Failure Detection System Research Articles

Individual and Team Communication in a Dynamic Task

The purpose of this paper is to report the results of an experiment designed to study communication and its role in performance on a task shared between two human operators. Specifically, the experiment addressed the effects of system representations and communication on individual and team performance. As individuals, the operators had independent access to a computer interface that represented system information either in a graphic or an alphanumeric format. In the second phase of the research, operators were paired into teams to supervise the system. Two operators supervised the simulated fluid flow process, coordinating system control and failure detection responsibilities through their voice-mediated communication. Teams were comprised of operators using graphic, alphanumeric, or a combination of both graphic and alphanumeric information. Protocol analysis, specifically a think aloud methodology, was used in order to study the cognitive strategies used by individual operators to supervise the system, and to identify the types of communication that lead to the best performance. Performance comparisons were made between operators performing as individuals and in teams. Team communication was analyzed and related to system performance. Both the amount of communication and content of the communication was related to effective individual and team performance. Discussion focused on the application of these results to the design of advanced controls, operator interfaces, and operator communication.

On-Board Diagnostics of Vehicle Emission System Components: Review of Upcoming Government Regulation

This paper reviews the upcoming California Air Resources Board (CARB) requirements for on-board diagnostics of emissions related engine components. Its purpose is to disseminate this information to the controls community to stimulate new research. The CARB regulations are described and a literature review of possible solutions is presented. The state-of-the-art for current on-board engine diagnostics is reviewed and the desired characteristics of new on-board failure detection systems are outlined. This paper closes with a short review of advanced diagnostic techniques that show promise for resolving problems in this area.

Integration control and failure detection with application to the robot payload variation problem

The problem of integrated control and failure detection of robot payload variations during the execution of a task is investigated by incorporating failure diagnostic methods based on the novel 4-parameter control configuration. The combined controller and failure detection system is a complex design involving nontrivial combination of parameters related to two independent designs: the doubly coprime factorization-based compensator design and failure detection design. The combined design provides a generic setting to study fundamental tradeoffs present in a simultaneous control and diagnostic system. The PUMPA-560 robot arm is used as a case study for the combined design.

SATZAP: A Novel Approach to GPS Integrity*

Integrity continues to be a major issue preventing user acceptance of GPS. Substantial effort has therefore been invested in the pursuit of failure detection and warning system (FDWS) concepts to improve upon the existing integrity of the GPS positioning services. Notable progress has been made towards FDWS approaches, such as an independent GPS integrity channel (GIC) and receiver autonomous integrity monitoring (RAIM). This paper presents the conceptual outline of a new FDWS called “SATZAP.” It is a novel approach that takes maximum advantage of existing GPS resources. It utilizes the operational control system (OCS) and on-orbit Block II space vehicles (SVs) already built by the Department of Defense. Its user warning method is unique amongst FDWSs. Failed SVs would simply be turned “off” as far as the user is concerned. Turning an SV “off” within the integrity time limit is inherently compatible with all types of GPS receiver equipment; absolutely no user receiver modifications are required, and the warning cannot be misinterpreted.

Orthogonal series generalized likelihood ratio test for failure detection and isolation

A new failure detection and isolation algorithm for linear dynamic systems is presented. This algorithm, the Orthogonal Series Generalized Likelihood Ratio (OSGLR) test, is based on the assumption that the failure modes of interest can be represented by truncated series expansions. This assumption leads to a failure detection algorithm with several desirable properties. Computer simulation results are presented for the detection of the failures of actuators and sensors of a C-130 aircraft. The results show that the OSGLR test generally performs as well as the GLR test in terms of time to detect a failure and is more robust to failure mode uncertainty. However, the OSGLR test is also somewhat more sensitive to modeling errors than the GLR test.

