Search

Highlights: [•] A comprehensive model is proposed for quick and accurate fault detection and identification in health monitoring. [•] A principal component analysis scheme is designed to reduce dimensionality of data collected by COSMED K4b2 and improve efficiency in data transmission. [•] A new method which combines the K-means clustering and artificial neural networks is developed to classify data and accurately detect and isolate faults. [•] The mechanism provides the feasibility and increases the efficiency of the fault detection and identification for a system with multiple processes. [Copyright &y& Elsevier]

When joint actuator faults of the legged robot systems occur, it is important to estimate the states and unknown faults (inputs) accurately for the emergent operation of the control system to protect the robot per se and the surrounding environment. The traditional extended Kalman filter (EKF) is a routine method for state estimation, however, it requires that all the input data are measured. In this paper, an extended Kalman filter with fault reconstruction, referred to as EKF-FR, is proposed to estimate the states as well as unknown joint actuator faults (inputs) for the planar one-legged articulated hopping robot system, which owns the fundamental dynamic characteristics for advanced legged robots. The major merit of the proposed EKF-FR approach lies in the fact that the unknown faults (inputs) are reconstructed by designing an input-output linearization controller with asymptotical stability based on the nonlinear robot system, hence the proposed approach is with more accuracy in comparison with the ones based on the Jacobian linearized systems, such as the extended Kalman filter with unknown inputs (EKF-UI). Simulation results for two single-fault cases demonstrate the effectiveness of the proposed EKF-FR algorithm. [ABSTRACT FROM AUTHOR]

Abstract: As modern chemical processes evolve into extremely large and complex systems involving multiple third-party technologies, instruments, and software, the information necessary for process supervision is distributed across disparate sources. Supervision methods rely on models that are themselves based on these process descriptors. Therefore, any changes during operations have to be reflected to the supervision methods so that they can adapt suitably. In this paper, we propose an ontology, called OntoSafe, that provides the semantics for abnormal situation management. It enables the integration of all the information necessary for forming a cohesive judgment of the condition and state of the process. It also captures the underlying linkages so that changes in the process descriptors are propagated consistently. We illustrate the benefits of these using the ENCORE distributed intelligence system. The application to an offshore oil and gas production case study is also reported. [Copyright &y& Elsevier]

The increased complexity and scale of high performance computing and future extreme-scale systems have made resilience a key issue, since it is expected that future systems will have various faults during critical operations. It is also expected that current solutions for resiliency, mainly counting on checkpointing in hardware and applications, will become infeasible because of unacceptable recovery time for checkpointing and restarting. In this paper, we present innovative concepts for anomaly detection and identification, analyzing the duration of pattern transition sequences of an execution window. We use a three-dimensional array of features to capture spatial and temporal variability to be used by an anomaly analysis system to immediately generate an alert and identify the source of faults when an abnormal behavior pattern is captured, indicating some kind of software or hardware failure. The main contributions of this paper include the innovative analysis methodology and feature selection to detect and identify anomalous behavior. Evaluating the effectiveness of this approach to detect faults injected asynchronously shows a detection rate of above 99.9% with no occurrences of false alarms for a wide range of scenarios, and accuracy rate of 100% with short root cause analysis time. [ABSTRACT FROM AUTHOR]

On-line, real-time collection of measurements remains a key challenge in water quality monitoring and control due to unknown and varying quality of on-line sensor data. Today's data quality assessment is typically based on a comparison of sensor-based measurements and grab samples of the sampled solution taken next to the on-line analyzer and analyzed in a laboratory. In this work, internal data is used for fault detection and identification of a phosphate analyzer to inspect the measuring process itself. These internal data is shown to be information-rich with respect to the analyzer's status. Furthermore, this information is captured well by means of a newly developed method for qualitative analysis of time series. This method was developed with global optimality in mind and therefore lends itself to a probabilistic assessment of the qualitative representation of time series. [ABSTRACT FROM AUTHOR]

