Incremental Fault Research Articles

A relationship-aware calibrated prototypical network for fault incremental diagnosis of electric motors without reserved samples

Recently, incremental learning (IL) has been widely used in intelligent fault diagnosis of electronic machinery. Most of the typical IL methods have adopted the exemplar-replay strategy to retain the learned diagnostic knowledge. However, it is almost impossible to have infinite storage space to retain fault samples in practical industrial scenarios, which brings a significant challenge for actual industrial applications. To solve this issue, a novel Relationship-Aware Calibrated Prototypical Network (RACPN) is proposed for incremental fault diagnosis of electric motors, which retains learned diagnostic knowledge without requiring the storage of any fault samples from previous sessions. Firstly, a fault prototype calibration (FPC) method is employed to learn new diagnostic knowledge from new sessions. Secondly, a task-relationship representation (TRR), which stands for a method to represent the relationships between tasks, is utilized to enhance the maintenance of diagnostic knowledge across different sessions. Finally, a Gaussian Bayes classifier with Mahalanobis metric is adopted to enhance the inference reliability for classifying fault categories. Experiments conducted on two electrical motor datasets demonstrate the superiority and effectiveness of the proposed RACPN. The results validate that current signals as model input can achieve satisfactory diagnostic performance. The proposed RACPN is a promising tool for incremental fault diagnosis in electric motors.

Semi-supervised prototype network based on compact-uniform-sparse representation for rotating machinery few-shot class incremental fault diagnosis

Distinct from the ideal ‘closed-set’ assumption of comprehensive fault types and arbitrary data access, the entangled challenges of labeled instance paucity and fault increasing with operation are difficult to circumvent in industrial diagnostics. However, conventional deep learning-based diagnostic models lack the capacity for dynamic and incremental acceptance of new faults under limited labeled instances. When confronted with this more industrially realistic diagnostic task, their performance is constrained by overfitting, imbalance, and catastrophic forgetting, thus diminish the feasibility of engineering deployment. In this research, we define this task as a few-shot class incremental fault diagnosis (FS-CIFD) problem and develop a semi-supervised prototype network based on compact-uniform-sparse representation (CUS-SSPN) for this purpose. Structurally, CUS-SSPN incorporates the designed compact-uniform-sparse (CUS) composite loss, optimal transport (OT)-based semi-supervised strategy, and multi-level knowledge distillation (KD) strategy into the prototype network. With the CUS composite loss and OT-based semi-supervised strategy, CUS-SSPN can assign representative prototypes for each incremental fault and form an intra-class compact and inter-class uniform and sparse representation space. This allows the model to incrementally learn new faults while effectively suppressing the overfitting triggered by instance paucity and the biased diagnosis triggered by unbalanced labeling. Combining the designed multi-level KD strategy to convey category centroid information and classification decision knowledge from both prototype and logits levels, catastrophic forgetting of known faults can be effectively mitigated. Extensive FS-CIFD tasks are constructed on both publicly available dataset and practical wind turbine dataset to validate the effectiveness of CUS-SSPN and the feasibility of engineering diagnostics.

The Chaochou active fault in southern Taiwan: Characteristics and geomorphological significance as a reverse fault

The Chaochou Fault in southern Taiwan has long been recognized as an active fault, but its exact location is still uncertain. In this study, we focused on the middle part of the Chaochou Fault, where a flight of fluvial terraces was deformed and preserved. High-resolution aerial photos were first used to observe and map these terraces and their vicinity, followed by comprehensive field investigations including RTK-GPS surveys for high-resolution topographic profiles of the deformed terraces and OSL sample collections for obtaining the ages of the terraces. A series of active faults, consisting of Fault A (“FA”), Fault B (“FB”), and Fault Bb (“FBb”), are collectively named the Chaochou Active Fault Zone. The fault zone is recognized and characterized by range-facing scarps, frontal flexural scarps, and discontinuous slopes deformed by FA, FB, and FBb, respectively. The repeated activities of FA were recorded by the incremental fault scarp heights from young to old terraces. Considering the highest fault scarp height of 65.9 ± 6 m observed in T1 and its OSL age of 35.3 ± 4.3 ka, the long-term fault slip rate of FA is about 2 mm/yr.

