Model-based Condition Monitoring Research Articles

Adaptive Ensemble Learning for Machine Tool Prognostics from Meta-Feature-Based Context Information

Model-based condition monitoring systems are widely used for the maintenance planning of manufacturing systems. With further digitalization, these models are more dependent on data. However, most studies are limited to single product or machine working conditions. Context information, such as machine parameters, material information, or environmental parameters are often ignored, neglected, or directly fed into prediction models. In this study, we present a new method called MetaSPX that leverages the context information to adapt the weights of a prediction ensemble. Models are only trained on single multi-variate run-to-failure time series and then added to the ensemble. Spatial proximity calculations of meta-features of the context information, combined with cross-validation knowledge of the ensemble members, are the main drivers of this method. It is tested on a Turbofan data set and compared against a single support vector regression model and an averaging ensemble. The results show a 67% reduction of the average prediction error starting at only two ensemble members compared to the averaging ensemble.

Model-based condition monitoring of piezoelectric bending actuators

Model-based condition monitoring of piezoelectric bending actuators

A review: multiplicative faults and model-based condition monitoring strategies for fault diagnosis in rotary machines

A review: multiplicative faults and model-based condition monitoring strategies for fault diagnosis in rotary machines

Reconstruction of sleeper displacements from measured accelerations for model-based condition monitoring of railway crossing panels

Railway switches and crossings (S&C, turnouts) connect different track sections and create a railway network by allowing trains to change tracks. This functionality comes at a cost as the load-inducing rail discontinuities in the switch and crossing panels cause much larger degradation rates for S&C compared to regular plain line tracks. The high degradation rates make remote condition monitoring an interesting prospect for infrastructure managers to optimise maintenance and ensure safe operations. To this end, this paper addresses the development of tailored signal processing tools for condition monitoring using embedded accelerometers in crossing panels. Multibody simulations of the dynamic train–track interaction are used to aid the interpretation of the measured signals in a first step towards building a model-based condition monitoring system. An analysis is performed using sleeper acceleration measurement data generated by 100 000 train passages in eight crossing panels. Based on the given data, a novel frequency-domain displacement reconstruction method is developed and the robustness of the method with respect to encountered operational variability of the measured data is demonstrated. The separation of the track response into quasi-static and dynamic domains based on deformation wavelength regions is proposed as a promising strategy to observe the ballast condition and the crossing geometry condition, respectively.

Model-Based Condition Monitoring of the Sensors and Actuators of an Electric and Automated Vehicle

Constant monitoring of driving conditions and observation of the surrounding area are essential for achieving reliable, high-quality autonomous driving. This requires more reliable sensors and actuators, as there is always the potential that sensors and actuators will fail under real-world conditions. The sensitive condition-monitoring methods of sensors and actuators should be used to improve the reliability of the sensors and actuators. They should be able to detect and isolate the abnormal situations of faulty sensors and actuators. In this paper, a developed model-based method for condition monitoring of the sensors and actuators in an electric vehicle is presented that can determine whether a sensor has a fault and further reconfigure the sensor signal, as well as detect the abnormal behavior of the actuators with the reconfigured sensor signals. Through the simulation data obtained by the vehicle model in complex road conditions, it is proved that the method is effective for the state detection of sensors and actuators.

Multi-domain system level modeling approach for assessment of degradation behaviour under thermal and thermo-mechanical stress

Within the scope of this work, a multi-domain system modeling approach that is compatible with various electronical components and subsystems used in automotive and railway applications is to be developed. This approach serves the development of a hybrid, model-based condition monitoring of complex electronic and mechatronic systems. It aims at the implementation of condition monitoring in relevant applications of automotive and railway technology, e.g., safety-relevant electronic systems for train control and control units of electrified automobiles. Furthermore, the multi-physics modeling of the mentioned electronic components is performed using different object-oriented simulation environments, such as Simulink and OpenModelica11Simulink® and Modelica® are registered trademarks. in order to obtain as much information as possible that is relevant for the reliability assessment of the modeled systems. The approach is applied on two common example circuits used in complex electronic systems, a clocking circuit with a crystal oscillator and a full-wave bridge rectifier circuit. Moreover, the investigated circuits as well as the degradation behaviour under external loads during the operational lifetime in order to model the functionality are demonstrated.