Indoor air quality (IAQ) is an important factor, which can influence the health and comfort of passengers in subway stations. Various types of hazardous pollutants, such as particulate matters, remain accumulated in the subway space due to overcrowding and inadequate ventilation system. Subway stations are extremely crowded during rush hours and indoor air of the subway stations could be strongly affected, which in turn, affects passengers’ respiratory system. In this study, several key air pollutants data were collected every minute by the air sampler and tele-monitoring system (TMS) to effectively monitor and control IAQ in subway stations. The quality of the online measurement could decide the failure and success in environmental process assessment. Therefore, prompt detection of the occurrence of sensor faults and identification of those locations are of primary importance for efficient monitoring and control of IAQ. In this paper, Principal components analysis based approach is used to detect, identify and reconstruct the sensor faults in monitoring IAQ. Four types of sensor failures, namely, bias, drifting, complete failure and precision degradation are tested for monitoring IAQ. Several test results of a real subway TMS showed that the developed sensor validation technique can work well for the four kinds of sensor faults. [ABSTRACT FROM AUTHOR]

Abstract: To design an effective plant alarm system, it is essential to evaluate various performances of fault detection and identification (FDI) in emergencies. In this study, an operator model, which mimics the FDI behavior of a human operator with primary cognitive and executive capabilities, is developed as a virtual subject for supervising a chemical plant system. Analyzing the FDI tracks generated by the behavior simulation using the virtual subject under an alarm system makes it possible to evaluate the performance of the alarm system. In a case study, the evaluation method using the virtual subject is applied for designing an appropriate alarm system in a boiler-plant simulator. FDI performances of the original alarm system for all assumed malfunctions are evaluated and alarm signal selection and limit settings are reconsidered according to the evaluation results. The results of the case study demonstrate the usefulness of the model-based evaluation method. [Copyright &y& Elsevier]

The detection of motor faults at their incipient stage is gaining importance as it leads to increased reliability and reduced machine downtime. The stator current analysis has caught the attention of researchers as a mature and simple technique for induction motor fault detection and identification. In this paper, angular space vector analysis of the induction motor stator current for fault identification has been investigated. The tracking of spatial angular vector profile of stator current (Parke's vector) is used to identify the degrading health condition of induction motors. Any significant deviation in the shape of spatial angular vector is an indicator of the inset of irregularities -mechanical or electrical-in the induction motor. Three major types of induction motor faults-bearing fault, broken rotor bar fault, and unbalanced supply faults- have been experimentally investigated. The experimentation has been performed on a 3ø, 1.5 kW, 4 poles, 1440 RPM, ABB squirrel cage motor. The motor setup was mechanically loaded to operate at various loads. The TMS 320F420 DSP-based dSPACE DS 1104 control card has been used to carry out the experimentation. The softwares used include MATLAB® ver. 2006 and dSPACE controldesk. [ABSTRACT FROM AUTHOR]

This paper presents a method to increase the reliability of Unmanned Aerial Vehicle (UAV) sensor Fault Detection and Identification (FDI) in a multi-UAV context. Differential Global Positioning System (DGPS) and inertial sensors are used for sensor FDI in each UAV. The method uses additional position estimations that augment individual UAV FDI system. These additional estimations are obtained using images from the same planar scene taken from two different UAVs. Since accuracy and noise level of the estimation depends on several factors, dynamic replanning of the multi-UAV team can be used to obtain a better estimation in case of faults caused by slow growing errors of absolute position estimation that cannot be detected by using local FDI in the UAVs. Experimental results with data from two real UAVs are also presented. [ABSTRACT FROM AUTHOR]

