Machine Condition Monitoring Research Articles

A New Stator Resistance Estimation Technique for Vector-Controlled PMSM Drive

This article presents new stator resistance (R <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">s</sub> ) estimation techniques for a vector-controlled permanent magnet synchronous motor (PMSM) drive. These techniques give the indirect measurement of the temperature rise in the stator. An accurate estimation of the R <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">s</sub> will help in the condition monitoring of the PMSM machine. The various R <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">s</sub> estimation techniques proposed in this article are formulated using the model reference adaptive system (MRAS). Out of the various methods proposed, Þ - MRAS performs better than the remaining stator resistance estimation techniques since this is independent of the integrator, differentiator, and speed; offers robust and dynamic performance in all four modes of operation, including zero speed. The stability and sensitivity analysis is further carried out to ensure the robustness of the proposed estimation techniques. The proposed stator resistance estimation techniques are simulated in MATLAB/Simulink and experimentally validated through a dSPACE-1104-based laboratory prototype.

Deep multi-sensorial data analysis for production monitoring in hard metal industry

The industry practice of machining hard metal parts using CNC lathe turning machines is through grinding and milling procedures. The typical practice for quality control is through manual inspection, as automated solutions are difficult to integrate in production and do not reach the same level of accuracy. In this scope, the proposed system aims to automate the manufacturing process for the machine condition monitoring and 3D inspection of defective hard metal parts, by utilizing deep neural networks (DNNs) and investigating the defects on real production samples. Concretely, data are collected with (a) shop floor sensors, (b) high-resolution laser microprofilometer and (c) ultrasound scanner. The proposed system analyzes the collected data through AI models for quality control. Moreover, a fusion scheme is proposed to further improve accuracy. The system is validated on the classification of defective and non-defective samples, using metrics including accuracy, F-score, precision and recall for the performance evaluation.

Condition-Based Monitoring in Variable Machine Running Conditions Using Low-Level Knowledge Transfer With DNN

Traditional machine learning methods assume that training and testing data must be from the same machine running condition (MRC) and drawn from the same distribution. However, in several real-time industrial applications, this assumption does not hold. The traditional methods work satisfactorily in steady-state conditions but fail in time-varying conditions. In order to utilize time-varying data in variable MRCs, this article proposes a novel low-level knowledge transfer framework using a deep neural network (DNN) model for condition monitoring of machines in variable running conditions. The low-level features have been extracted in time, frequency, and time–frequency domains. These features are extracted from the source data to train the DNN. The trained DNN-based parameters are then transferred to another DNN, which is modified according to the low-level features extracted from the target data. The proposed approach is validated through three case studies on: 1) the air compressor acoustic data set; 2) the Case Western Reserve University bearing data set; and 3) the intelligent maintenance system bearing data set. The prediction accuracy obtained for the above case studies is as high as 100%, 93.07%, and 100%, respectively, with fivefold cross-validation. These real-time results show considerable improvement in the prediction performance using the proposed approach. <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Note to Practitioners</i> —Condition-based monitoring schemes are widely applicable to rotating machines in various industries since they operate in tough working situations, and consequently, unpredicted failures occur. These unpredicted failures may cause perilous accidents in the industries. CBM systems prevent such failures, which results in the reduction of equipment damage and, hence, increases machinery lifetime. Modern industries are so complex and generating huge data, and these data can be collected using sensors, but placing a large number of sensors is difficult and expensive for different but similar kinds of faults in industries. This also increases the cost due to additional sensors and circuits. In this article, the authors have proposed a novel low-level knowledge transfer framework using the deep neural network (DNN)-based method for condition monitoring of machines in variable running conditions. Low-level features have been extracted to reduce the computations of DNN drastically with improved performance. This article also considered additional faults in the target domain, which is more practical in real-time applications. The proposed scheme has been validated with three case studies on acoustic and vibration signatures.

