LPV model-based robust diagnosis of flight actuator faults

A linear parameter-varying (LPV) model-based synthesis, tuning and assessment methodology is developed and applied for the design of a robust fault detection and diagnosis (FDD) system for several types of flight actuator faults such as jamming, runaway, oscillatory failure, or loss of efficiency. The robust fault detection is achieved by using a synthesis approach based on an accurate approximation of the nonlinear actuator–control surface dynamics via an LPV model and an optimal tuning of the free parameters of the FDD system using multi-objective optimization techniques. Real-time signal processing is employed for identification of different fault types. The assessment of the FDD system robustness has been performed using both standard Monte-Carlo methods as well as advanced worst-case search based optimization-driven robustness analysis. A supplementary industrial validation performed on the AIRBUS actuator test bench for the monitoring of jamming, confirmed the satisfactory performance of the FDD system in a true industrial setting.     

Multi-scale kernel Fisher discriminant analysis with adaptive neuro-fuzzy inference system (ANFIS) in fault detection and diagnosis framework for chemical process systems

Fault detection and diagnosis (FDD) framework is one of safety aspects that is important to the industrial sector to ensure its high-quality production and processes. However, the development of FDD system in chemical process systems could have difficulties, e.g. highly nonlinear correlation within the variables, highly complex process, and an enormous number of sensors to be monitored. These issues have encouraged the development of various approaches to increase the effectiveness and robustness of the FDD framework, such as the wavelet transform analysis, where it has the advantage in extracting the significant features in both time and frequency domain. It has motivated us to propose an extension work of the multi-scale KFDA method, where we have modified it with the implementation of Parseval’s theorem and the application of ANFIS method to improve the performance of the fault classification. In this work, through the implementation of Parseval’s theorem, the observation of fault features via the energy spectrum and effective reduction in DWT analysis data quantity can be accomplished. The extracted features from the multi-scale KFDA method are used for fault diagnosis and classification, where multiple ANFIS models were developed for each designated fault pattern to increase the classification accuracy and reduce the diagnosis error rate. The fault classification performance of the proposed framework has been evaluated using a benchmarked Tennessee Eastman process. The results indicated that the proposed multi-scale KFDA-ANFIS framework has shown the improvement with an average of 87.02% in classification accuracy over the multi-scale PCA-ANFIS (78.90%) and FDA-ANFIS (70.80%).

     

Fault diagnosis for heat pumps with parameter identification and clustering

For reducing the energy consumption of heat pumps, fault detection and diagnosis (FDD) is fundamental. The FDD system presented is based on a gray-box process model, the parameters of which are identified online. The faults are classified from the parameters using clustering methods. Known clustering techniques have been simplified and new “vector clustering” techniques have been developed for classifying gradual faults. The FDD system has been tested in various real applications, for one of which the results are presented in this work. The contribution lies on the application side with a software tool developed for the fully automated training process.     

Fault detection and diagnosis using statistical control charts and artificial neural networks

In order to operate a successful plant or process, continuous improvement must be made in the areas of safety, quality and reliability. Central to this continuous improvement is the early or proactive detection and correct diagnosis of process faults. This research examines the feasibility of using cumulative summation (CUSUM) control charts and artificial neural networks together for fault detection and diagnosis (FDD). The proposed FDD strategy was tested on a model of the heat transport system of a CANDU nuclear reactor.The results of the investigation indicate that a FDD system using CUSUM control charts and a radial basis function (RBF) neural network is not only feasible but also of promising potential. The control charts and neural network are linked by using a characteristic fault signature pattern for each fault which is to be detected and diagnosed. When tested, the system was able to eliminate all false alarms at steady state, promptly detect six fault conditions, and correctly diagnose five out of the six faults. The diagnosis for the sixth fault was inconclusive.     

基于测试性的电子系统综合诊断与故障预测方法综述

故障诊断与预测技术是故障预测与健康管理(PHM)中的两大关键技术.依据电子系统的故障模式与机理,结合测试性设计分析理论,提出了一种基于测试性的电子系统综合诊断与故障预测方法框架.对国内外综合诊断与故障预测方法进行了分类与总结,从基于测试性的嵌入式诊断、基于信号处理的智能故障诊断、基于测试性的故障预测3个方面论述了电子系统综合诊断与故障预测方法.最后分析了制约电子系统综合诊断与故障预测的因素,并探讨了未来的发展趋势.     