Abstract: Oscillations are a common type of plant-wide disturbances whose effects propagate to many units and thus may impact overall process performance. It is important to detect and diagnose such oscillations early in order to rectify the situation. Many frequency domain tools such as the power spectrum and spectrum envelope methods are capable of detecting the oscillation frequency. However, few methods are available for locating the root cause which is the main objective of oscillation diagnosis. This paper proposes a new method to diagnose the root cause of plant-wide oscillations using the adjacency matrix. A novel feature of the new method is that it utilizes the information in the process flowsheet. The method is not data-based and it can be carried out without using any data. However, this method complements the data based methods very well and it is best used in combination with other data-based methods to provide powerful diagnosis of plant-wide oscillations. This paper is the subject of a newly proposed complete procedure for detection and diagnosis of plant-wide oscillation. Two industrial case studies are also presented to demonstrate the applicability of the proposed procedure. [Copyright &y& Elsevier]

This work deals with modeling and fault detection and identification for robot manipulator. We have used for a dynamical system a hybrid approach. The model is decomposed into two parts: first, a certain part modeled using classical analytical theory and it is preferable to be linear. Second, an uncertain part representing the nonlinearities neglected in the first part, which is modeled using neuro-fuzzy modeling. Both analytical redundancy and neuro-fuzzy modeling are used to improve robustness. The analytical redundancy is used to generate residuals for the fault detection and location procedure. The neuro-fuzzy modeling is used to model modeling errors and faults, which allows performing the robustness and the sensitivity. Thanks to neurofuzzy modeling the errors of modeling are compensated and the faults are well identified as it is shown through the results of simulation. [ABSTRACT FROM AUTHOR]

Ideally, when faults happen, the closed-loop system should be capable of maintaining its present operation. This leads to the recently studied area of fault-tolerant control (FTC). This paper addresses soft computing and signal processing based active FTC for benchmark process. Design of FTC has three levels: Level 1 comprises a traditional control loop with sensor and actuator interface and the controller. Level 2 comprises the functions of online fault detection and identification. Level 3 comprises the supervisor functionality. Online fault detection and identification has signal processing module, feature extraction module, feature cluster module and fault decision module. Wavelet analysis has been used for signal processing module. In the feature extraction module, feature vector of the sensor faults has been constructed using wavelet analysis, sliding window, absolute maximum value changing ratio and variance changing ratio as a statistical analysis. For the feature cluster module, the self-organizing map (SOM), which is a subtype of artificial neural network has been applied as a classifier of the feature vector. As a benchmark process three-tank system has been used. Control of the three-tank system is provided by fuzzy logic controller. Faults are applied to three level sensors. Sensor faults represent incorrect reading from the sensors that the system is equipped with. When a particular fault occurs in the system, a suitable control scheme has been selected on-line by supervisor functionality to maintain the closed-loop performance of the system. Active FTC has been achieved by switch mode control using fuzzy logic controller. Simulation results show that benchmark process has maintained acceptable performance with FTC for the sensor faults. As a result, when the system has sensor faults soft computing and signal processing based FTC helps for the best performance of the system. [ABSTRACT FROM AUTHOR]

This article presents a new scheme for incipient fault detection and its identification in transformers. The new approach is actually based on adaptive modeling of transformers using the transmission line method (TLM) obtained from the hysteresis model. The adaptive TLM observer representing no-load, quarter-load, half-load, and rated-load conditions is used for faults detection. The continuous wavelet transform (CWT) is performed on residuals that are obtained by comparing real system currents and calculated TLM observer currents in order to extract the features for fault identification. An adaptive fuzzy reasoning technique is used to identify incipient faults in the transformer. The sum of CWT coefficients of residuals is applied to the adaptive fuzzy rule-based decision-making unit to indicate the type of faults. The main advantage of the suggested scheme is that different types of incipient faults in the transformer can be correctly identified. The test results verify the effectiveness of the suggested method. [ABSTRACT FROM AUTHOR]

Abstract: In this paper, a novel fault detection and identification (FDI) scheme for a class of nonlinear systems is presented. First of all, an augment system is constructed by making the unknown system faults as an extended system state. Then based on the ESO theory, a novel fault diagnosis filter is constructed to diagnose the nonlinear system faults. An extension to a class of nonlinear uncertain systems is then made. An outstanding feature of this scheme is that it can simultaneously detect and identify the shape and magnitude of the system faults in real time without training the network compared with the neural network-based FDI schemes. Finally, simulation examples are given to illustrate the feasibility and effectiveness of the proposed approach. [Copyright &y& Elsevier]