A new feature boosting based continual learning method for bearing fault diagnosis with incremental fault types

Rotating machinery can unexpectedly generate many new fault types under changing operating conditions. The capability of fault diagnostic models to adapt and acquire knowledge regarding new fault types is increasingly vital. However, traditional deep learning-based fault diagnosis models often encounter the challenge of catastrophic forgetting when facing new fault types. Numerous attempts were trapped with the stability–plasticity dilemma when tackling this phenomenon. In this study, we draw inspiration from the gradient boosting algorithm and propose a feature boosting based continual learning method. This method allows the diagnostic model to continuously and adaptively acquire knowledge of new fault types. Initially, the concept of gradient boosting is employed to construct an initial fault diagnostic model. Then, new modules are continuously extended dynamically for the initial diagnostic model to fit the residuals between the actual label and the output of the initial diagnostic model. Finally, to maintain the backbone of the fault diagnostic model as a single one, redundant parameters and feature dimensions are removed using an effective distillation strategy. Experimental results demonstrate that the feature boosting based continual learning method effectively mitigates catastrophic forgetting and enhances the plasticity of the fault diagnosis model, outperforming other existing methods.

A Lifelong Learning Method Based on Generative Feature Replay for Bearing Diagnosis With Incremental Fault Types

Deep learning (DL)-based fault diagnosis models need to collect sufficient fault information for each fault type to ensure high-precision diagnosis. Some unexpected and new fault types will inevitably appear in actual conditions, which are called incremental fault types or class increment. Traditional DL models require the costly collection of all known data for retraining; while the use of new fault data may lead to catastrophic forgetting of old tasks. To solve the problem of bearing diagnosis with incremental fault types, a lifelong learning method based on generative feature replay (LLMGFR) is proposed in this study. A feature distillation method is put forward in this method to avoid forgetting in the feature extractor. The generator is trained to produce old task features. The generated features are mixed with real features of the current task to solve the imbalance problem and catastrophic forgetting of the classifier effectively. According to incremental fault diagnosis cases, LLMGFR can learn constantly and adaptively in dynamic environments with incremental fault types.

Feature Clustering-Based Network for Industrial Process Diagnosis With Incremental Fault Types

Feature Clustering-Based Network for Industrial Process Diagnosis With Incremental Fault Types

Incremental Fault Diagnosis Method Based on Metric Feature Distillation and Improved Sample Memory

Incremental learning-based fault diagnosis systems (IFD) are widely used because of their ability to handle constantly updated fault data and types. However, the catastrophic forgetting problem remains the most crucial contemporary challenge facing IFD. This paper proposes an incremental fault diagnosis method based on metric feature distillation (MFD) and improved sample memory to solve this problem. First, the metric feature distillation is designed with metric learning methods and feature distillation. It uses distillation and triplet loss to constrain the network parameters of old and new tasks in the same feasible region, effectively alleviating catastrophic forgetting. Then, for a small amount of data that can be stored scenario, an improved sample memory strategy is introduced to reduce catastrophic forgetting further, called the center and hard sample memory (CAHM). It can better preserve the global information of the data, reducing the forgetting of old data information that needs to be preserved during the training process. Experimental results on CWRU and MFPT datasets verify the proposed method’s effectiveness.

Broad auto-encoder for machinery intelligent fault diagnosis with incremental fault samples and fault modes

Intelligent fault diagnosis (IFD) has been a widely concerned topic in the field of prognostics and health management. Existing machinery IFD approaches are generally developed based on the one-time learning manner. Therefore, they are powerless to deal with the data stream issue in which new fault samples and fault modes will be progressively collected for model training. To overcome this drawback, this paper proposes a broad auto-encoder (BAE) with incremental learning capabilities for on-line IFD of machinery. The BAE is constructed by stacking a series of auto-encoders in the width direction. Then, the output weight matrix of the BAE is calculated by the ridge regression algorithm. After that, the capabilities of sample-incremental learning and class-incremental learning are developed, so that the BAE can easily update itself to accommodate the new fault samples and fault modes without model retraining. With the two incremental learning capabilities, the BAE can be first trained using limited historical fault samples, and then incrementally learn new diagnosis knowledge from the newly coming fault samples and fault modes. In this way, the BAE will be more and more powerful over time. Finally, the proposed BAE is applied to diagnose faults for high-speed train wheelset bearings and disc components. The results show that the proposed BAE offers an efficient solution for machinery IFD to deal with the continuous data stream issue.

