Search

Two performance parameters are particularly important for the assessment of structural health monitoring (SHM) systems, i.e., their damage detection capabilities and risk of false positive indications due to varying environmental and operational conditions (EOCs). A reduced ratio of false-positive indications can be of significant importance for particular applications, for example, in aerospace, where the costs of unplanned maintenance procedures can be very high. In such cases, the reduction of the false calls ratio can be critical for the possibility of the practical application of the system, apart from damage detection efficiency and system costs. Among various sensor technologies, PZT networks are proven to be one of the most universal approaches to SHM, and they were successfully applied in different scenarios. Moreover, many EOCs which may have an impact on the risk of false positive indications have been identified. Over the years, different approaches to the influence of EOCs compensation have been proposed. Compensation methods can be tailored to the particular way in which a given measurement condition, for example, ambient temperature, alters signals acquired by the PZT network or can be formulated to be also applied in the more general case. In the paper, a method for enhancement of damage detection efficiency under influence of EOCs of general nature is proposed. The particular measurement condition affecting signals acquired by PZT sensors neither needs to be measured, which could be hard in some cases, but also nor even have to be identified. The efficiency of the proposed compensation algorithms is verified based on the example of experimental results obtained under varying temperatures.

With the development of concepts of industry 4.0, condition monitoring techniques are changing. Large amounts of generated data require diagnostic procedures to be automated, which drives the need for new and better methods of autonomous interpretations of vibration condition monitoring data. However, if new methods are to be operational, they need to be verified under real industrial conditions and compared with well-established expert-based diagnostic techniques. This article introduces the novel algorithm of data preprocessing for the nearest-neighbor-based anomaly detection. This approach is validated on real industrial machinery in a series of case studies. The population of over-hung centrifugal fans, employed in the same industrial process, were monitored continuously according to the proposed methodology for an extended time period. Piezoceramic accelerometers were used to register time-domain vibration data. The data were processed to extract several signal features to serve as inputs to the anomaly detection algorithm. The novel solution is compared to the well-established condition monitoring approach. Presented data include not only the intact state of machinery but also a machine breakdown case and serious deterioration of the machine condition. The influence of maintenance work is also presented in the article. Authors show the data-driven approach to condition monitoring, which can be used as one of many predictive maintenance techniques.

The paper presents work related to nonlinear system parameters identification. The research is focused on systems with hysteretic stiffness characteristics. The identification procedure is developed with use of artificial neural networks. The presented method assumes two separate clusters of neural networks, which are supported by additional signal processing block. Such approach gives an advantage over the conventional identification methods due to its small restrictions. The validation process considers structural responses in time and frequency domains as well as the restoring force plane of the dynamic structure. First, verification of the identification method is performed on the numerical simulation of the system with hysteretic stiffness. Next, the identification of the real dynamic system with contact-related nonlinearity is carried out. The steel samples with contacting surfaces were used in the experiment. Electromagnetic shaker was used to excite the structure and enforce a relative shear motion between surfaces in contact. The system response was recorded using the Polytec laser vibrometer.

In the last few decades, there has been a significant increase in interest in developing, constructing, and using structural health monitoring (SHM) systems. The classic monitoring system should, by definition, have, in addition to the diagnostic module, a module responsible for monitoring loads. These loads can be measured with piezoelectric force sensors or indirectly with strain gauges such as resistance strain gauges or FBG sensors. However, this is not always feasible due to how the force is applied or because sensors cannot be mounted. Therefore, methods for identifying excitation forces based on response measurements are often used. This approach is usually cheaper and easier to implement from the measurement side. However, in this approach, it is necessary to use a network of response sensors, whose installation and wiring can cause technological difficulties and modify the results for slender constructions. Moreover, many load identification methods require the use of multiple sensors to identify a single force history. Increasing the number of sensors recording responses improves the numerical conditioning of the method. The proposed article presents the use of contactless measurements carried out with the help of a high-speed camera to identify the forces exiting the object.