This paper introduces a statistical fault detection and identification (FDI) scheme for aircraft systems, which uses flight attitude data rather than information from purposely developed physical or virtual sensors. The scheme is based on the modelling of relationships among the considered data via stochastic Time-dependent Functionally Pooled Non-linear AutoRegressive with Exogenous excitation (TFP-NARX) representations. These are globally valid inside a flight regime and under various considered environmental conditions, thanks to the pooling technique used for their identification. Moreover, due to the TFP-NARX coefficients being a function of time-dependent quantities, high modelling accuracy is achieved. The scheme's operation involves identifying nominal TFP-NARX models of relationships among the attitude data from an aircraft operating in a healthy state. Owing to fault occurrence, these relationships may change. Then, an in-flight comparison of the nominal and the current aircraft dynamics provides fault-related information, which is statistically evaluated for FDI purposes. The scheme's performance and robustness are assessed with numerous flights conducted throughout a flight regime under various manoeuvring settings and turbulence levels. [ABSTRACT FROM AUTHOR]

A scheme is proposed to detect, identify, and estimate failures, including abrupt total, partial, and multiple failures, in a dynamic system. The new approach, named 1M³ L, is developed based on variable-structure multiple-model estimation, which allows to improve performance by online adaptation. It uses an interacting multiple model estimator for fault detection and identification but the maximum likelihood estimator for estimating the extent of failure. It provides an effective and integrated framework for fault detection, identification, and state estimation. For two aircraft examples, the proposed approach is evaluated and compared with hierarchical multiple-model approaches and a widely used single-model residual-based generalized likelihood ratio approach in terms of detection and estimation performance. The results show that the IM³ L provides not only fast detection and proper identification, but also good estimation of the failure extent as well as robust state estimation. [ABSTRACT FROM AUTHOR]

Abstract: The use of autonomous helicopters in civilian applications requires the improvement of safety conditions to avoid potential accidents. Fault detection and isolation (FDI) plays an important role in this context. This paper presents an actuator and sensor FDI system for small autonomous helicopters. Fault detection is accomplished by evaluating any significant change in the behaviour of the vehicle with respect to the fault-free behaviour, which is estimated by using observers. Several types of faults have been considered. The effectiveness of the proposed approach is demonstrated by means of experimental results and simulations. [Copyright &y& Elsevier]

Abstract: Engine diagnostics is a typical multi-sensor fusion problem. It involves the use of multi-sensor information such as vibration, sound, pressure and temperature, to detect and identify engine faults. From the viewpoint of evidence theory, information obtained from each sensor can be considered as a piece of evidence, and as such, multi-sensor based engine diagnosis can be viewed as a problem of evidence fusion. In this paper we investigate the use of Dempster–Shafer evidence theory as a tool for modeling and fusing multi-sensory pieces of evidence pertinent to engine quality. We present a preliminary review of Evidence Theory and explain how the multi-sensor engine diagnosis problem can be framed in the context of this theory, in terms of faults frame of discernment, mass functions and the rule for combining pieces of evidence. We introduce two new methods for enhancing the effectiveness of mass functions in modeling and combining pieces of evidence. Furthermore, we propose a rule for making rational decisions with respect to engine quality, and present a criterion to evaluate the performance of the proposed information fusion system. Finally, we report a case study to demonstrate the efficacy of this system in dealing with imprecise information cues and conflicts that may arise among the sensors. [Copyright &y& Elsevier]

Abstract: A support vector machine (SVM) approach to the classification of transients in nuclear power plants is presented. SVM is a machine-learning algorithm that has been successfully used in pattern recognition for cluster analysis. In the present work, single- and multiclass SVM are combined into a hierarchical structure for distinguishing among transients in nuclear systems on the basis of measured data. An example of application of the approach is presented with respect to the classification of anomalies and malfunctions occurring in the feedwater system of a boiling water reactor. The data used in the example are provided by the HAMBO simulator of the Halden Reactor Project. [Copyright &y& Elsevier]

