Analytical Redundancy Technique Research Articles

Fault Detection in Aircraft Flight Control Actuators Using Support Vector Machines

Future generations of flight control systems, such as those for unmanned autonomous vehicles (UAVs), are likely to be more adaptive and intelligent to cope with the extra safety and reliability requirements due to pilotless operations. An efficient fault detection and isolation (FDI) system is paramount and should be capable of monitoring the health status of an aircraft. Historically, hardware redundancy techniques have been used to detect faults. However, duplicating the actuators in an UAV is not ideal due to the high cost and large mass of additional components. Fortunately, aircraft actuator faults can also be detected using analytical redundancy techniques. In this study, a data-driven algorithm using Support Vector Machine (SVM) is designed. The aircraft actuator fault investigated is the loss-of-effectiveness (LOE) fault. The aim of the fault detection algorithm is to classify the feature vector data into a nominal or faulty class based on the health of the actuator. The results show that the SVM algorithm detects the LOE fault almost instantly, with an average accuracy of 99%.

Deep Learning-Aided Synthetic Airspeed Estimation of UAVs for Analytical Redundancy With a Temporal Convolutional Network

A synthetic air data system (SADS) is an analytical redundancy technique that is crucial for unmanned aerial vehicles (UAVs) and is used as a backup system during air data sensor failures. Unfortunately, the existing state-of-the-art approaches for SADS require GPS signals or high-fidelity dynamic UAV models. To address this problem, a novel synthetic airspeed estimation method that leverages deep learning and an unscented Kalman filter (UKF) for analytical redundancy is proposed. Our novel fusion-based method only requires an inertial measurement unit (IMU), elevator control input, and airflow angles while GPS, lift/drag coefficients, and complex aircraft dynamic models are not required. Additionally, we demonstrate that our proposed temporal convolutional network (TCN) is a more efficient model for airspeed estimation than the renowned models, such as ResNet or bidirectional long short-term memory (LSTM). Our deep learning-aided UKF was experimentally verified on long-duration real flight data and has promising performance compared with the state-of-the-art methods. In particular, it is confirmed that our proposed method robustly estimates the airspeed even under dynamic flight conditions where the performance of conventional methods is degraded.

Fault Monitoring and Accommodation of the Heat Exchanger Parameters of Triga-Mark II Nuclear Research Reactor using Model-Based Analytical Redundancy

One of the major challenges in instrumentation is to identify wrong data (signal) measurements and perform their validation. This can be done by regularly ensuring a correct operation of the different process components, particularly those having great importance for safety, in order to detect, isolate and identify any possible degradation or fault. This operation, known as on-line fault monitoring, should be done as early as possible, before any fault causes failure in equipment which can lead to the downtime of the plant and even to severe catastrophes and disasters. Therefore, these consequences influence negatively on productivity, availability and environment.At Triga-Mark II nuclear research reactor, the heat exchangers are provided for removing generated heat from the reactor pool water throw cooling circuits. Therefore, the monitoring of the evolution of certain thermo hydraulic parameters is necessary to ensure the safety of the reactor.Among several developed techniques, analytical redundancy has been recognized as an effective method for fault monitoring. It is the process of identifying a faulty instrument in a system through a comparison of its output to an estimate data. This estimation is based on the model and the measurements provided by the data acquisition chains of the existing sensors during all the operating modes of the installation. In our case, we are limited to mathematical models and Kalman filter approaches.In this paper we review the state of the fault monitoring and some model based analytical redundancy techniques for the heat exchanger, and present experimental results on their application to temperatures and flow rates of the cooling system of Triga-Mark II research reactor core.

Analytical Redundancy Design for Aeroengine Sensor Fault Diagnostics Based on SROS-ELM

Analytical redundancy technique is of great importance to guarantee the reliability and safety of aircraft engine system. In this paper, a machine learning based aeroengine sensor analytical redundancy technique is developed and verified through hardware-in-the-loop (HIL) simulation. The modified online sequential extreme learning machine, selective updating regularized online sequential extreme learning machine (SROS-ELM), is employed to train the model online and estimate sensor measurements. It selectively updates the output weights of neural networks according to the prediction accuracy and the norm of output weight vector, tackles the problems of singularity and ill-posedness by regularization, and adopts a dual activation function in the hidden nodes combing neural and wavelet theory to enhance prediction capability. The experimental results verify the good generalization performance of SROS-ELM and show that the developed analytical redundancy technique for aeroengine sensor fault diagnosis based on SROS-ELM is effective and feasible.