A Lifelong Learning Method for Gearbox Diagnosis With Incremental Fault Types

Deep learning (DL)-based fault diagnosis models have to collect the most comprehensive data of mechanical fault types to ensure reliability. In real scenarios, due to complex, variable operating conditions, machines often generate unexpected faults that lead to an increment of fault types, causing the diagnosis model to be invalid. Therefore, the data of new fault types are needed to retrain the model. However, DL models suffer from catastrophic forgetting when incrementally learning new classes. To solve the problem of the diagnosis of increasing fault types, a lifelong learning method for fault diagnosis (LLMFD) is proposed in this paper under the lifelong learning paradigm. The key of LLMFD is a proposed dual-branch aggregation networks (DBANets) framework that is combined with reserved exemplars to learn the new fault types without forgetting the old ones. In DBANets, each residual block layer has a dynamic block and a steady block to solve the stability–plasticity dilemma in lifelong learning. The aggregation weights are adopted to balance stability and plasticity. LLMFD is applied to a diagnosis case of incremental fault types. Results verify that LLMFD is superior to other lifelong learning methods and has satisfactory robustness.

Model Evolution Mechanism for Incremental Fault Diagnosis

As the key issue of equipment operation and maintenance, fault diagnosis has attracted extensive attention, in which deep learning is often applied to the construction of fault diagnosis models. However, the changing operation state of industrial systems will lead to the decline of model performance. Therefore, it is necessary and urgent to explore the dynamic update of models. This work establishes an incremental fault diagnosis (IFD) framework, and further proposes a model evolution mechanism based on adaptive knowledge distillation and representative exemplar selection (MEMAR). Firstly, when new samples pour in, knowledge distillation (KD) is used to mine the potential correlation between old and new samples, with adaptive constraints on parameter optimization for model updating. Secondly, genetic algorithm (GA) is applied to selectively retain representative exemplar subsets, so as to enhance the model’s ability to deal with dynamic data. Moreover, in order to further verify the effectiveness of MEMAR, the discrimination performance of the fault diagnosis model after multi-stage incremental learning is tested. The experimental results show that the proposed MEMAR can reduce the storage space of old samples and training time. Besides, it enables the fault diagnosis model based on deep learning to have excellent diagnostic performance for new and old samples.

Online Monitoring of Transmission Line Operation Based on Power Internet of Things

Current monitoring methods of transmission line operation status can not obtain real-time data of distributed distribution transmission line, which leads to a large error in monitoring results. Therefore, a multi-state on-line monitoring method based on power Internet of Things is proposed. Using the gateway of power internet of things to set up network control access mode, build edge computing model, and using AD chip of ADS8365W5300 to obtain the real-time data of massive distributed distribution network, then make a decision on the fault after processing and analyzing the data. This paper constructs an edge computing model which can complete the data processing and analysis in the edge node, and designs the deployment of the edge computing model. By evaluating the faults in the dynamic incremental fault set, the risk state of transmission line in the danger control area is obtained, and a multi-state on-line monitoring method is designed. The experimental results show that the proposed method can monitor the transmission line running state accuratel

Relative Structural Complexity of Plate‐Boundary Fault Systems Controls Incremental Slip‐Rate Behavior of Major Strike‐Slip Faults