A vehicle health monitoring system based on analytical redundancy, is developed for automated passenger vehicles. A residual generator and a residual processor are designed together to detect and identify actuator and sensor faults of the Buick LeSabre rapidly. The residual generator includes fault detection filters and parity equations. It uses the control commands and sensor measurements to generate the residuals, which have a unique static pattern in response to each fault. Then, the residual processor interrogates the residuals by matching them to one of several known patterns. It computes the probability of each hypothesis conditioned on the history of residuals. The fault detection latency is reduced by integrating the design of the residual generator and the residual processor. The vehicle health monitoring system is evaluated in real-time on a Buick LeSabre. The vehicle sensor and actuator faults are simulated artificially by the computer or created manually by the driver. In one experiment, a real intermittent sensor fault occurred and was immediately detected and identified. The real-time evaluation demonstrates that the vehicle health monitoring system can detect and identify actuator and sensor faults under various disturbances and uncertainties with almost minimal detection latency. [ABSTRACT FROM AUTHOR]

A multiple fault detection and identification method based on fully decoupled parity equations for dynamic systems with known linear and unknown nonlinear terms is presented. The fully decoupled parity equation vectors is derived and it is shown that the residuals generated from it are decoupled, from other faults and the unknown nonlinear term and are sensitive only to specific actuator or sensor faults. The condition for the existence of the equation is also given. From the residuals generated from the fully decoupled parity equation, the faults are estimated using the recursive least-squares method. The performance of the proposed method is illustrated by applying it to detect, isolate, and identify faults in a simulated dc motor. [ABSTRACT FROM AUTHOR]

Abstract: Control systems that utilize switched linear controllers have proven to be useful (and, in some cases, essential) for accomplishing certain control objectives in particular classes of plants. These controllers are often digital in nature and, as such, are subject to internal hardware malfunctions (faults). In this paper, we present a systematic methodology for constructing embeddings to protect switched linear controllers against hardware faults that corrupt their internal state. Our methodology is based on replacing the original controller with a redundant (higher dimensional) controller that preserves the functionality of the original controller while enabling error detection and correction. More importantly, this methodology allows an external mechanism to detect and identify transient state-transition faults through non-concurrent (e.g. periodic) parity checks. The resulting error detection and correction procedures can then be performed periodically, thereby relaxing the reliability requirements and overhead associated with the checking mechanism. [Copyright &y& Elsevier]

Abstract: By use of the idea of fault detection and identification, this Letter proposes a new scheme to resolve the problem of chaotic private communication. From the point of view of fault detection and identification the scalar message signal hidden in the chaotic systems can be regarded as the component fault signal, thereby it can be detected and recovered using the model-based methods of fault detection and identification. The famous Duffing oscillator is used to illustrate and verify the effectiveness of this scheme. [Copyright &y& Elsevier]

Wastewater treatment plants (WWTP) are notorious for poor data quality and sensor reliability due to the hostile environment in which the measurement equipment has to function. In this paper, a structured residual approach with maximum sensitivity (SRAMS) based on the redundancy of the measurements is used to detect, identify and reconstruct single and multiple sensor faults in a single reactor for high activity ammonia removal over nitrite (SHARON) process. SRAMS is based on inferences, which are insensitive to the faults in the sensor of interest and sensitive to faults in the other sensors. It is used for four types of sensor failure detection: bias, drift, complete failure and precision degradation. The application of sensor validation shows that single and multiple sensor faults can be detected and that the fault magnitude and fault type can be estimated by the reconstruction scheme. This sensor validation method is not limited by the type or application of the considered sensors. The methodology can thus easily be applied for sensor surveillance of other continuously measuring sensors and analysers. [ABSTRACT FROM AUTHOR]

