https://www2.econ.iastate.edu/tesfatsi/DeepLearningInNeuralNetworksOverview.JSchmidhuber2015.pdf

Deep learning in neural networks

A review on machinery diagnostics and prognostics implementing condition-based maintenance

This book is downloadable from http://www.irisa.fr/sisthem/kniga/. Many monitoring problems can be stated as the problem of detecting a change in the parameters of a static or dynamic stochastic system. The main goal of this book is to describe a unified framework for the design and the performance analysis of the algorithms for solving these change detection problems. Also the book contains the key mathematical background necessary for this purpose. Finally links with the analytical redundancy approach to fault detection in linear systems are established. We call abrupt change any change in the parameters of the system that occurs either instantaneously or at least very fast with respect to the sampling period of the measurements. Abrupt changes by no means refer to changes with large magnitude; on the contrary, in most applications the main problem is to detect small changes. Moreover, in some applications, the early warning of small - and not necessarily fast - changes is of crucial interest in order to avoid the economic or even catastrophic consequences that can result from an accumulation of such small changes. For example, small faults arising in the sensors of a navigation system can result, through the underlying integration, in serious errors in the estimated position of the plane. Another example is the early warning of small deviations from the normal operating conditions of an industrial process. The early detection of slight changes in the state of the process allows to plan in a more adequate manner the periods during which the process should be inspected and possibly repaired, and thus to reduce the exploitation costs.

Detection of abrupt changes: theory and application

There is an increasing demand for dynamic systems to become safer and more reliable This requirement extends beyond the normally accepted safety-critical systems such as nuclear reactors and aircraft, where safety is of paramount importance, to systems such as autonomous vehicles and process control systems where the system availability is vital It is clear that fault diagnosis is becoming an important subject in modern control theory and practice Robust Model-Based Fault Diagnosis for Dynamic Systems presents the subject of model-based fault diagnosis in a unified framework It contains many important topics and methods; however, total coverage and completeness is not the primary concern The book focuses on fundamental issues such as basic definitions, residual generation methods and the importance of robustness in model-based fault diagnosis approaches In this book, fault diagnosis concepts and methods are illustrated by either simple academic examples or practical applications The first two chapters are of tutorial value and provide a starting point for newcomers to this field The rest of the book presents the state of the art in model-based fault diagnosis by discussing many important robust approaches and their applications This will certainly appeal to experts in this field Robust Model-Based Fault Diagnosis for Dynamic Systems targets both newcomers who want to get into this subject, and experts who are concerned with fundamental issues and are also looking for inspiration for future research The book is useful for both researchers in academia and professional engineers in industry because both theory and applications are discussed Although this is a research monograph, it will be an important text for postgraduate research students world-wide The largest market, however, will be academics, libraries and practicing engineers and scientists throughout the world

Robust Model-Based Fault Diagnosis for Dynamic Systems

Fault diagnosis in dynamic systems using analytical and knowledge-based redundancy—a survey and some new results

The control of systems that involve friction presents interesting challenges. Recent research has focused on detailed modelling of friction phenomena in order to use robust on-line friction compensation procedures, attempting to cancel out the friction force effect in the feedback control of a mechanical or mechatronic system. However, the friction modelling problem remains a very difficult challenge and this article proposes a new approach to friction compensation which is based on the theory of robust fault estimation. The friction forces acting in a dynamic system can be viewed as actuator faults with time-varying characteristics to be estimated and compensated within an output feedback fault-tolerant control (FTC) scheme, so that the limitations arising from the use of a friction model are obviated. The friction (fault) estimation problem is hence embedded inside a control system with required stability, and performance robustness. This can be a significant advantage over well-known model-based friction compensation methods in which detailed modelling of friction phenomena is essential and for which robustness with respect to friction characteristics is difficult to achieve using non-linear models.