Predictive Fault Diagnosis System for Intelligent and Robust Health Monitoring

This paper describes work related to the Predictive Fault Diagnosis System for Intelligent and Robust Health Monitoring, which exists as a solution to complete failure detection, identification, and prognostics (FDI&P) in health monitoring applications. Several advanced FDI analytical redundancy techniques have been applied for such a purpose, with the most notable being a compound method comprised of optimal filtering, statistical analysis, and neuro-fuzzy algorithms that is able to detect and diagnose both known and unknown failures. Although this scheme has been proven to be quite successful for systems that can be well described in a state space representation, current research has shown the viability in extending the process to highly complex systems with considerable nonlinearities while still maintaining the FDI capabilities. This paper highlights the utility of these algorithms for determining failures in (1) a known reusable liquid rocket engine model and (2) an unknown input-output relation in a fluid flow testbed. Other research has focused on prognostic capabilities provided by a neural architecture enhanced with fuzzy logic using rule-based knowledge. An example of using data to construct fuzzy rules for determining the remaining useful life of components is provided to give insight into the process.

Fault detection and isolation of PEM fuel cell system based on nonlinear analytical redundancy

This paper presents a procedure dealing with the issue of fault detection and isolation (FDI) using nonlinear analytical redundancy (NLAR) technique applied in a proton exchange membrane (PEM) fuel cell system based on its mathematic model. The model is proposed and simplified into a five orders state space representation. The transient phenomena captured in the model include the compressor dynamics, the flow characteristics, mass and energy conservation and manifold fluidic mechanics. Nonlinear analytical residuals are generated based on the elimination of the unknown variables of the system by an extended parity space approach to detect and isolate actuator and sensor faults. Finally, numerical simulation results are given corresponding to a faults signature matrix.

Analysis of order of redundancy relation for robust actuator fault detection

This paper presents a new approach, called robust nonlinear analytic redundancy (RNLAR) technique to actuator fault detection for input-affine nonlinear multivariable dynamic systems. Robust fault detection is important because of the universal existence of model uncertainties and process disturbances in most systems. This paper characterizes the order of redundancy relation for nonlinear systems in terms of robustness. A theorem is proposed that state an increase in the order of redundancy relation increases the robustness in the sense of a performance index defined in this paper. Experimental results on a PUMA 560 robotic arm are presented to demonstrate the application of the theorem.

Robust nonlinear analytic redundancy for fault detection and isolation in mobile robot

A robust nonlinear analytical redundancy (RNLAR) technique is presented to detect and isolate actuator and sensor faults in a mobile robot. Both model-plant-mismatch (MPM) and process disturbance are considered during fault detection. The RNLAR is used to design primary residual vectors (PRV), which are highly sensitive to the faults and less sensitive to MPM and process disturbance, for sensor and actuator fault detection. The PRVs are then transformed into a set of structured residual vectors (SRV) for fault isolation. Experimental results on a Pioneer 3-DX mobile robot are presented to justify the efiectiveness of the RNLAR scheme.

Robust Fault Detection of a Robotic Manipulator

In this paper, a new robust fault detection technique for robotic manipulators is developed. The new approach, called robust nonlinear analytic redundancy (RNLAR) technique, detects both sensor and actuator faults in a robotic manipulator The proposed RNLAR technique can compensate for the effects of model-plan-mismatch (MPM) and process disturbance. A nonlinear primary residual vectors (PRV) design method to detect faults is proposed where the PRVs are highly sensitive to faults and less sensitive to MPM and process disturbance. Experimental results on a PUMA 560 are presented to justify the effectiveness of the RNLAR scheme.

ROBUST FAULT DETECTION AND ISOLATION IN MOBILE ROBOT

Robust nonlinear analytical redundancy (RNLAR) technique is used to detect and isolate actuator and sensor faults in a mobile robot. Both model-plant-mismatch (MPM) and process disturbance are considered during fault detection. The RNLAR is used to design primary residual vectors (PRV), which are highly sensitive to the faults and less sensitive to MPM and process disturbance, for sensor and actuator fault detection. The PRVs are transformed into set of structured residual vectors (SRV) for fault isolation. Experimental results on a Pioneer 3-DX are presented to justify the effectiveness of the RNLAR scheme.

Analytical Redundancy Techniques for Fault Detection in an Active Heavy Vehicle Suspension

This paper describes the design and implementation of an intelligent fault monitoring system for a prototype heavy vehicle active suspension system. A controller architecture has been designed, to facilitate safe start up and shutdown of the active control. This paper develops a method for monitoring the complete system, by dividing it into submodels which are more straightforward to analyse. The aim is to define a practical method for detecting faults, taking into account the nonlinearities in the real vehicle.