The capabilities of ceramic PZT transducers, allowing for elastic wave excitation in a broad frequency spectrum, made them particularly suitable for the Structural Health Monitoring field. In this paper, the approach to detecting impact damage in composite structures based on harmonic excitation of PZT sensor in the so-called pitch–catch PZT network setup is studied. In particular, the repeatability of damage indication for similar configuration of two independent PZT networks is analyzed, and the possibility of damage indication for different localization of sensing paths between pairs of PZT sensors with respect to damage locations is investigated. The approach allowed for differentiation between paths sensitive to the transmission mode of elastic wave interaction and sensitive reflection mode. In addition, a new universal Bayesian approach to SHM data classification is provided in the paper. The defined Bayesian classifier is based on asymptotic properties of Maximum Likelihood estimators and Principal Component Analysis for orthogonal data transformation. Properties of the defined algorithm are compared to the standard nearest-neighbor classifier based on the acquired experimental data. It was shown in the paper that the proposed approach is characterized by lower false-positive indications in comparison with the nearest-neighbor algorithm.

The paper introduces the concept of Instantaneous Circular Pitch Cycle Map (ICPCM) as a tool for analysis and diagnosis of modulation signatures in vibrations generated by Epicyclic Gearbox (EGs). The method is based on the representation of a vibration signal as a 3-D map illustrating selected signal properties. Within the method, a novel concept of a Resampling Relative Order (RRO) is introduced. The paper illustrates how the ICPCM might serve as an auxiliary tool for proper interpretation of spectra, envelope spectra, and time-synchronous averaged representations. The idea of the method is presented on a simulated signal, while its practical performance is illustrated on a real signal from a multi-stage EG.

Condition monitoring is an indispensable element related to the operation of rotating machinery. In this article, the monitoring system for the parallel gearbox was proposed. The novelty detection approach is used to develop the condition assessment support system, which requires data collection for a healthy structure. The measured signals were processed to extract quantitative indicators sensitive to the type of damage occurring in this type of structure. The indicator’s values were used for the development of four different novelty detection algorithms. Presented novelty detection models operate on three principles: feature space distance, probability distribution, and input reconstruction. One of the distance-based models is adaptive, adjusting to new data flowing in the form of a stream. The authors test the developed algorithms on experimental and simulation data with a similar distribution, using the training set consisting mainly of samples generated by the simulator. Presented in the article results demonstrate the effectiveness of the trained models on both data sets.

In this article, the recommendation system for processing signals was presented. It contains a database and two fuzzy modules composed within the system. Based on the contextual knowledge provided by the user, collected database, and fuzzy rules, the system suggests processing methods and features. The article presents an evaluation of the proposed system on a two-stage gearbox dataset. The system results are a list of recommended processing methods and extracted features, which allow for a more accurate data classification.

Data scarcity is a significant problem when it comes to designing machine learning systems for structural health monitoring applications, especially those based around data-hungry algorithms and methods, such as deep learning. Synthetic data generation could potentially alleviate this problem, lowering the number of measurements that need to be acquired in slow and often expensive conventional experiments. Such synthesis can be done by Generative Adversarial Networks, potentially creating unlimited synthetic samples recreating the original data distribution. While most of the research about these networks is centered around using them on image data, they have also been applied to audio waves - going as far as successfully synthesizing human speech. This suggests that these networks should also apply to synthesizing time-domain signals in various fields of structural health monitoring, guided waves in particular, as they are in many ways similar to audio wave signals. This work proposes an adaptation of style-based GAN architecture to time-domain signal generation, and presents its viability for guided waves synthesis, utilizing a database of signals collected in series of pitch-catch experiments on a composite plate.

Interpretation of the data acquired from guided-wave-based measurements often utilizes machine learning. However, creating effective machine learning models generally requires a significant amount of data - which in the case of guided waves are costly and time-consuming to acquire. This limitation significantly reduces the application perspective of many advanced machine learning algorithms, most notably deep learning. The problem of data scarcity has been partially addressed in the field of computer vision via the usage of generative adversarial neural networks. These generate synthetic data samples, matching the real data distribution. Aside from images, generative adversarial networks have also been applied to synthesize audio data - with recent advances going as far as successfully synthesizing human speech. These developments suggest that they may be applicable for generating guided waves data - as fundamentally the problem is in many ways similar to that presented by audio waves. This work explores the capabilities of generative adversarial neural networks in the area of guided-wave signal synthesis. The used database was acquired in a series of pitch-catch experiments in which various sensor locations were utilized, and is significantly extended both in terms of sensor locations and data available from each sensor pair. Lastly, the resultant synthesized data is evaluated by qualitative signal comparison.