High-speed train fault detection with unsupervised causality-based feature extraction methods

With the development of smart sensors, large amount of operating data collected from a complex system as a high-speed train providing opportunities in efficient and effective fault detection and diagnosis (FDD). The data brings also challenges in the FDD modelling process, since the various signals may be redundant, useless and noisy for the FDD modelling of a specific sub-system. The data-driven methods suffer also from the curse of dimensionality. Feature dimension reduction can reduce the dimension of the monitoring dataset and eliminate the useless information. Different from the classical methods based on the correlation among variables, recent studies have shown that causality-based methods can make the FDD model more explanatory and robust. From the adjacency matrix of the causal network diagram, three unsupervised causality-based feature extraction methods for FDD in the braking system of a high-speed train are proposed in this paper. By constructing the causal network diagram among the raw monitoring feature variables through the causal discovery algorithm, the proposed methods extract informative features based on the causal adjacency matrix or the full causal adjacency matrix proposed in this work. These methods are adopted for fault detection with real dataset collected from the braking system in a high-speed train to verify their effectiveness. The experimental results show that the proposed causality-based feature extraction methods are effective and have certain advantages in comparison with the classical correlation-based methods. Especially, the feature extraction method based on the correlation matrix constructed from full causal adjacency matrix achieves better and stable results than the benchmark methods in the experiment.     

Integration of fault diagnosis and control based on a trade-off between fault detectability and closed loop performance

This paper presents a novel methodology for simultaneous optimal tuning of a fault detection and diagnosis (FDD) algorithm and a feedback controller for a chemical plant in the presence of stochastic parametric faults. The key idea is to propagate the effect of time invariant stochastic uncertainties onto the measured variables by using a Generalized Polynomial Chaos (gPC) expansion and the nonlinear first principles’ model of the process. A bi-level optimization is proposed for achieving a trade-off between the fault detectability and the closed loop process variability. The goal of the outer level optimization is to seek a trade-off between the efficiency of detecting a fault and the closed loop performance, while the inner level optimization is designed to optimally calibrate the FDD algorithm. The proposed method is illustrated by a continuous stirred tank reactor (CSTR) system with a fault consisting of stochastic and intermittent variations in the inlet concentration. Beyond achieving improved trade-offs between fault detectability and control, it is shown that the computational cost of the gPC model based method is lower than the Monte Carlo type sampling based approaches, thus demonstrating the potential of the gPC method for dealing with large problems and real-time applications.     

测控装备面向协同智能诊断系统应用研究

随着测控装备集成度的提高,单一的诊断模式已很难准确定位装备出现的故障,需要多个领域知识共同对诊断对象进行决策融合.本文根据某型测控装备实际,设计了以故障树诊断为基础的专家系统,结合博弈论相关理论对多个知识库进行诊断结果的融合.最终将诊断结果纳入装备的日常维护保养,提高了装备的综合保障能力.     

未知环境中移动机器人故障诊断与容错控制技术综述

以我国月球探测为研究背景,以轮式移动机器人为研究对象,介绍了在外星球表面等未知环境中进行深空探测的移动机器人的故障模型和传感器误差模型,分析了未知环境中移动机器人故障诊断与容错控制的特点.在此基础上综述了国内外在该领域的研究进展和主要方法,包括基于多模型的方法、基于粒子滤波器的方法、基于传感器信息融合的方法以及层次容错结构等. 最后,总结了该领域待解决的几个难点问题,并对该研究领域的发展趋势进行了展望.     

Fault detection and diagnosis of an industrial steam turbine using a distributed configuration of adaptive neuro-fuzzy inference systems

The issue of fault detection and diagnosis (FDD) has gained widespread industrial interest in process condition monitoring applications. An innovative data-driven FDD methodology has been presented in this paper on the basis of a distributed configuration of three adaptive neuro-fuzzy inference system (ANFIS) classifiers for an industrial 440 MW power plant steam turbine with once-through Benson type boiler. Each ANFIS classifier has been developed for a dedicated category of four steam turbine faults. A preliminary set of conceptual and experimental studies has been conducted to realize such fault categorization scheme. A proper selection of four measured variables has been configured to feed each ANFIS classifier with the most influential diagnostic information. This consequently leads to a simple distributed FDD system, facilitating the training and testing phases and yet prevents operational deficiency due to possible cross-correlated measured data effects. A diverse set of test scenarios has been carried out to illustrate the successful diagnostic performances of the proposed FDD system against 12 major faults under challenging noise corrupted measurements and data deformation corresponding to a specific fault time history pattern.     