A hierarchical multiple model adaptive control of discrete-time stochastic systems for sensor and actuator uncertainties

A hierarchical multiple model adaptive control (MMAC) is described for discrete-time stochastic systems with unknown sensor and actuator parameters, where the decentralized structure consists of a central processor and of m local processors which do not communicate between each other. A major assumption in this study is that the central and any local stations have different knowledge of the hypotheses on the unknown parameters. This leads to a flexible design algorithm for passively adaptive control strategies. Furthermore, the coordinator algorithm in evaluating the global a posteriori probability is relatively simple to implement. The result is applied to the design problem of an instrument failure detection and identification (FDI) system.

Development of fuel failure detection system for a high temperature gas cooled reactor. IV

For pt.III see ibid., vol.36, p.1261, (1989). A fuel failure detection (FFD) system using a wire-precipitator has been developed for a high-temperature gas-cooled reactor. For actual application of the FFD system, the response data to a precipitator for noble-gas fission products (FPs) released from coated particle fuels were measured under various in-core irradiating conditions. The dependency of precipitator counting rates on the fuel temperature and the reactor power was obtained by analyzing measured data. A model of the response function of a precipitator to the noble-gas FP nuclides was developed and its validity has been examined. By integrating these results, a state equation for the fuel failure detection can be determined. The response of a Ge gamma-ray spectrometer and a precipitator for a failed fuel sample was measured. The response curves of long-life nuclides are different from those of short-life nuclides. Using this difference, the fuel failure detection can be carried out.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

Authors

Authors

Use of neural networks for sensor failure detection in a control system

The use of the back-propagation neural network for sensor failure detection in process control systems is discussed. The back-propagation paradigm and traditional fault detection algorithms such as the finite integral squared-error method and the nearest-neighbor method are discussed. The algorithm is applied to the internal model control structure for a first-order linear time-invariant plant subject to high model uncertainty. Compared with traditional methods, the back-propagation technique is shown to be able to discern accurately the supercritical failures from their subcritical counterparts. The use of online adapted back-propagation fault detection systems in nonlinear plants is also investigated.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

Bayesian Approach to Optimal Input Design for Failure Detection and Diagnosis

This paper analyzes the sequential decision problem related to failure detection and diagnosis in dynamical systems when an input to the data generating mechanism is available to the decision-maker. The sequential decision rules are augmented by an input design rule. It is shown that some important properties of the decision rules related to the classical case (where no input is available) still hold: The terminal decision rule is a collection of fixed sample size Bayes rules and the problem when the sample size is not restricted can be approximated asymptotically.

「システム異常の検知技術」特集号の編集に寄せて

「システム異常の検知技術」特集号の編集に寄せて

A Decentralized Multiple Model Adaptive Filtering for Discrete-Time Stochastic Systems

A decentralized multiple model adaptive filter (MMAF) is proposed for linear discrete-time stochastic systems. The structure of decentralized multiple model studied here is based on introducing a global hypothesis for the global model and a local hypothesis for the local model, where it is assumed that the former hypothesis includes the latter one as a partial element. Algorithms for the decentralized MMAFs in unsteady and steady-state are derived using recent results in decentralized Kalman filtering. The results can be applied in designing a system for sensor failure detection and identification (FDI). An example is included to illustrate the characteristics of such a FDI system for the estimation of lateral dynamics of the hydrofoil boat.

Duality between failure detection and radar/optical maneuver detection

The two distinct technical specialty areas of failure detection in stochastic dynamical systems and the detection and tracking of target maneuvers by radar or optics are identified as two separate disciplines that have achieved results by using modeling techniques. These techniques can be cross-applied to the other specialty area, which is revealed to have in common the same underlying mathematical formulation of the even detection problem. >

Static Shape Estimation and Failure Detection in Large Spaceborne Antennas Via Computer Vision

The application of computer vision to static shape estimation and failure detection in large spaceborne antenna systems is explored with the aid of a physical model at the Ground Antenna Test Facility of the Jet Propulsion Laboratory (JPL), Pasadena, California. First, a mathematical description of various relationships between the viewed antenna components and their images is established. Then, algorithms for estimating the static shapes of the elastic supporting structure of the antenna are presented. The performance of these algorithms is determined experimentally using the JPL antenna model and a camera system with binary grey levels. An assessment of the limitations and practicality of this approach is given. The paper concludes with a discussion of the application of computer vision to detect and diagnose failures and abnormal conditions in spaceborne antenna systems.