Modelling, simulation, and experimental verification of a pendulum-flywheel vibrational energy harvester

Vibration energy harvesting has been a popular topic in recent years. This technology is promising in developing self-powered sensor nodes for health condition monitoring of machines or structures, especially in remote areas. This study proposes a pendulum-flywheel vibration energy harvester based on the electromagnetic energy conversion mechanism. The harvester has two motion modes, namely the pendulum mode and eccentric flywheel mode, and can switch between the two modes automatically in response to external excitations.We first establish a theoretical model and fabricate a prototype of the harvester for evaluating its performance. Then, experimental and theoretical methods are employed to estimate the parameters of the model, such as the dipole moment of magnets, the mechanical damping coefficients, and the optimal resistance of the external electrical load. The typical trajectories of different motion modes, the frequency response characteristics, and the influence factors on the basins of attraction of the harvester are studied with the theoretical model. It is found that the small magnet distance can broaden the frequency band and enlarge the amplitude of the dynamic responses of the system. This finding provides us with an approach to control the performance of harvester and enables it to have stronger adaptability to variant ambient vibration in nature. Finally, laboratory tests are performed to validate the theoretical model. The experimental data verified the assumption that the rotation speed of the pendulum and the induced electromotive voltage have a linear relationship. Experimental and numerical simulation results show that the errors between them in most cases are less than 10% when the excitation displacement is small and have a slight increase with the excitation displacement. In the experiments, this harvester achieves a maximum power of 16.3 mW, exhibiting good performance in comparison with the-state-of-the-art pendulum-based harvesters.

Condition monitoring and intelligent diagnosis of rolling element bearings under constant/variable load and speed conditions

Extensive research has been conducted for intelligent fault diagnosis and prognosis of rolling element bearings, a vital component in every rotating machinery, and many robust and reliable techniques have been developed thus far. The majority of the proposed approaches, however, are established for constant operational conditions and therefore encounter difficulties when working conditions vary, which is common in industrial applications. The reason is that many characteristics of a normal state of a system in one working condition might be similar to the characteristics of a defected one in another working condition. The aim of this paper is to develop a method that can differentiate between different health states of machinery, regardless of load and speed conditions. For this purpose, a newly proposed approach, namely spectral amplitude modulation (SAM), is employed to highlight various components of a signal with different energy levels. Subsequently, the impulsivity of these extracted signals’ envelope spectrum is computed to quantify their cyclostationarity level. These quantities could be further utilized as the input variables of machine learning algorithms for automated and intelligent diagnosis of bearings. In this paper, two methods for data classification, namely support vector machine (SVM) and subspace k-nearest neighbors, are employed. Moreover, the computed impulsiveness of signals contains information about the health state of machinery and therefore could be employed as a health indicator for online condition monitoring of machines. To thoroughly assess the potential of the proposed method for condition monitoring and intelligent diagnosis of machinery in constant and highly variable working conditions, it is implemented on data collected from three distinct test rigs, namely the IMS, PoliTo and FEMTO data sets. The damages on bearings in those experiments have different severity levels, types, and they are located on different components of the bearings. In addition to localized defects, distributed faults, which are advanced and critical stages of defects, are also studied in this research. This type of defect is more difficult to detect and has been largely overlooked due to the fact that the characteristics of its signals are different from localized ones and similar to other modulation sources.

RETRACTED ARTICLE: An intelligent Context Based Multi-layered Bayesian Inferential predictive analytic framework for classifying machine states

Proactive fault diagnosis is a burning issue in condition monitoring of machines. Intelligent methods prove to be promising solutions for designing predictive analytic frameworks for fault diagnosis and machine state classification. The competency of machine learning algorithms in handling large voluminous data has marked them as a natural solution for developing intelligent framework that proactively classifies the machine states. The paper proposes a novel Context Based Multi-layered Bayesian Inferential (CBMBI) predictive analytic framework, which is motivated by MisMatch Negativity (MMN) and Predictive Coding. The CBMBI framework is augmented with a new hyperparameter (context) that greatly reduces the misclassification rate. The performance of the framework is analysed with Case Western Reserve University 6205-2RS JEM SKF dataset. The profound results reveal that the proposed framework shows 97% accuracy and 94% F1-score which is relatively higher than the state of art technique.

