In recent years, autoencoder has been widely used for the fault diagnosis of mechanical equipment because its excellent performance in feature extraction and dimension reduction; however, original only limited ability due to lack label information. To solve this issue, study proposes a distance stack (FD-SAE) rolling bearing diagnosis. Compared with existing methods, FD-SAE stronger faster network convergence speed. By analyzing characteristics data, it is found that there are evident...

The paper deals with the robust fault detection problem for a class of discrete-time linear Markovian jump systems an unknown input. By using general observer-based filter as residual generator, design is formulated H∞-filtering problem, in which matrices are dependent on system mode, i.e. generator well. main objective to make error between and (or, more generally, weighted fault) small possible. A sufficient condition solve this established terms feasibility certain matrix inequalities...

This technical note deals with the problem of finite horizon H <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">∞</sub> fault detection for linear discrete time-varying systems. Using an observer-based filter (FDF) residual feedback as a generator, new formulation is first presented. Then design -FDF converted into minimum certain quadratic form. Through building relationship Krein space projection and applying innovation analysis, sufficient...

With the fast development of science and industrial technologies, fault diagnosis identification has become a crucial technique for most applications. To ensure system safety reliability, many conventional model based methods have been proposed. However, with increase in complexity uncertainty engineering system, it is not feasible to establish accurate mathematical models time. Rotating machinery, due its mechanical structure transmission mechanics, within this category. Thus, data-driven...

An improved algorithm with feature selection and neural network classification is proposed in this paper to investigate the fault diagnosis problem of rotating machinery. The vectors are constructed by extracting time- frequency-domain characteristics overall machine under multiple operating conditions. To strengthen diagnostic ability, an sensitive improving distance evaluation (DE) method reconstructing a low-dimensional sample selectively chosen parameters from multidimensional vectors....

Fault detection (FD) plays an important role in guaranteeing system safety and reliability for unmanned aerial vehicles (UAVs). This paper focuses on developing alternative approach to FD airspeed sensor UAVs by using data from gyros, accelerometers, global positioning system, wind vanes. Based the kinematics model of UAV, estimator is proposed provide analytical redundancy information above-mentioned sensors, which are commonly implemented UAVs. filter process independent measurement...

This paper studies the problem of fault detection for linear discrete time-varying systems with multiplicative noise in finite-horizon, where our main object is to provide an optimal filter (FDF) design scheme such that stochastic sensitivity/robustness ratio diagnosis maximized sense probability 1. An operator-aided optimization approach and a generalized FDF are proposed solutions issues derived operator forms. The relationships among deduced explicitly revealed via methodology. parameter...

In this article, a new event-triggered H <sub xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">i</sub> /H xmlns:xlink="http://www.w3.org/1999/xlink">∞</sub> optimization fault detection (FD) scheme is proposed for linear discrete time systems. A system with varying sampling periods first presented to describe the dynamics of an system. Based on this, observer-based filter (FDF) constructed as residual generator so that generated signal completely...

Event-triggered fault diagnosis has attracted tremendous research attention in the last decade due to its superiority improving utilisation efficiency of communication resources. Different from traditional works time-driven, event-triggered schemes are used determine whether current measurement output should be released detection filter, while sensor data not satisfying a predefined triggering condition will discarded directly. As such, on been challenging issue and many outstanding results...

Fault diagnosis is the key technology to guarantee reliability and safety of satellite attitude control systems. Although model-based methods have achieved good fault performance, various factors (such as closed-loop design, model nonlinearity, external disturbances) in systems still bring challenges isolation. Meanwhile, data-driven can assist for based on residual signals (generated from methods) input/output data relieve difficulty. Based above motivations, a novel dual-driven approach...

The robust fault detection problems for time-delay LTI systems with unknown inputs are studied. basic idea of our study is to express the robustness as well sensitivity residual signals faults in terms L/sub 2-/gain and, based on it, formulate design filters an optimization problem. Applying H/sub /spl infin// technique, LMI approach solution problem then proposed. main results consist derivation existence conditions filter and further optimal

This note considers the problem of exponential stability for time-delay systems. In terms linear matrix inequalities, a new sufficient condition is obtained. Based on this, an improved upper bound decay rate can be easily calculated. When time-varying norm-bounded parameter uncertainties appear, robust uncertain systems also provided. The reduced conservatism proposed conditions illustrated via two numerical examples.