A Comprehensive Review on Signal-Based and Model-Based Condition Monitoring of Wind Turbines: Fault Diagnosis and Lifetime Prognosis

Wind turbines play an increasingly important role in renewable power generation. To ensure the efficient production and financial viability of wind power, it is crucial to maintain wind turbines’ reliability and availability (uptime) through advanced real-time condition monitoring technologies. Given their plurality and evolution, this article provides an updated comprehensive review of the state-of-the-art condition monitoring technologies used for fault diagnosis and lifetime prognosis in wind turbines. Specifically, this article presents the major fault and failure modes observed in wind turbines along with their root causes, and thoroughly reviews the techniques and strategies available for wind turbine condition monitoring from signal-based to model-based perspectives. In total, more than 390 references, mostly selected from recent journal articles, theses, and reports in the open literature, are compiled to assess as exhaustively as possible the past, current, and future research and development trends in this substantial and active investigation area.

Estimation of the railway equivalent conicity under different contact adhesion levels and with no wheelset sensorization

The running stability of a railway vehicle, which is highly relevant for operational safety, strongly depends on the shapes of the profiles of wheel and rail, which change during the progress of wear. The model-based condition monitoring of a key parameter, such as the wheelsets equivalent conicity, is one of the approaches to realize condition-based maintenance and ensure railway vehicles reliability and running safety. A constrained nonlinear estimation technique is developed in this paper concerning the monitoring of the equivalent conicity. Suitable modelling of the railway vehicle dynamics is presented for the model-based estimator synthesis to perform the equivalent conicity monitoring employing only the carbody and bogie frame measurements, taking into account the track adhesion level variation. This approach allows avoiding the wheelset sensorization affected by a harsh and vibrational environment, thus reducing the possibility to produce inaccurate measurements. Another direct consequence of this approach is the reduction of costs relative to vehicle sensorization. Moreover, the possibility to constraint the estimated vehicle states and parameters allows considering their operative acceptance limits.

DYNAMIC INVESTIGATION OF A PERMANENT MAGNET SYNCHRONOUS MOTOR FOR FAULTY OPERATIONS

Industrial efficacy of electrical energy is an important feature for our daily life. To provide efficient and reliable operation, new types of electrical machines and their condition monitoring have gain importance. In this paper, a model based condition monitoring and fault evaluation application for permanent magnet synchronous motors (PMSM) is presented. Possible fault conditions are realized on mathematical model and their results are discussed. In order to observe the dynamic behavior of PMSM, its mathematical model is implemented into MATLAB / Simulink as a reference. Then, faulty conditions are created by altering specific model parameters to imitate possible faults. The behavior of the motor is observed and compared with the reference output to contribute the dynamic effects of faulty operations.

Model-Based Condition Monitoring: State-Space Solution for Counter-Current Heat Exchanger

One of the key challenges to model condition monitoring is; how to suitable match the level of detail of the model description of the diagnostic requirements. To do so, a counter-current heat exchanger of the tire factory was investigated. Moreover, a two-stage grey-box modelling was deployed. The modelling is deployed in a discrete state space, relying on a combination of prior knowledge, along with the recorded data. The nonlinear model structure of heat exchanger is derived considering the use of the prior knowledge; thereby, it is converted into linearized form. The discretization of the linearized model is applied. The optimization procedure based on recorded data is adopted to estimate the unknown physical parameters in state space system matrices. The well-known N4SID subspace identification method is used to compare the obtained results.

Model-Based Condition Monitoring for the Detection of Failure of a Ball Bearing in a Centrifugal Pump

Centrifugal pumps are employed as important rotating machineries that play a vital role in sugar production plants. Untimely bearing failure, impeller damages, wearing out of sealing system and cavitation are the main faults that occur in centrifugal pump which need to be detected in advance for uninterrupted production activity in sugar plants. In such condition, the rolling bearing damage is a primary indicator of the possible major issues of the centrifugal pumps. The objective of this paper is to report a novel method that has been applied for the detection of the faulty bearing of a centrifugal pump used as a boiler feed pump in sugar production unit. A case study particularly of a deep groove ball bearing used to support non-drive end of pump is presented. Theoretical formulation of response parameters such as vibration velocity and frequency has been carried out using dimensional analysis. Subsequently, the in situ experiments have been performed using multichannel vibration analyzer (Brand—ADASH, Model-A4400 VA4 Pro) for validation.