AbstractA comparison of published incremental fault slip rates from four major strike‐slip faults with the proximity and number of other active faults in the surrounding plate boundary systems shows that the behavior of the primary fault is correlated with the structural complexity of its tectonic setting. To do this, we characterize the relative structural complexity of the fault network surrounding a fault location of interest by defining the coefficient of complexity, which quantifies the density and displacement rates of the faults in the plate‐boundary network at specified radii around the site of interest. We show that the relative constancy of incremental slip rates measured along primary faults of the Alpine, San Andreas, North Anatolian, and Dead Sea fault systems reflects the proximity, number, and activity of their close neighbors. Specifically, faults that extend through more structurally complex plate boundary fault systems exhibit more irregular slip behavior than faults that pass through simpler settings. We suggest that these behaviors are likely a response to more complex stress interactions within more structurally complicated regional fault systems, as well as possible temporal changes in fault strength and/or kinematic interactions amongst mechanically complementary faults within a system that collectively accommodates overall relative plate motion. Our results provide a potential means for better evaluating the future behavior of large plate‐boundary faults in the absence of well‐documented incremental slip‐rate behavior, and may help improve the use of geological slip‐rate data in probabilistic seismic hazard assessments.

Incremental Fault Analysis: Relaxing the Fault Model of Differential Fault Attacks

This article presents a new fault analysis technique against cryptographic devices called the incremental fault analysis (IFA), which can be adapted into fault attacks using more traditional differential fault analysis (DFA) techniques in order to increase their feasibility under more practical fault injection conditions. Many previous attack methods require precise fault injection techniques such as clock glitching. By contrast, IFA is compatible with a more practical overclocking fault injection technique in which a cryptosystem is stressed at a constant level throughout the entire encryption, and this constant stress level is then increased between consecutive encryptions. It is observed that as new faults occur incrementally between increased stress levels, they often become superimposed upon faults first appearing at lower stress levels. IFA exploits these incremental fault differentials to deduce the cipher key more rapidly. Attacks were tested using practical fault injection methods on the advanced encryption standard (AES) both with and without IFA applied. Using IFA, allowed cipher keys to be retrieved with a success rate of 100% from 10 times less faulty ciphertexts and 6.4 times less computational time, requiring 16, 86, and 43 ciphertexts on average for AES-128, AES-192, and AES-256, respectively.

How Do Variably Striking Faults Reactivate During Rifting? Insights From Southern Malawi

AbstractCrustal extension is commonly thought to be accommodated by faults that strike orthogonal and obliquely to the regional trend of the minimum compressive stress (σ3). Activation of oblique faults can, however, be conceptually problematic as under Andersonian faulting, it requires preexisting crustal weaknesses, high fluid pressures, and/or stress rotations. Furthermore, measurements of incremental fault displacements, which are typically used to identify oblique faulting, do not necessarily reflect regional stresses. Here, we assess oblique faulting by calculating the stress ratio (σ3/σ1, where σ1 is the maximum compressive stress), slip tendency, and effective coefficient of friction (μs′) required to reactivate variably striking normal faults under different trends of σ3. We apply this analysis to NW and NNE striking active faults at the southern end of the Malawi Rift, where NE‐SW, ENE‐WSW, E‐W, and SE‐NW σ3 trends have previously been proposed. A uniform σ3 trend is inferred for this region as recent joints sets do not rotate along the rift. With a NE‐SW trending σ3, NW‐striking faults are well oriented, however, NNE‐striking faults require μs′ < 0.6 to reactivate. This is inconsistent with a lack of frictionally weak phyllosilicates detected in the fault zone rocks. With an ENE‐WSW to E‐W trending σ3, all faults can reactivate at μs′ > 0.55. These σ3 trends are also comparable to a focal mechanism stress inversion, regional joint orientations, and previously reported geodetically derived extension directions. We therefore conclude that unlike typical models of oblique rifting, the southern Malawi Rift consists of faults that all strike slightly oblique to σ3.