Friction compensation as a fault-tolerant control problem

The validity of applying frequency domain techniques to the identification of aircraft system dynamics is investigated using a simplified linear analytic model of a helicopter. A closed-loop identification of the helicopter model, with the longitudinal and lateral motions decoupled, is described. Measurement noise effects are included in the simulations. Input signals are applied to excite the system modes. Frequency response estimates are obtained by spectral analysis, and the coherence function is evaluated to indicate the likely accuracy. The transfer function coefficients are obtained by a nonlinear least-squares curve-fitting technique. With prior knowledge of the system order, the advantages of using a low peak-factor harmonic test signal for identification instead of the more usual frequency sweeps are demonstrated.

Comparison of test signals for aircraft frequency domain identification

© The Institution of Engineering and Technology 2017. This study proposes a fault estimation (FE)-based fault-tolerant control (FTC) strategy to maintain system reliability and achieve desirable control performance for a 3-DOF helicopter system with both actuator drift and oscillation faults and saturation. The effects of the faults and saturation are combined into a composite non-differentiable actuator fault function, which is approximated by a differentiable function and estimated together with the system state using a non-linear unknown input observer. An adaptive sliding mode controller based on the estimates is developed to compensate the effects of the faults and saturation. Taking into account the bi-directional robustness interactions between the FE and FTC functions, an integrated design approach is proposed to obtain the observer and controller gains in a single step, so as to achieve robust overall FTC system performance. In fault-free cases, the proposed strategy can be considered as a new approach for anti-windup control to compensate the effect of input saturation. Comparative simulations are provided to verify the effectiveness of the proposed design under different actuator fault scenarios.

https://ietresearch.onlinelibrary.wiley.com/doi/epdf/10.1049/iet-cta.2016.1602

Integrated fault-tolerant control for a 3-DOF helicopter with actuator faults and saturation

In this paper, a neuro-fuzzy and de-coupling fault diagnosis scheme (NFDFDS) is proposed for fault detection and isolation (FDI) of nonlinear dynamic systems. In this approach, powerful approximation and reasoning capabilities of neuro-fuzzy models are combined with the de-coupling capabilities of optimal observers to perform reliable fault detection and isolation. The neuro-fuzzy model presented here is a special form of Takagi–Sugeno (TS) fuzzy model used to represent the system by a fuzzy fusion of local linear sub-models. The necessary condition for the application of this FDI scheme is that this special form of the TS model can represent the nonlinear system, which is true for many practical systems. It is shown that if all the local models are stable and the corresponding local observers converge to the local models it can be expected that the global model is stable and the corresponding global observer will converge to the nonlinear input–output system. An application of FDI for an electro-pneumatic valve actuator in a sugar factory is presented. Key issues of finding a suitable structure for detecting and isolating nine realistic actuator faults are described. Copyright © 2004 John Wiley & Sons, Ltd.

Neuro-fuzzy uncertainty de-coupling: a multiple-model paradigm for fault detection and isolation

Tensor Product Distributed Compensation (TPDC) is a recently established controller design framework, that links TP model transformation and Parallel Distributed Compensation (PDC) framework. TP model transformation converts different models to a common representational form: the TP model form. The primary aim of this paper is to investigate the approximation capabilities of TP model forms, because the universal applicability of TPDC framework strongly relies on it. We point out that the set of functions that can be approximated arbitrarily well by TP forms with bounded number of components lies no-where dense in the set of continuous functions. Consequently, in a class of control problems this property necessitates the usage of tradeoff techniques between the accuracy and the complexity of the TP form, which is easily feasible within the TPDC framework unlike in analytic models.

Ron J. Patton

Papers

Friction compensation as a fault-tolerant control problem

Comparison of test signals for aircraft frequency domain identification

Integrated fault-tolerant control for a 3-DOF helicopter with actuator faults and saturation

Neuro-fuzzy uncertainty de-coupling: a multiple-model paradigm for fault detection and isolation

Approximation properties of TP model forms and its consequences to TPDC design framework