Delay-dependent fault detection and diagnosis using B-spline neural networks and nonlinear filters for time-delay stochastic systems

A new fault detection and diagnosis (FDD) scheme is studied in this paper for the continuous-time stochastic dynamic systems with time delays, where the available information for the FDD is the input and the measured output probability density functions (PDFs) of the system. The square-root B-spline neural networks is used to formulate the output PDFs with the dynamic weightings. As a result, the concerned FDD problem can be transformed into a robust FDD problem subjected to a continuous time uncertain nonlinear system with time delays. Delay-dependent criteria to detect and diagnose the system fault are provided by using linear matrix inequality (LMI) techniques. It is shown that this new criterion can provide higher sensitivity performance than the existing result. Simulations are given to demonstrate the efficiency of the proposed approach.     

旋翼飞行机器人故障诊断与容错控制技术综述

对故障诊断和容错技术的发展过程进行了简要概述，以旋翼飞行机器人为研究对象，在分析了旋翼飞行机器人故障诊断与容错控制特点的基础上，介绍了当前国内外在该领域的研究进展和主要方法．最后，总结了该领域待解决的难点问题，并指出了该研究领域的发展趋势．     

基于迭代学习的离散线性时变系统故障诊断

针对一类离散线性时变系统的故障诊断问题,提出一种新的故障检测与估计算法.该算法通过引入虚拟故障构建离散故障跟踪估计器,在选取的优化时域内,利用估计器输出和系统实际输出产生的残差信号,采用迭代学习算法来调节虚拟故障,使虚拟故障逼近系统中实际发生的故障,从而达到对系统故障诊断的目的.该方法不仅能检测出系统不同类型的故障,还可以实现对故障信号的精确估计.仿真结果验证了所提出方法的有效性.     

Bearing fault diagnosis based on combined multi-scale weighted entropy morphological filtering and bi-LSTM

With the development of industry and technology, mechanical systems’ safety has strong relations with the diagnosis of bearing faults. Accurate fault diagnosis is essential for the safe and stable operation of rotating machinery. Most former research depends too much on the fault signal specificity and learning model’s choices. To overcome the disadvantages of lacking intrinsic mode function (IMF) modal aliasing, low degree of discrimination between data of different fault types, high computational complexity. This paper proposes a method that combines multi-scale weighted entropy morphological filtering (MWEMF) signal processing and bidirectional long-short term memory neural networks (Bi-LSTM). The developed rolling bearing fault diagnosis strategy is then implemented to different databases and potential models to demonstrate the greatly improved system’s ability to reconstruct the time-to-frequency domain characteristics of fault signature signals and reduce learning cost. After verification, the classification accuracy of the proposed model reaches 99%.

     

Development and application of sliding mode LPV fault reconstruction schemes for the ADDSAFE Benchmark

This paper describes the development and the evaluation of a robust sliding mode observer fault detection scheme applied to an aircraft benchmark problem as part of the ADDSAFE project. The ADDSAFE benchmark problem which is considered in this paper is the yaw rate sensor fault scenario. A robust sliding mode sensor fault reconstruction scheme based on an LPV model is presented, where the fault reconstruction signal is obtained from the so-called equivalent output error injection signal associated with the observer. The development process includes implementing the design using AIRBUS׳s the so-called SAO library which allows the automatic generation of flight certifiable code which can be implemented on the actual flight control computer. The proposed scheme has been subjected to various tests and evaluations on the Functional Engineering Simulator conducted by the industrial partners associated with the ADDSAFE project. These were designed to cover a wide range of the flight envelope, specific challenging manoeuvres and realistic fault types. The detection and isolation logic together with a statistical assessment of the FDD schemes are also presented. Simulation results from various levels of FDD developments (from tuning, testing and industrial evaluation) show consistently good results and fast detection times.     