Condition Monitoring of Machines Using Fused Features From EMD-Based Local Energy With DNN

Several data-driven methods such as signal processing and machine learning exist separately to analyze non-linear and non-stationary data but their performance degrades due to insufficient information in the real-time application. In order to improve the performance, this paper proposes a novel feature extraction method using fusion of hand-crafted (low-level) features and high-level features, followed by feature extraction/selection on fused features. Local energy-based hand-crafted features have been derived from empirical mode decomposition, and high-level features have been extracted from the deep neural network. A method is also proposed for reduction of massive data points in the samples. The proposed scheme has studied the effect of variation in the number of extracted/selected features. The effectiveness of the proposed scheme is validated through three case studies: a) on acoustic dataset collected from the reciprocating type air compressor, b) on vibration dataset collected from deep groove ball bearing, and c) on steel plate faults dataset. The classification accuracy on acoustic dataset are obtained as high as 100.0%, 99.78%, and 99.78% using the random forest, linear support vector machine, and radial basis function support vector machine, respectively, with 5-fold cross-validation. Similarly, on vibration dataset obtained accuracies are 100.0%. The proposed scheme has been compared with ten conventional methods on five-fold cross-validation. These experimental results show considerable improvement in the prediction performance of machine conditions using the proposed scheme.

Adaptive Channel Weighted CNN With Multisensor Fusion for Condition Monitoring of Helicopter Transmission System

Deep learning-based multi-sensor fusion approach has been widely used for machine condition monitoring. Complementary information from different physical sensors or the same sensors on multiple locations of the monitored target can effectively improve the accuracy of condition monitoring. While, how to distinguish importance of every single sensor for condition monitoring is rarely researched. To address this gap, an adaptive channel weighted convolutional neural network (ACW-CNN) is proposed in this paper to investigate importance of different sensors in fusion approach. The designed ACW layer in a deep neural network can calibrate sensors weight according to sensors importance. The recalibrated channel weights can be used as the guidance for sensor position optimization. Furthermore, a new loss function, that is, Focal Loss, is introduced into ACW-CNN to deal with class imbalance generated in real helicopter transmission system (HTS) monitoring. To validate the effectiveness of the proposed ACW-CNN for condition monitoring, two case studies including condition monitoring to gearbox transmission system and helicopter transmission system are carried out. The results of comparative experiments show the proposed method, that is ACW-CNN with Focal Loss, is superior to other methods in classification accuracy, G-mean, and F-measure for condition monitoring of HTS.

Vibration Analysis of Shaft Misalignment Using Machine Learning Approach under Variable Load Conditions

The Industry 4.0 revolution is insisting strongly for use of machine learning-based processes and condition monitoring. In this paper, emphasis is given on machine learning-based approach for condition monitoring of shaft misalignment. This work highlights combined approach of artificial neural network and support vector machine for identification and measure of shaft misalignment. The measure of misalignment requires more features to be extracted under variable load conditions. Hence, primary objective is to measure misalignment with a minimum number of extracted features. This is achieved through normalization of vibration signal. An experimental setup is prepared to collect the required vibration signals. The normalized time domain nonstationary signals are given to discrete wavelet transform for features extraction. The extracted features such as detailed coefficient is considered for feature selection viz. Skewness, Kurtosis, Max, Min, Root mean square, and Entropy. The ReliefF algorithm is used to decide best feature on rank basis. The ratio of maximum energy to Shannon entropy is used in wavelet selection. The best feature is used to train machine learning algorithm. The rank-based feature selection has improved classification accuracy of support vector machine. The result obtained with the combined approach are discussed for different misalignment conditions.

A nonparametric health index and its statistical threshold for machine condition monitoring

Machine condition monitoring uses monitoring data to evaluate machine health conditions and conduct condition-based maintenance. Nowadays, kurtosis, entropy, Gini index and smoothness index are popular indices for machine condition monitoring and they fall into a unified framework. A problem is that, if monitoring data do not follow a particular assumed parametric health index, the parametric index is not fully useful for machine condition monitoring. Another problem is that parametric health indices lack their associate statistical thresholds at a significance level for machine condition monitoring. In this paper, statistical modeling and statistical analysis of normalized square envelope spectrum are proposed to construct a nonparametric health index and its associate statistical threshold at a significance level for machine condition monitoring. An illustrative bearing run-to-failure example showed that the proposed nonparametric health index and its statistical threshold can assess degradation well without needing a specific parametric form and abnormal and faulty datasets.