Development of fuel failure detection system for a high temperature gas cooled reactor. III

Part II reference unavailable. A precision-type precipitator has been developed for high-sensitivity detection of fuel failure in a high-temperature gas-cooled reactor. The fuel failure detection (FFD) method is based on selective detection of short-life fission products to eliminate the background effect caused by long-life fission products in the primary circuit. Although this precipitator has a limited length of wire in the body equal to that of the conventional type, it can control the movement of the wire precisely, preventing the collection of fission products on the same part of the wire in each round. Unlike the conventional type, the precision precipitator can greatly reduce the background counts of the precipitation wire. This is very effective for selective detection of short-life fission products in the primary coolant helium gas. Tests showed background counts from 10% to below 0.4%. Therefore, an extremely slight change in the amount of fission products released from coated particles can be detected by the FFD system at a failure rate of 10/sup -5/.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

Expert system for sensor failure detection of aircraft

In this paper, an expert system for the sensor failure detection and isolation of flight test system is introduced. In order to detect and locate faults of an aircraft's sensors with sufficient robustness to parameter changes and noises, an accurate discrete-time model is presented and separated-bias algorithm is used for residual sequence generation. The expert system is used for the difficult task of failure isolation and flight decision making. The results of simulation and actual application show that the expert system for the FDI(failure detection and isolation) of flight test instruments can declare the faulty and locate the failures correctly. This expert system is suitable not only to the flight regime of low angle of attack but also to the flight regime of high angle of attack.

Fault simulation and test generation—an overview

A design or test engineer may proudly boast ‘I have a fault coverage of 96%’. What does he or she mean by that and how does it relate to yield in production or detection of system failures? The paper explains some basic concepts with regard to fault simulation and indicates the relationship between fault coverage and yield.

関数ベクトル空間手法による線形離散形動的システムの異常検出

関数ベクトル空間手法による線形離散形動的システムの異常検出

Effect of model uncertainty on failure detection: the threshold selector

The performance of all failure detection, isolation, and accommodation (DIA) algorithms is influenced by the presence of model uncertainty. The authors present a unique framework to incorporate a knowledge of modeling error in the analysis and design of failure detection systems. A concept is introduced called the threshold selector, which is a nonlinear inequality whose solution defines the set of detectable sensor failure signals. It identifies the optimal threshold to be used in innovations-based DIA algorithms. The optimal threshold is shown to be a function of the bound on modeling errors, the noise properties, the speed of DIA filters and the classes of reference and failure signals. The size of the smallest detectable failure is also determined. The results are applied to a multivariable turbofan jet engine example, which demonstrates improvements compared to previous studies.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

Failure detection in dynamic systems with modeling errors

This paper determines the theoretical limits of failure detectability in systems with modeling errors. Such errors are the main obstacle to accurate failure detection in aircraft, jet engines, power generation systems, and most other systems that include mechanical components and/or thermofluid processes. The paper then introduces the reachable measurement intervals (RMI) failure detection method that performs at the limit of detectability. RMI is capable of detecting the smallest theoretically detectable failures in systems with modeling errors, and is implementable in real time on microprocessors. The method is illustrated on a jet engine example. Nomenclature a — entry of the matrix A A = system matrix b = entry of the matrix B B = input matrix Bj = input matrix equal to the identity matrix C = output matrix C{ = output matrix for additional measurements D = direct transmission matrix H = Pontryagin's function K — matrix of gains for additional measurements L = vector of Lagrange multipliers m = entry of the matrix M M = matrix of Lagrange multipliers associated with uncertain entries of A n = entry of the matrix N N = matrix of Lagrange multipliers associated with uncertain entries of B p = discrete-time index corresponding to current time P = performance index for optimization q - vector of system parameters / = time t0 — current time ts = length of the processing window T - sampling time u = input vector x = state vector y = output vector z - vector of additional measurements Subscripts d = discrete e = uncertain and uncorrelated n = nominal v = uncertain