The development of a machine’s condition monitoring system is often a challenging task. This process requires the collection of a sufficiently large dataset on signals from machine operation, context information related to the operation conditions, and the diagnosis experience. The two referred problems are today relatively easy to solve. The hardest to describe is the diagnosis experience because it is based on imprecise and non-numerical information. However, it is essential to process acquired data to develop a robust monitoring system. This article presents a framework for a system dedicated to recommending processing algorithms for condition monitoring. It includes a database and fuzzy-logic-based modules composed within the system. Based on the contextual knowledge provided by the user, the procedure suggests processing algorithms. This paper presents the evaluation of the proposed agent on two different parallel gearboxes. The results of the system are processing algorithms with assigned model types. The obtained results show that the algorithms recommended by the system achieve a higher accuracy than those selected arbitrarily. The results obtained allow for an average of 5 to 14.5% higher accuracy.

Many engineering structures operate under non-stationary conditions resulting in signals containing time-varying frequency components. Analysis of such signals in time or frequency domain naturally yields smeared results. Typically implemented solution to this problem is the angle and order domain analysis. However, in some scenarios, acquisition of speed signal may be not available, due to practical limitations. If the speed fluctuations are limited, or recorded vibrational signal is free from environmental and measurement disturbances, it is possible to extract the speed signal from raw time waveforms. State-of-the-art methods include e.g. semi-automatic tracking of the ridges of the Short-Time Fourier Transform for speed signal reconstruction. However, these methods can fail if the time-varying frequency components cross with each other. Moreover, it is very common that complicated structures operating in non-stationary conditions are controlled by sophisticated electronic systems, which can introduce constant frequency noise to the measuring equipment. This can lead to some interference of these time-invariant and time-variant frequency components. The paper presents a simple and easy way to use a tool for hand selection and tracking of desired components, thus making complex time-frequency filtration easy. Finally, instantaneous frequency is calculated with use of Centre Moments of Frequency to extract the speed signal. The presented method shows how a particular tool developed within 3-dimensional image processing branch could be implemented in analysis of 2-dimensional signals characterized by time-frequency varying components.

Reinforcement learning provides an edge in many problems of mechanical engineering. That is because its ability to solve problems on-the-fly based on interactions with initially unknown environment. In this paper the reinforcement learning approach to identification, adaptive control and structural change detection is presented. The approach consists of training feedforward artificial neural networks in a task of adaptive control of an unknown and timevariable system. The nets are used to build system model and propose optimal control prescription to actively suppress vibration of a selected system element. The difference between expected and obtained quality of outcome is used as a damage indicator - in a novelty-detection-based framework. A series of numerical experiments for a 5DOF system is used to evaluate method’s performance. It is shown that the method is able to operate for both impulse and white noise external excitation and can work despite random and significant changes of system parameters.

Active vibration suppression is a well explored area when it comes to simple problems, however as the problem complexity grows to a time variant system, the amount of researched solutions drops by a large margin, which is further increased with the added requirement of very limited knowledge about the controlled system. These conditions make the problem significantly more complicated, often rendering classic approaches suboptimal or unusable, requiring a more intelligent approach - such as utilizing soft computing. This work proposes a Artificial Neural Network (ANN) Model Predictive Control (MPC) scheme, inspired by horizon techniques which are used for MPC. The proposed approach aims to solve the problem of active vibration control of an unknown and largely unobservable time variant system, while attempting to keep the controller fast by introducing several methods of reducing the amount of calculations inside the control loop - which with proper tuning have no negative impact on the controller’s performance. The proposed approach outperforms the multi-input Proportional-Derivative (PD) controller preoptimized using a genetic algorithm.