Novel desiccant-based very low humidity indicator for condition monitoring in miniaturized hermetic packages of active implants

Abstract Hermetic packaging is a suitable means to achieve long lifetime for implantable medical devices. In order to determine the time to failure of the hermetic packages, the prediction of lifetime by helium leakage measurements prior to implantation is state of the art. However, these methods are not applicable to packages with very small internal cavity volumes, as these established methods reach their resolution limits in this case. In these packages, even the smallest amounts of water can lead to corrosion or short circuits and thus to a malfunctioning system with unexpected behavior. Online monitoring of humidity offers the advantage that critical conditions are detected before a failure occurs. A typical maximum limit is 5000 ppm water vapor inside the package, which corresponds to a relative humidity of around 8.1 %RH at 37 °C. This paper presents the design and implementation of capacitive micromachined humidity detectors with a desiccant-based dielectric in different designs. These allow the electrical detection of moisture absorption, especially in the very low humidity regime ≤ 8 %RH. For this humidity range design-dependent sensitivities of 0.08 pF/%RH and 0.25 pF/%RH were achieved. Presented sensor devices facilitate active monitoring of current moisture content in the package cavity and at the same time provide a buffer for ingressing moisture, thus increasing the overall lifetime of the system. The current sensors are designed for miniaturized active medical implants, but can also be transferred to distributed sensors in harsh environments for Internet of Things applications or machine condition monitoring in Industry 4.0 scenarios.

Condition Monitoring for Railway Point Machines Based on Sound Analysis and Support Vector Machine

Railway point machines (RPMS) are one of the key equipments in the railway system to switch different routes for the trains. Condition monitoring for RPMs is a vital measure to keep train operation safe and reliable. Taking convenience and low cost into consideration, a novel intelligent condition monitoring method for RPMs based on sound analysis is proposed. Time-domain and frequency-domain features are obtained, and normalized using z-score standardization method to eliminate the influences of different dimensions. Binary particle swarm optimization (BPSO) is utilized to select the most significant discrimination feature subset. The effects of the selected optimal features are verified using Support vector machine (SVM), 1-Nearest neighbor (1NN), Random forest (RF), and Naive Bayes (NB). Experiment results indicate SVM performs best on identification accuracy and computing cost compared with the other three classifiers. The identification accuracies on normal switching and reverse switching processes reach 100% and 99.67%, respectively, indicating the feasibility of the proposed method.

Machine monitoring system: a decade in review

In recent years, significant advancements have been achieved in the domain of machine monitoring as witnessed from the beginning of the new industrial revolution (IR) known as IR 4.0. This new revolution is characterized by complete automation, and an increase in the technological deployments and advanced devices used by various systems. As a result, considerable advancement has been reported by researchers and academia around the world by adopting and adapting the new technology related to the machine condition monitoring problem. For these reasons, it is important to highlight new findings and approaches in machine condition monitoring based on the wireless sensor solution and machine learning signal processing methodology that are relevant to assist in advancing this new revolution. This article presents a comprehensive review on tool condition monitoring (TCM), tool wear, and chatter based on vibration, cutting force, temperature, surface image, and smart label monitoring parameters from signal acquisition, signal processing methodology, and decision-making, particularly for the milling process. The paper also provides a brief introduction to the manufacturing industries and computer numerical control (CNC) machine tool demand and a review of machine monitoring and countries involvement in machine monitoring, as well as the approaches and a survey of machine monitoring. The aim is to contribute to this rapidly growing field of machine condition monitoring research by exploring the latest research findings on the solution approach for milling machine process monitoring, to help expedite future research, and to give some future direction that needs to be considered as a path to produce a standardized CNC machine platform.