EMD-CNN model based condition monitoring and fault warning for power overhead optical cable communication

Optical fiber communication has become the main communication mode of power communication in China. As the carrier and foundation, the stability and security of power communication optical cable is directly related to the reliability and security of power communication network in China. In this paper, the idea of in-depth learning is introduced into the condition monitoring of communication optical cable, taking the all-dielectric self-supporting optical cable (ADSS) as an example. By transforming the image into picture set and choosing the appropriate depth neural network model to analyse the picture, the condition monitoring of communication optical cable can be realized.

Model-Based Condition Monitoring of a Vanadium Redox Flow Battery

The safe, efficient and durable utilization of a vanadium redox flow battery (VRB) requires accurate monitoring of its state of charge (SOC) and capacity decay. This paper focuses on the unbiased model parameter identification and model-based monitoring of both the SOC and capacity decay of a VRB. Specifically, a first-order resistor-capacitance (RC) model was used to simulate the dynamics of the VRB. A recursive total least squares (RTLS) method was exploited to attenuate the impact of external disturbances and accurately track the change of model parameters in realtime. The RTLS-based identification method was further integrated with an H-infinity filter (HIF)-based state estimator to monitor the SOC and capacity decay of the VRB in real-time. Experiments were carried out to validate the proposed method. The results suggested that the proposed method can achieve unbiased model parameter identification when unexpected noises corrupt the current and voltage measurements. SOC and capacity decay can also be estimated accurately in real-time without requiring additional open-circuit cells.

An On-Board Model-Based Condition Monitoring for Lithium-Ion Batteries

A model-based condition monitoring for lithium-ion (Li-ion) batteries involves estimating critical model parameters (e.g., capacity and impedance) and operational states [e.g., state of charge (SOC) and state of health]. This is important to design high-performance and safety-critical battery systems. Moreover, the tuning parameters of the condition monitoring algorithms significantly influence the performance of the algorithms. This paper proposes a real-time model-based condition monitoring for Li-ion batteries based on a real-time second-order resistor–capacitor electrical circuit battery model. The proposed method consists of an extended Kalman filter based online parameter identification and a smooth variable structure filter based state estimation. The two filters are systematically integrated to cooperate with each other and named hybrid filter (HF). Furthermore, a proposed particle swarm optimization based tuning parameter optimization is employed to find the optimal tuning parameters of the HF. The proposed method is compared with the dual extended Kalman filter (DEKF) via simulations and validated by experiments using a low-priced microcontroller. The results show that the proposed method can effectively choose the tuning parameters and has less computational complexity and higher accuracy of the SOC and capacity estimation than those of the conventional DEKF. Therefore, the proposed HF can be suitable for use in real-time embedded battery management system applications.

Velocity measurement-based friction estimation for railway vehicles running on adhesion limit: swarm intelligence-based multiple models approach

Model-based condition monitoring is an increasingly important area for rail transportation. The key elements of such condition monitoring methodologies are low-cost vehicle sensors and intelligent algorithms. In this study, a swarm intelligence-based multiple models approach is proposed to detect different friction conditions by using velocity measurements of a railway vehicle. In this case of application, estimated parameter is the maximum friction coefficient. Additionally, proposed methodology is tested experimentally by using the measurements taken from a tram wheel test stand. Multiple mathematical models of the test stand are created with different maximum friction coefficients, whereas all initial conditions and other system parameters are same for each model. Therefore, comparison of the output of each model with measurements is considered to interpret the parameter value of the model, which best represents the system, is selected as parameter estimate. Unlike the traditional multiple models approach, a swarm intelligence-based evolution of the models is proposed. Experiments carried out on the test stand reveal that the proposed methodology is promising to be used as an on-board friction condition monitoring tool for railway vehicles with traction. Furthermore, it can be considered to detect weather conditions since friction conditions change due to the weather events such as rain, ice, snowfall, condensation of water droplets, and leaves on the line and it can be used as an auxiliary system for intelligent traction and high adhesion control systems.