Abstract: In this paper, a robust fault detection and identification approach based on an adaptive observer is developed for uncertain continuous linear time-invariant systems with multiple discrete time-delays in both states and outputs. State and output faults of bias type that may evolve slowly or abruptly are considered, and the delay system is disturbed by unstructured bounded unknown inputs. Based on the scheme of [Trunov, A. B., & Polycarpou, M. M. (2000). Automated fault diagnosis in nonlinear multivariable systems using a learning methodology. IEEE Transactions on Neural Networks, 11, 91–101], a novel adaptive observer for detecting and estimating faults in the considered system is constructed, and robustness with respect to unknown inputs and sensitivity to faults of the detecting scheme are rigorously analyzed. The fault estimate and the state estimation error are then proved to be uniformly bounded. Finally, simulations of a heating process demonstrate that the proposed approach can detect the faults shortly after the occurrences without any false alarm and can approximate the faults with desired accuracy. [Copyright &y& Elsevier]

State-transition faults in digital sequential systems, such as finite-state logic controllers, have traditionally been handled by embedding the given system into a larger one, in a way that preserves the state evolution of the original system while enabling an external mechanism to concurrently perform checks to detect, identify and correct errors. In this paper, we develop a methodology for systematically constructing embeddings of one-hot encoded finite-state machines (FSMs) in a way that allows the external mechanism to capture transient state-transition faults via checks that are performed in a non-concurrent manner (e.g., periodically). More specifically, by employing coding techniques over finite fields, we completely characterize an appropriate class of redundant FSM embeddings and its corresponding non-concurrent error-detecting/identifying capabilities. These embeddings can be used to construct a redundant version of the given one-hot encoded FSM so that the external mechanism can detect and identify errors due to past state-transition faults based on an analysis of the current, possibly corrupted FSM state. As a result, the proposed error detection and identification approach relaxes the stringent requirements on the reliability of the checker and avoids the slowdown associated with concurrent checking. [Copyright &y& Elsevier]

On-line monitoring of penicillin cultivation processes is crucial to the safe production of high-quality products. In the past, multiway principal component analysis (MPCA), a multivariate projection method, has been widely used to monitor batch and fed-batch processes. However, when MPCA is used for on-line batch monitoring, the future behavior of each new batch must be inferred up to the end of the batch operation at each time and the batch lengths must be equalized. This represents a major shortcoming because predicting the future observations without considering the dynamic relationships may distort the data information, leading to false alarms. In this paper, a new statistical batch monitoring approach based on variable-wise unfolding and time-varying score covariance structures is proposed in order to overcome the drawbacks of conventional MPCA and obtain better monitoring performance. The proposed method does not require prediction of the future values while the dynamic relations of data are preserved by using time-varying score covariance structures, and can be used to monitor batch processes in which the batch length varies. The proposed method was used to detect and identify faults in the fed-batch penicillin cultivation process, for four different fault scenarios. The simulation results clearly demonstrate the power and advantages of the proposed method in comparison to MPCA. [Copyright &y& Elsevier]

A fault detection and identification algorithm, called optimal stochastic fault detection filter, is determined. The objective of the filter is to detect a single fault, called the target fault, and block other faults, called the nuisance faults, in the presence of the process and sensor noises. The filter is derived by maximizing the transmission from the target fault to the projected output error while minimizing the transmission from the nuisance faults. Therefore, the residual is affected primarily by the target fault and minimally by the nuisance faults. The transmission from the process and sensor noises is also minimized so that the filter is robust with respect to these disturbances. It is shown that the filter recovers the geometric structure of the unknown input observer in the limit where the weighting on the nuisance fault transmission goes to infinity. Further, the asymptotic behavior of the filter near the limit is determined by using a perturbation method. Filter designs can be obtained for both time-invariant and time-varying systems. [Copyright &y& Elsevier]