Temporal evolution of fault systems in the Upper Jurassic of the Central German Molasse Basin: case study Unterhaching

The structural evolution of faults in foreland basins is linked to a complex basin history ranging from extension to contraction and inversion tectonics. Faults in the Upper Jurassic of the German Molasse Basin, a Cenozoic Alpine foreland basin, play a significant role for geothermal exploration and are therefore imaged, interpreted and studied by 3D seismic reflection data. Beyond this applied aspect, the analysis of these seismic data help to better understand the temporal evolution of faults and respective stress fields. In 2009, a 27 km2 3D seismic reflection survey was conducted around the Unterhaching Gt 2 well, south of Munich. The main focus of this study is an in-depth analysis of a prominent v-shaped fault block structure located at the center of the 3D seismic survey. Two methods were used to study the periodic fault activity and its relative age of the detected faults: (1) horizon flattening and (2) analysis of incremental fault throws. Slip and dilation tendency analyses were conducted afterwards to determine the stresses resolved on the faults in the current stress field. Two possible kinematic models explain the structural evolution: One model assumes a left-lateral strike slip fault in a transpressional regime resulting in a positive flower structure. The other model incorporates crossing conjugate normal faults within a transtensional regime. The interpreted successive fault formation prefers the latter model. The episodic fault activity may enhance fault zone permeability hence reservoir productivity implying that the analysis of periodically active faults represents an important part in successfully targeting geothermal wells.

Influence of normal fault growth and linkage on the evolution of a rift basin: A case from the Gaoyou depression of the Subei Basin, eastern China

ABSTRACT It is widely acknowledged that the growth and linkage of a boundary fault affects the evolution of the accompanying sedimentary basin and exerts a strong influence on its sedimentary characteristics. In this study, we attempt to discuss the evolution of a boundary fault and rift basin by using seismic interpretation, incremental fault throw, and growth rate–distance (G–d) profile, which are more suitable parameters than the expansion index in rift basins where footwall strata are missing because of erosion or a depositional hiatus. Therefore, the Gaoyou depression is taken as an example. The Gaoyou depression is the largest hydrocarbon-bearing depression in the Subei Basin, which is a representative rift basin in eastern China. The depression is bounded to the south by the Jiangdu–Wubao fault zone, which includes the Zhen 1–Wu 1, Zhen 2, and Wu 2 faults. Using high-resolution, three-dimensional seismic data, the Zhen 1–Wu 1 fault can be subdivided into four segments based on its G–d profile: the Shaobo, Fanchuan, Liuwushe, and Liulu segments. Growth rates indicate that the Zhen 1–Wu 1 fault and the Zhen 2 and Wu 2 faults developed successively from the Late Cretaceous to the Miocene during rifting of the Gaoyou depression and that development of the Zhen 2 and Wu 2 faults was controlled by the Zhen 1–Wu 1 fault. The four segments of the Zhen 1–Wu 1 fault were linked with each other by the time the Zhen 2 and Wu 2 faults formed in the Late Paleocene. An evolution model of the Gaoyou depression is set up in this study. Integrating the fault activity with previous sedimentological studies, we suggest that the evolution of the Gaoyou depression was controlled by the Jiangdu–Wubao fault zone rather than by a single fault. Successive slip of the Zhen 1–Wu 1 fault and the Zhen 2 and Wu 2 faults caused subsidence and migration of the depocenters. This study presents a case investigation of the effects of boundary faults on rift basins in eastern China.

Revisiting Vulnerability Analysis in Modern Microprocessors

The notion of Architectural Vulnerability Factor (AVF) has been extensively used to evaluate various aspects of design robustness. While AVF has been a very popular way of assessing element resiliency, its calculation requires rigorous and extremely time-consuming experiments. Furthermore, recent radiation studies in 90 nm and 65 nm technology nodes demonstrate that up to 55 percent of Single Event Upsets (SEUs) result in Multiple Bit Upsets (MBUs), and thus the Single Bit Flip (SBF) model employed in computing AVF needs to be reassessed. In this paper, we present a method for calculating the vulnerability of modern microprocessors -using Statistical Fault Injection (SFI)- several orders of magnitude faster than traditional SFI techniques, while also using more realistic fault models which reflect the existence of MBUs. Our method partitions the design into various hierarchical levels and systematically performs incremental fault injections to generate vulnerability estimates. The presented method has been applied on an Intel microprocessor and an Alpha 21264 design, accelerating fault injection by 15×, on average, and reducing computational cost for investigating the effect of MBUs. Extensive experiments, focusing on the effect of MBUs in modern microprocessors, corroborate that the SBF model employed by current vulnerability estimation tools is not sufficient to accurately capture the increasing effect of MBUs in contemporary processes.