Experimental Test of a Two-Stage Kalman Filter for Actuator Fault Detection and Diagnosis of an Unmanned Quadrotor Helicopter

This paper addresses the problem of Faut Detection and Diagnosis (FDD) of a quadrotor helicopter system in the presence of actuator faults. To this end a Two-Stage Kalman Filter (TSKF) is used to simultaneously estimate and isolate possible faults in each actuator. The faults are modelled as losses in control effectiveness of rotors. Three fault scenarios are investigated: loss of control effectiveness in one single actuator, simultaneous loss of control effectiveness in all motors, and loss of control effectiveness in three motors with different magnitudes. The developed FDD algorithm is evaluated through experimental application to an unmanned quadrotor helicopter testbed available at the Department of Mechanical and Industrial Engineering of Concordia University, called Qball-X4. The obtained results show the effectiveness of the proposed FDD method.     

Actuator and Sensor Fault Detection and Failure Prediction for Systems with Multi-dimensional Nonlinear Partial Differential Equations

This paper presents a new model-based fault detection and failure prediction framework for a class of multi-input and multi-output (MIMO) nonlinear distributed parameter systems (DPS) described by partial differential equations (PDE) with actuator and sensor faults. The fault functions cover both abrupt and incipient faults. A Luenberger type observer is used to monitor the health of the DPS as a detection observer on the basis of the nonlinear PDE representation of the system and by utilizing only the measured output vector. By taking the difference between measured and estimated outputs, a residual signal is generated for fault detection. If the detection residual exceeds a predefined threshold, a fault is claimed to be active. Once an actuator or a sensor fault is detected, an appropriate fault parameter update law is developed to learn the fault dynamics online with the help of an additional measurement. Later, an explicit formula is introduced to estimate the time-to-failure in the presence of an actuator/sensor fault by utilizing the limiting values of the output vector along with the estimated fault parameter vector. Eventually, the effectiveness of the proposed detection and prediction framework is demonstrated on a nonlinear process.

     

基于不确定性的故障预测方法综述

故障预测是实现视情维修策略的基础. 不确定性问题在故障预测中普遍存在, 对此, 总结了基于不确定性的故障预测方法的关键问题, 并以不确定性属性的特点将现有故障预测方法分为基于随机性、模糊性、灰性及混合不确定性等4 类. 综述了各类方法的研究现状与不足, 并展望了基于不确定性的故障预测方法的发展趋势, 探讨了基于区间不确定性的故障预测方法的可行性.

     

Fuzzy model-based fault detection and diagnosis for a pilot heat exchanger

This article addresses the design and real-time implementation of a fuzzy model-based fault detection and diagnosis (FDD) system for a pilot co-current heat exchanger. The design method is based on a three-step procedure which involves the identification of data-driven fuzzy rule-based models, the design of a fuzzy residual generator and the evaluation of the residuals for fault diagnosis using statistical tests. The fuzzy FDD mechanism has been implemented and validated on the real co-current heat exchanger, and has been proven to be efficient in detecting and isolating process, sensor and actuator faults.     

Discrete Event System Framework for Fault Diagnosis with Measurement Inconsistency: Case Study of Rogue DHCP Attack

Fault detection and diagnosis (FDD) facilitates reliable operation of systems. Various approaches have been proposed for FDD like Analytical redundancy (AR), Principal component analysis (PCA), Discrete event system (DES) model etc., in the literature. Performance of FDD schemes greatly depends on accuracy of the sensors which measure the system parameters. Due to various reasons like faults, communication errors etc., sensors may occasionally miss or report erroneous values of some system parameters to FDD engine, resulting in measurement inconsistency of these parameters. Schemes like AR, PCA etc., have mechanisms to handle measurement inconsistency, however, they are computationally heavy. DES based FDD techniques are widely used because of computational simplicity, but they cannot handle measurement inconsistency efficiently. Existing DES based schemes do not use Measurement inconsistent (MI) parameters for FDD. These parameters are not permanently unmeasurable or erroneous, so ignoring them may lead to weak diagnosis. To address this issue, we propose a Measurement inconsistent discrete event system (MIDES) framework, which uses MI parameters for FDD at the instances they are measured by the sensors. Otherwise, when they are unmeasurable or erroneously reported, the MIDES invokes an estimator diagnoser that predicts the state(s) the system is expected to be in, using the subsequent parameters measured by the other sensors. The efficacy of the proposed method is illustrated using a pumpvalve system. In addition, an MIDES based intrusion detection system has been developed for detection of rogue dynamic host configuration protocol (DHCP) server attack by mapping the attack to a fault in the DES framework.