The article presents a novel approach to vibration suppression that can be used in cases when system under consideration is unknown. The approach is based on a reinforcement-learning of artificial neural network fed with positions and velocities of the system components. The learning routine is designed on the basis of a 1+1 evolutionary scheme with concept of persistence of efficient steps. The performance of the four variants of the algorithm is evaluated on the basis of a simulated 3-DOF system.

One of the important issues related to the applications of Lamb waves for structural health monitoring is their undesired sensitivity to variation of environmental conditions. Temperature is the main factor that can affect wave propagation and hence significantly reduce performance of a SHM system. Therefore, there is a need for development of robust monitoring methods with low sensitivity to temperature variations. This paper is aimed at verification of efficiency of four methods designed for damage detection using Lamb wave measurements performed in variable environmental conditions. The methods investigated in the comparison are the following: optimal baseline selection approach, the damage index based on a signal alignment with respect to instantaneous phase, and a group measurement approach capable of distinguishing local damage-related changes from temperature-induced global ones. The fourth method relies on fusion all these solutions simultaneously. The methods’ ability to damage detection is compared using a specimen that is subjected to large temperature changes. It is found that although all the methods have their strengths and weaknesses, a cooperation of all solutions allows for significant increase of the damage detection efficiency.

Neural networks are commonly recognized tools for the classification of multidimensional data obtained in structural health monitoring (SHM) systems. Their configuration for a given scenario is, however, a challenging task, which limits the possibilities of their practical applications. In this article the authors propose using the neural network ensemble approach for the classification of SHM data generated by guided wave sensor networks. The overproduce and choose strategy is used for designing ensembles containing different types and sizes of neural networks. The proposed method allows for a significant increase of the state assessment reliability, which is illustrated by the results obtained from the practical industrial case of a full-scale aircraft test. The method is verified in the process of detecting fatigue cracks propagating in the aircraft load-carrying structure. The long-term experiments are performed in variable environmental conditions with a net of structure-embedded piezoelectric sensors.

In the paper, optical-based measurement methods for calculating the deflection and vibration of overhead lines are presented. The authors describe the state of the art in the field of non-contact examination of static and dynamic overhead transmission line characteristics, and propose concepts of vision-based measurement systems for both static and dynamic states of a structure. The system devoted to static measurements is based on a digital SLR camera and image-processing software used for the acquisition and interpretation of data. The digital image correlation method, implemented in Wiz2D software, is applied to compute the displacement of the transmission line cable with respect to a known baseline position. Dynamic characteristics of the structure are obtained using a stereo-vision system consisting of two high-speed digital cameras. Corresponding points in two video sequences of the vibrating cable are tracked using TEMA software. 3D positions in a camera frame of reference and displacements are computed using a 3D reconstruction method. The paper describes two series of experiments conducted on a lab setup and the obtained results are examined and discussed.

The paper presents up-to-date multidisciplinary review of scientific knowledge and industrial guidelines concerning the issue of technical assessment of epicyclic gearboxes on the basis of vibrational measurement. The paper covers norms, patents, signal processing methods, technical features, diagnostic methodologies, and high-level soft computing. The review classifies recognized papers according to different aspects, like case difficulty, applied methods or studied objects. Moreover, the paper discusses recent development of diagnostic techniques taking into account changing operational conditions and requiring powerful data processing units not available hitherto. Since in recent years many partial review papers have been published in different journals, the current paper presents their individual scope. Because diagnostics of epicyclic gearboxes generally covers and extends the classical signal processing techniques and classical machine condition monitoring functionality, the current paper might serve as a descriptive database of sources of information and bibliography or as a reference to individual techniques.

Summary Deflection curve can be used to detect and localize damage in civil engineering structures. In this paper, a vision-based method applied for in-plane displacement field measurement of cantilever beams is presented. The deflection curve of the analyzed structure is computed by means of the digital image correlation. Damage is introduced into the structure. Resulting deflection curves are used as an input to the novel damage detection algorithms: line segments method and voting method. The algorithms are then compared with the second derivative method and assessed through the probability of detection procedure. Copyright © 2015 John Wiley & Sons, Ltd.

