning system. Further, by using SPC to construct the dynamic control limits, false alarm rate is reduced. And this paper mainly focuses on the cold machine failure in the component failure and its soft fault detection. This approach has been tested on a simulation model of the “Sino-German Energy Conservation Demonstration Center” building heat pump air-conditioning system in Shenyang, China, and the results show that the Kalman filter and SPC fault detection method is simple and highly efficient with a low false alarm rate, and it can deal with the difficulties caused by the extreme environment and the non-linear influence of the parameters, and what's more, it provides a good foundation for dynamic fault diagnosis and fault prediction analysis.

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.

Dynamic Eccentricity Fault Diagnosis in Induction Motors Using Finite Element Method and Experimental Tests

This paper presents the dynamic eccentricity (DE) in the squirrel cage induction machine obtained by the simulation using finite element method (FEM) and the experimental tests. Motor current signature analysis (MCSA) through the power spectral density (PSD) is used to monitor the low frequency components related to the rotor faults. In order to generalise the study, a comparison was performed between faulty machines and healthy, under different load conditions. The experimental tests show the encountered difficulties in the detection of the dynamic eccentricity under various load conditions because the spectral representation shows the existence of low frequency components for the healthy case, which are superimposed with the indicator components. These low frequency components offer an alternative way to the monitoring of eccentricity air-gap when the load is balanced.

Dynamic eccentricity fault diagnosis in induction motors using finite element method and experimental tests

Dynamic eccentricity fault diagnosis in induction motors using finite element method and experimental tests

Fault diagnosis for a class of active suspension systems with dynamic actuators’ faults

In this paper, a new fault diagnosis and fault-tolerant control method for a class of active suspension systems with actuator faults is proposed. The considered actuators have uncertain dynamic characteristics, which are the electromagnetic actuators made up with a motor control system and a ball screw transmission mechanism. To detect such suspension system actuator faults, dynamic fault diagnosis observers are designed for the actuators to estimate the possible faults. The actuators are analyzed to first and second order dynamic models, respectively, whose output can be measured but the rate is non-measurable. Then, the fault diagnosis method is developed for these two kinds of models to obtain the fault information. Using the fault estimation and adaptive control technique, a robust fault-tolerant controller is constructed to guarantee the performance of the rail vehicles in the faulty case. Finally, using the parameters of a practical suspension system, a simulation study is conducted to show the effectiveness of the proposed method.