As modern process plants become more complex, the ability to detect and identify the faulty operation of pneumatic control valves is becoming increasingly important. In this work, a neural network pattern classifier is employed to carry out fault diagnosis and identification upon the actuator of a Fisher–Rosemount 667 industrial process valve. The network is trained with experimental data obtained directly from a software package that comes with the valve. This has eliminated the need for additional instrumentation of the valve. Using this software, tests are carried out to obtain experimental parameters associated with the valve performance for incorrect supply pressure, diaphragm leakage, and vent blockage faults. Specifically, the valve signature and dynamic error band tests are used to directly obtain lower and upper bench sets, minimum, maximum, and average dynamic errors, as well as the dynamic linearity. Additionally, valve deadband and hysteresis are measured graphically from the available valve signature plots for each faulty condition. The relationships between these parameters, for each fault, form signatures that are subsequently learned by a multilayer feedforward network trained by error back-propagation. The test results show that the resulting network has the ability to detect and identify various magnitudes of each fault. It is also observed that a smaller network with a shorter training time results when the valve deadband and hysteresis are included in the training data. Thus, the extra effort required to extract these parameters from the valve signature plots is justified. [Copyright &y& Elsevier]

Maintenance of process equipment is one of the inescapable tasks associated with the operation of process plants and, until relatively recently, it was implemented either on a routine basis or after the failure of equipment. Attitudes are changing and now many organizations are adopting methods for identifying incipient faults, so that maintenance can be scheduled before there has been a failure which would lead to loss of production and spoilage of raw material. These methods include schemes in which all the measurement signals in a plant are gathered and, in addition to providing the input to the various control systems, they are compared with model-based values for individual sections throughout the entire process, so that a warning or an alarm can be raised, according to the seriousness of the observed discrepancy. Although these methods represent a considerable advance on ‘routine maintenance’ and ‘repair after failure’, their effectiveness is limited by the quality and detail of the information that is gathered. Each process measurement should be regarded as a ‘window’ through which the operation of the process can be observed. However, the majority of existing process measurement systems have been designed to provide the input signal to a process control system, for which the primary requirement is a good average value. Typically, this is obtained by filtering out all the components of the sensor signal having frequencies greater than about 5 Hz, with the result that the ‘window’ is only translucent, rather than transparent, and much of the detailed information which is available at the interface between the process and the sensor is lost. If, on the other hand, the measurement system has a wide frequency response, then much of the information that is available at the interface between the process and the sensor can be gathered and analysed to provide a clear view of the operation and status of both the process and the measurement system. This concept involves two significant departures from the conventional approach to process measurements and to the utilization of measurement information. In the first place, the measurement systems should be designed to have the widest possible frequency response, so that the maximum information regarding the measured process parameter that is present at the interface between the process and the sensor is gathered. Secondly, instead of conditioning the primary sensor signal to provide a good average value for process control, it is analysed using well established techniques to provide not only the signal for process control but also information on which predictive maintenance procedures can be based. [ABSTRACT FROM AUTHOR]

• DC side faults detection of the solar photovoltaic power plants. • Using Student's T-Test for outlier detection in photovoltaic power plants. • String level fault detection in large scale solar power plants. • Comparison of actual and simulated solar power plant for fault detection. A new fault detection system is proposed in this study for large-scale grid-tied PV power plants. The fault detection system performs string level comparison of DC power of Actual PV Plant and a simulated PV plant, referred as Theoretical PV Plant. The comparison is performed with a statistical tool which distinguishes a faulty condition and identifies the nature of the fault. This statistical tool is used as outlier detector based on the William Gosset's (Student's) T-Test. The Theoretical PV Plant generates power in a simulation tool using inputs of irradiance and PV module temperature, both obtained from sensors at the Actual PV Plant. The string power of Actual PV Plant is also obtained in the simulation from the data logger in real-time environment. These string powers of Actual and Theoretical PV Plants are compared to find a faulty condition. A GUI is developed where the fault alarms appear on Real-time Status Monitor whenever a fault occurs in the Actual PV Plant. The proposed fault detection system has been validated on a 125 kWp grid-connected PV plant. [ABSTRACT FROM AUTHOR]