In this paper a vision-based method for measuring vertical deflection of large engineering structures under load is presented in example of crane working arm deflection measurements. The deflection is computed by means of a digital image correlation technique embedded in the vision-based system. The results of measurements performed on the vision system were compared with data obtained using a Leica Laser Tracker system. The comparison revealed proper functioning of the developed vision system in industrial conditions.

The authors provide a method based on ensembles of artificial neural networks (ANNs) that, fed with supervised training data, are able to estimate rotational speed of the machine under investigation. Ensembles of various ANNs trained using different algorithms provide reliability of the method and protection from unfortunate initial weights distribution. The method is validated on the basis of data gathered in large amount of experiments performed in epicyclic gearbox test bed

Typically, if order analysis of vibration signal is expected, a speed sensor (phase marker) or an encoder are installed on the shaft. However, in some practical scenarios, the speed information recorded in parallel to the vibration signal acquisition is not available; yet, it is still required. In this case, one is forced to use a raw vibration signal to extract the information about so-called instantaneous phase or instantaneous frequency of a selected component, and—if required—scale to a selected shaft. In recent years, few different techniques for speed recovery have been proposed, each one with different assumptions and each implementing more or less complexed mathematical apparatus. The current paper proposes a guidance how to select a suitable method on the basis of the visual deduction about signal characteristics with the implication on selection of the easiest and most automatized method sufficient for analysed case.

SUMMARY This paper presents a novel approach to the problem of health monitoring of aircraft structures using Lamb waves. Piezoelectric sensors, embedded in the aircraft sheathing, generate Lamb waves with the aim to monitor the structural integrity of complex structure parts. The ultrasonic signals obtained from the sensor pairs arranged in pitch-catch configuration are used for the calculation of a number of different damage indices. The damage indices are then used as inputs for a classifier employing an artificial neural network (ANN) that is trained to perform structure condition assessment. Efficiency of the ANN classifier trained on artificial data generated from the numerical simulations performed using linear interaction simulation approach is investigated. The resulting classification results are compared with those obtained for the ANN trained on experimental data from the real specimens. Copyright © 2014 John Wiley & Sons, Ltd.

Lamb waves (LW) are used for damage detection and health monitoring due to the long range propagation ability and sensitivity to structural integrity changes as well as their robustness in different applications. However, due to the dispersive character and multimode nature of LWs, analysis of the acquired ultrasonic signals is very complex. It becomes even more difficult when applied to a complex structure, for instance, an aircraft component with riveted joints and stringers characterized by difficult geometries. Therefore, numerous approaches to the evaluation of damage indices have been proposed in the literature. In this paper, comparison a number of damage indices applied to LWs testing in aircraft aluminum panels. The damage indices, known from the literature have been selected from the application point of view. Artificial neural network has been used for the classification of fatigue cracks and artificial damages induced in the specimens taken from a real aircraft structure. Article presents results of simulation, data analysis and data classification obtained using selected and dedicated neural network. The main aim of the presented research was to develop signal processing and signal classification methods for an aircraft health monitoring system. The article presents a part of the research carried out in the project named SYMOST.

This paper presents an approach to damage growth monitoring and early damage detection in the structure of PZL - 130 ORLIK TC II turbo-prop military trainer aft using the statistical models elaborated by the Polish Air Force Institute of Technology (AFIT) and the network of the sensors attached to the structure. Drawing on the previous experiences of the AFIT and AGH in structural health monitoring, the present research will deploy an array of the PZT sensors in the structure of the PZL -130 Orlik TC II aircraft. The aircraft has just started Full Scale Fatigue Test (FSFT) that will continue up to 2013. The FSFT of the structure is necessary as a consequence of the structure modification and the change of the maintenance system - the transition to Condition Based Maintenance. In this paper, a novel approach to the monitoring of the aircraft hot-spots will be presented. Special attention will be paid to the preliminary results of the statistical models that provide an automated tool to infer about the presence of damage and its size. In particular, the effectiveness of the selected signal characteristics will be assessed using dimensional reduction methods (PCA) and the so-called averaged damage indices will be delivered. Moreover, the results of the signal classification based on the neural network will be presented alongside the numerical model of the wave propagation. The work contains selected information about the project scope and the results achieved at the preliminary stage of the project