A Fault Diagnostics Algorithm for Differential Brake Control System

This paper presents a novel analytical redundancy technique for fault diagnostics of a differential brake control system. The proposed diagnostic algorithm builds upon an adaptive observer to generate residuals and identify faults. By judiciously utilizing the residual information, it is possible to detect faults at the component level rather than the system level. In addition, the adaptive observer is shown to effectively combat with the perennial nuisances of fault diagnostic algorithms, sensitivity to disturbance and false alarms. Hardware-in-the-loop simulation results show how reliably the proposed diagnostic algorithm detects faults in the differential brake control system, which is expected to play a crucial role in active vehicle safety.

Analytical redundancy relations for fault detection and isolation in algebraic dynamic systems

This paper extends the analytical redundancy techniques, well developed for linear systems, to FDI in non-linear dynamic systems modeled by polynomial differential algebraic equations. The residual evaluation form is developed for sensor, actuator and process component faults. Fault detectability and isolability conditions are given, and the design of robust structured residuals is addressed.

Analytical redundancy relations for fault detection and isolation in algebraic dynamic systems

This paper extends the analytical redundancy techniques, well developed for linear systems, to FDI in non-linear dynamic systems modeled by polynomial differential algebraic equations. The residual evaluation form is developed for sensor, actuator and process component faults. Fault detectability and isolability conditions are given, and the design of robust structured residuals is addressed.

Sensor Fault Diagnosis for an Internal Combustion Engine Based on Neural Networks

In this paper the authors propose an approach to sensor fault diagnosis for an automotive engine based on analytical redundancy techniques. The diagnosis architecture utilizes artificial neural network models for the estimation of engine variables that can be used in the detection and isolation of incipient faults. The considered neural network models have been tuned and tested by means of an experimental setup, These models have been used to implement a fault detection and diagnosis architecture based on a NPERG scheme (Nonlinear Parity Equation Residual Generation), which has been tested with encouraging results.

In-situ calibration of boiler instrumentation using analytic redundancy

This paper presents a broadly useful diagnostic methodology, which engineers and plant managers can use to detect system faults and find the in-situ operating characteristics of boilers when some metered data are either missing or obviously erroneous. The methodology is able to analyse conflicting measurements and utilize analytic redundancy (AR) to deduce the measurement or measurements that are substantially in error without shutting down the plant and recalibrating all instrumentation. The application of the AR technique in evaluating boiler performance is presented through a case study. Copyright © 2001 John Wiley & Sons, Ltd.

Fault Isolation in Hybrid Systems Combining Model Based Diagnosis and Signal Processing

Abstract Sensor-rich systems typically employ extensive signal processing techniques for fault detection and isolation tasks. Sensor-poor systems, on the other hand, require system models and analytical redundancy techniques to make diagnostic inferences. The increasing availability of inexpensive, batch-fabricated micro-controllers and MEMS sensors enables deployment of a multitude of sensors and microprocessors for control and diagnosis of embedded systems. We develop a diagnosis method that combines model-based diagnosis with signal processing techniques to address the challenges in diagnosing complex systems with hybrid discrete/continuous behaviors and to reduce the computational requirements by focusing the signal processing algorithms. We demonstrate the approach on problems in reprographic copier paper path diagnosis.

Design of Structured Residuals in Algebraic Dynamic Systems

This paper extends the analytic redundancy technique, widely developed for linear systems, to the FDI of non-linear dynamic systems modeled by polynomial differential algebraic equations.The existence of parity relations is briefly recalled and a definition of D-detectability is introduced. The paper focuses on the design of robust structured residuals.

Instrument fault detection and isolation: state of the art and new research trends

This paper presents the current state-of-the-art of residual generation techniques adopted in instrument fault detection and isolation. Both traditional and innovative methods are described with their advantages and their limits. The improvement of analytical redundancy technique performances for better dealing with high-dynamics systems and/or with online applications is pointed out as the most interesting need to focus the research efforts.

Combining expert system and analytical redundancy concepts for fault-tolerant flight control

A technique for rule-based fault-tolerant flight control is presented. The objective is to define methods for designing control systems capable of accommodating a wide range of aircraft failures, including sensor, control, and structural failures. A software architecture that integrates quantitative analytical redundancy techniques and heuristic expert system concepts for the purpose of in-flight, real-time fault tolerance is described. The resultant controller uses a rule-based expert system approach to transform the problem of failure accommodation task scheduling and selection into a problem of search. Control system performance under sensor and control failures using linear discrete-time deterministic simulations of a tandem-rotor helicopter's dynamics is demonstrated. It is found that the rule-based control technique enhances existing redundancy management systems, providing smooth integration of symbolic and numeric computation, a search-based decision-making mechanism, straightforward system organization and debugging, an incremental growth capability, and inherent parallelism for computational speed.