Incremental Fault Diagnosis for Nonlinear Processes

A new faults classification method based on on-line independent support vector machine (OISVM) is proposed for fault diagnosis in nonlinear processes. Fault diagnosis can be taken as a pattern recognition problem. As most processes are intrinsically nonlinear, support vector machines (SVMs) are one of the most popular and promising classification algorithms. The fatal drawbacks of standard SVM is the computing overhead grows with the number of training samples, where as training samples from real industrial processes are increasing with time grows. An incremental fault diagnosis approach based on OISVM is proposed in this work. Some related problem such as variable selection and parameter setting are also discussed in this work. Simulation results on the Tennessee Eastman process (TEP) demonstrate the effectiveness of the proposed method.

Active tectonics in southern Xinjiang, China: Analysis of terrace riser and normal fault scarp degradation along the Hotan‐Qira Fault System

The northern piedmont of the western Kunlun mountains (Xinjiang, China) is marked at its easternmost extremity, south of the Hotan‐Qira oases, by a set of normal faults trending N50E for nearly 70 km. Conspicuous on Landsat and SPOT images, these faults follow the southeastern border of a deep flexural basin and may be related to the subsidence of the Tarim platform loaded by the western Kunlun northward overthrust. The Hotan‐Qira normal fault system vertically offsets the piedmont slope by 70 m. Highest fault scarps reach 20 m and often display evidence for recent reactivations about 2 m high. Successive stream entrenchments in uplifted footwalls have formed inset terraces. We have leveled topographic profiles across fault scarps and transverse abandoned terrace risers. The state of degradation of each terrace edge has been characterized by a degradation coefficient τ, derived by comparison with analytical erosion models. Edges of highest abandoned terraces yield a degradation coefficient of 33 ± 4 m2. Profiles of cumulative fault scarps have been analyzed in a similar way using synthetic profiles generated with a simple incremental fault scarp model. The analysis shows that (1) rate of fault slip remained essentially constant since the aggradation of the piedmont surface and (2) the occurrence of inset terraces was synchronous at all studied sites, suggesting a climate‐driven terrace formation. Observation of glacial and periglacial geomorphic features along the northern front of the western Kunlun range indicates that the Qira glaciofluvial fan emplaced after the last glacial maximum, during the retreat of the Kunlun glaciers (12–22 ka). The age of the most developed inset terrace in uplifted valleys is inferred to be 10 ± 3 ka, coeval with humid climate pulses of the last deglaciation. The mass diffusivity constant (k=τ/T, being time B.P.) in the Hotan region is determined to be 3.3 ± 1.4 m2/103 years, consistent with other estimates in similar climatic and geologic environments of western China. These results imply a minimum rate for the Tarim subsidence of 3.5 ± 2 mm/yr. If Western Kunlun overthrusts the Tarim platform on a crustal ramp dipping 40°–45° to the south, it would absorb at least 4.5 ± 3 mm/yr of convergence between western Tibet and Tarim.

Dynamic Leakage of Faults During Differential Depletion: Theory, Models, and Examples from the Niger Delta

Previous studies of fault sealing have addressed possible fault leakage during secondary migration due to the effects of increased hydrocarbon-water capillary pressure, fracturing, or small-scale incremental fault movements. Of equal importance to production geologists is the failure and leakage of faults during field development due to differential depletion of adjacent fault blocks. This paper examines the unique problems associated with this dynamic leakage of faults. It is theoretically shown that the fault sealing mechanism, and the extent of the seal, directly influences the failure process which in turn results in a variety of favorable and unfavorable effects on field development. The qualitative models give considerable insight into such aspects as oil-column expansion and resaturation losses, interfault block aquifer support (with important implications to material balance calculations), possible leakage or spillage of oil across faults, and potential fault failure during (re)injection projects. Examples of dynamic fault leakage are presented from selected fields of the Niger delta.