A Condition Monitoring Methodology Based on a Non-Parametric System Identification Approach

Abstract The paper presents a non-parametric approach to a general model-based scheme of the condition monitoring of a technology process. Within such an approach, the total inference on a technology process status is implemented by current comparing preliminary nominal indicators obtained, which correspond to the nominal technology process model with corresponding characteristics obtained by current observations. A consistent measure of stochastic dependence of random processes is proposed to be used as a technology process behavior indicator. Such a measure is the maximal correlation function. It consistently captures the actual nonlinear dependence between random processes, while conventional measures of dependence based on the dispersion and, all the more so, ordinary product correlation functions do not. Meanwhile, there are known examples demonstrating that actual dependence between model variables may be nonlinear even if the regression of a variable onto another one is linear. Within the case, such dependence can be properly described by maximal correlation ultimately. Such a consideration justifies applying consistent measures of dependence when applied to the model-based condition monitoring methodology.

A novel wind turbine condition monitoring method based on cloud computing

With the development of condition monitoring technology, the data collected by sensors are voluminous and much faster than before. The cloud computing technology is a good solution for big data processing, it is therefore very suitable to be applied in the condition monitoring of the wind turbine, especially for data-driven model-based condition monitoring methods. In order to solve this problem, a novel wind turbine condition monitoring method based on cloud computing is proposed in this paper. A data-driven model-based condition monitoring (CM) method by using hierarchical extreme learning machine (H-ELM) algorithm is adopted to achieve fault detection of the gearbox in the wind turbine, which has better performance than traditional ELM method. Then, compressed sensing (CS) method is applied to compress the first hidden layer output that will be uploaded to the cloud for further calculation. The proposed method is not only able to detect the faults effectively, but also considering data upload quantity reduction and data security. The case study validates the effectiveness of the proposed method. Consequently, it is effective and can also enhance economic benefit and operating efficiency of the wind farm.

Model-based condition monitoring of guiding rails in electro-mechanical systems

Abstract This paper presents a model-based approach for the condition monitoring of guiding rails in vertical transportation systems. A physics-based model of the system is used to estimate friction forces by means of an augmented Extended Kalman Filter (EKF). This estimation is compared to the friction forces predicted by a more detailed model of the system. This comparison, provides quantitative knowledge of the condition of the system. The difference between the estimated and the expected friction is continuously tracked, allowing an early assessment of the system’s degradation. The method is validated using an scaled test bench of a vertical transportation system. It is shown that, using only off-the-shelf sensors, a model-based approach could simplify the condition monitoring of guiding rails, which generally requires costly and complex devices.

Observer design and optimization for model-based condition monitoring of the wind turbine rotor blades using genetic algorithm

In this paper a model based approach for condition monitoring and diagnosis of faults in the wind turbine rotor blades is described. The investigation is focused on creating a fault-free reference system, the design and optimization of the controller/observer based on the Disturbance Accommodating Control theory and the multi-objective genetic optimization algorithm in order to further enhance the condition monitoring and faults diagnosis. Test cases are presented by injecting a parameter change, e.g. pitch angle, as an error to the non-linear wind turbine simulation model in order to demonstrate that the proposed method is able to detect the fault on the blades.

An unscented Kalman filter-based rolling radius estimation methodology for railway vehicles with traction

Monitoring the conditions of railway vehicle systems plays an important role in the maintenance of safety and performance of railway vehicles. Rolling radius is one of the properties that should be monitored continuously for the predictive maintenance of a railway vehicle since it changes with time due to wheel wear. In this study, a model-based condition monitoring methodology, which is based on an unscented Kalman filter, is proposed. The model includes the torsional dynamics of an independently rotating tram wheel with a traction motor and a contact model. The rolling radius is estimated by considering the traction effort of the motor and the angular velocity measurements. The proposed methodology is tested on a tram wheel test stand (roller rig), which has a wheel on roller configuration. First, a mathematical model is validated by the measurements taken from the test stand. Second, the unscented Kalman filter is applied as a parameter estimator. The results demonstrate that the proposed scheme is a promising option to be used in the predictive condition monitoring of the wheel profile for traction vehicles.