Entropy Measures in Machine Fault Diagnosis: Insights and Applications

Entropy, as a complexity measure, has been widely applied for time series analysis. One preeminent example is the design of machine condition monitoring and industrial fault-diagnostic systems. The occurrence of failures in a machine will typically lead to nonlinear characteristics in the measurements, caused by instantaneous variations, which can increase the complexity in the system response. Entropy measures are suitable to quantify such dynamic changes in the underlying process, distinguishing between different system conditions. However, notions of entropy are defined differently in various contexts (e.g., information theory and dynamical systems theory), which may confound researchers in the applied sciences. In this article, we have systematically reviewed the theoretical development of some fundamental entropy measures and clarified the relations among them. Then, typical entropy-based applications of machine fault-diagnostic systems are summarized. Furthermore, insights into possible applications of the entropy measures are explained, as to where and how these measures can be useful toward future data-driven fault diagnosis methodologies. Finally, potential research trends in this area are discussed, with the intent of improving online entropy estimation and expanding its applicability to a wider range of intelligent fault-diagnostic systems.

FPGA implementation of a wireless sensor node with a built-in ADALINE neural network coprocessor for vibration analysis and fault diagnosis in machine condition monitoring

Abstract Industry is a very attractive research field for wireless sensor network (WSN) applications. This is demonstrated by the creation of a special category of these networks dedicated to industrial applications, called industrial wireless sensor networks (IWSN). The sensor node, the main component of the network, must have several characteristics, such as a very high processing speed, reliable results, communication capabilities, and reduced transmission time. In this article, we outline the results of replacing the fast Fourier transform (FFT) processing of vibrational signals with an artificial adaptive linear element (ADALINE) neural network in order to extract the harmonics of the signals and thus detect faults in rotating machines. In addition, a MicroBlaze soft-core processor and an nRF24L01+ transmitter was chosen to manage various tasks within the node and the data exchange between the nodes of the network. A Digilent Cmod A7 platform with an Artix-7 FPGA circuit from Xilinx was selected to implement the different blocks that constitute the wireless node. Practical tests showed that the choice of the ADALINE enabled us to achieve the desired results by reducing the processing time to 7.478 ms, which is a reduction of time of approximately 85% compared to results obtained in scientific research. In addition, we have reduced the number of transmitted packets to only two. These results will have a positive impact on the performance of the node, with measurements using a periodic measurement methodology showing that the lifetime of the node can reach up to 17 h.

Development of a speed invariant deep learning model with application to condition monitoring of rotating machinery

The application of cutting-edge technologies such as AI, smart sensors, and IoT in factories is revolutionizing the manufacturing industry. This emerging trend, so called smart manufacturing, is a collection of various technologies that support decision-making in real-time in the presence of changing conditions in manufacturing activities; this may advance manufacturing competitiveness and sustainability. As a factory becomes highly automated, physical asset management comes to be a critical part of an operational life-cycle. Maintenance is one area where the collection of technologies may be applied to enhance operational reliability using a machine condition monitoring system. Data-driven models have been extensively applied to machine condition data to build a fault detection system. Most existing studies on fault detection were developed under a fixed set of operating conditions and tested with data obtained from that set of conditions. Therefore, variability in a model’s performance from data obtained from different operating settings is not well reported. There have been limited studies considering changing operational conditions in a data-driven model. For practical applications, a model must identify a targeted fault under variable operational conditions. With this in mind, the goal of this paper is to study invariance of model to changing speed via a deep learning method, which can detect a mechanical imbalance, i.e., targeted fault, under varying speed settings. To study the speed invariance, experimental data obtained from a motor test-bed are processed, and time-series data and time–frequency data are applied to long short-term memory and convolutional neural network, respectively, to evaluate their performance.

Tracking Operation Status of Machines through Vibration Analysis using Motif Discovery.

Industrial revolution 4.0 is inevitable for developing countries and created an urgency for many small and medium enterprises (SME) manufacturers to modernize their operations. Unfortunately, the process of modernization is expensive and it may not be justifiable for SMEs. This work is part of our existing effort in developing retrofitting sensors solution to help in reducing the cost and risk of SMEs moving towards Industry 4.0. This work focuses on tracking operation of machine through vibration data. Vibration analysis is not new and there are many existing works especially in the area of machine condition monitoring. This work is different from machine condition monitoring because the aim of this work is to track operation status of machines. The operation status of machines are crucial for manufacturers to manage maintenance and estimate the throughput of their operation. However, the main challenge of tracking operation status is that there are no prior knowledge on the machine’s vibration. The success of Motif Discovery has provided an opportunity to distinguish and identify the operation status of a machine from its vibration with minimal interference to the machine’s operation. This work puts Matrix Profile to test; using 15 sets of vibration data from 2-speed (high and low speed) industrial fan and exhaust hood, with the period for each collection at 10 minutes. From the experimental results, this work showed that it is possible to track the operation status of a machine through its vibration data and the accuracy can be as high as 99%.

Automatic Condition Monitoring of Industrial Machines Using FSA-Based Hall-Effect Transducer

A reliable and swift real-time condition monitoring for an industrial machine is a need of the hour when dealing with critical uncontrollable machine failures such as stator winding insulation failure, which is more prominent while working in robust conditions. Early detection of stator winding faults, especially turn-turn fault, occurred mainly as a short-circuit between the winding of the same phase due to the insulation failure, which can prevent fault propagation and protect winding from further insulation damages. This paper presents steady and transient state analysis of three-phase induction motor using fault signature analyzer based hall-effect transducer under various fault conditions. A numerical model with stator winding faults has been implemented, taking into account two major faults such as stator turn-turn (STTF) and stator phase-neutral fault (SPhN), using finite element method (FEM). FEM obtains the symmetrical current components of the motor under different fault conditions using finite element analysis (FEA). The machine torque has also been modeled and compared in a healthy and faulty condition. The numerical finite element method (FEM) model could identify a healthy and faulty induction machine with 95% accuracy using finite element analysis (FEA) and the same was validated using experimental setup results under the same operating conditions.

Intelligent wood machining monitoring using vibration signals combined with self-organizing maps for automatic feature selection

Data-driven models were developed for monitoring the power consumption and wood quality during the lumber manufacturing process. The study proposes hybrid models using vibration signals combined with self-organizing maps (SOMs) for cutting power and waviness prediction in the circular sawing process of Douglas-fir wood under very high feed speed conditions. The acquired vibration signals were fed into SOMs with different topologies for data mapping and automatic sensory feature selection, which were fed into the adaptive neuro-fuzzy inference system (ANFIS) or multilayer perceptron (MLP) neural network (NN) for cutting power and waviness prediction. The monitoring performance of the hybrid SOM-MLP NN and SOM-ANFIS models was compared. The frequency response of vibration signals and its correlation with the cutting parameters as well as the cutting power and waviness were discussed. The study shows that the developed SOM models to select the optimal features from the vibration signals could accurately predict cutting power and waviness when combined with a machine learning model. The prediction performance is highly dependent on the optimal choice of SOM topology. Having a poor choice of SOM topology, fine-tuning the architecture of ANFIS and MLP NN is crucial and greatly impacts the monitoring performance. Based on the obtained results, SOM is recommended for the automatic feature in the machining or tool condition monitoring, where due to process high variability, manual feature selection is a challenging task. SOM combined with an ANN or ANFIS makes a powerful intelligent model for monitoring complex processes such as wood circular sawing.

WP-DRnet: A novel wear particle detection and recognition network for automatic ferrograph image analysis

Abstract Ferrography plays an important role in wear analysis for machine condition monitoring, in which effective and efficient wear particle analysis is regarded as a crucial pre-requisite. An automatic wear particle detection and classification process is developed here using a cascade of two convolutional neural networks and a support vector machine (SVM) classifier. The neural networks are used for particle detection and recognition while particle classification is conducted in the SVM. This structure ensures that the computation expense is reduced and the accuracy is improved. The proposed network is verified using a large number of ferrograph images. Results show that high classification accuracies are obtained. Furthermore, the proposed approach can be further developed and applied in online machine condition monitoring applications.