Failure Diagnosis System Research Articles

Timing and predictors of death in pediatric patients with multiple organ system failure.

To describe the timing of onset of organ system failure, multiple organ system failure diagnosis, and the subsequent death in children admitted to a pediatric intensive care unit (ICU). Second, to identify independent risk markers of death in pediatric patients with multiple organ system failure. Review of a database. Pediatric ICU within a tertiary care center. We analyzed the pediatric ICU course of 777 consecutive patients aged < 18 yrs. Eighty-five (10.9%) of 777 children had multiple organ system failure, defined as the simultaneous occurrence of at least two organ system failures. Of 85 children, 37 (43.5%) were postoperative cardiac surgery patients and 48 (56.5%) patients were in the ICU for other reasons. The diagnostic criteria for multiple organ system failure were met on the day of admission by 73 (86%) of 85 patients. The maximum number of organ system failures occurred within 72 hrs in 74 (87%) children. The mortality rate for all patients with multiple organ system failure was 50.6%. Thirty-eight (88.4%) of deaths occurred within 7 days after the diagnosis of multiple organ system failure. Survival analysis was comparable for both postoperative cardiac surgery patients and patients with other diagnoses. Multivariate analysis identified three factors as independent risk markers of death in pediatric patients with multiple organ system failure: maximum number of simultaneous organ system failures during the pediatric ICU stay: odds ratio, 55.9 (95% confidence interval, 7.9 to 396.1); age < or = 12 months: odds ratio, 17.1 (95% confidence interval, 1.8 to 158.7); and the Pediatric Risk of Mortality (PRISM) score on the day of admission: odds ratio, 1.25 (95% confidence interval, 1.1 to 1.5). The mortality rate associated with multiple organ system failure in pediatric patients is high. The maximum number of simultaneous organ system failures during pediatric ICU stay, age < or = 12 months, and the PRISM score on the day of admission are independent risk markers of death. Diagnosis of multiple organ system failures, development of maximum number of organ system failures, and deaths occur remarkably early after pediatric ICU admission; the rationale for using prophylactic therapy under such circumstances is unclear.

Fault diagnosis and computer integrated manufacturing systems

This paper examines one aspect of human interaction with computer-integrated systems, that of fault diagnosis or troubleshooting. The complexity (and attendant unreliability) of the new manufacturing systems has meant that fault diagnosis has become an increasing proportion and an integral part of operators' jobs. Establishing and maintaining high levels of diagnostic accuracy and efficiency is important for a variety of reasons. One is that equipment downtime is expensive, another is that errors of diagnosis might, under some circumstances, be a threat to both safety and quality. As yet there has been little research into human fault diagnosis in CIM. This paper draws on research from other industries, such as power generation and chemical production and explores the application of that work to fault-finding in computer-integrated discrete production systems. The focus is on the factors that influence individual fault-finding behaviors and efficiencies. Where possible, the aim is to identify critical human-machine interaction design principles and to highlight research questions yet to be addressed in contemporary manufacturing environments, with respect to the diagnosis of system failures. >

Tool failure diagnosis in milling using a neural network

Due to the complex machining processes, it is difficult to detect tool failure in unmanned machining successfully. In order to solve this important problem, this paper describes a new microcomputer-based diagnosis system for monitoring tool failure in milling operations. A multi-layer feed-forward neural network with the back-propagation algorithm is introduced for processing the characteristic patterns of the cutting force signal in milling operations. The neural network with learning ability can recognise tool failure even under varying cutting conditions. Experimental results are presented to verify the feasibility of this intelligent tool failure diagnosis system in milling operations.

A fault dictionary based expert system for failure diagnosis in a multiple-PCB environment

This paper describes a failure-diagnosis expert system for printed circuit boards which exploits functional test data. The data is output from automatic test equipment which is used to test every board subsequent to manufacture. The use of a conventional machine-learning technique produced only limited success due to the very large search space of failure reports. This also ruled out the use of a conventional knowledge-based approach. In addition, there was a requirement to track changes in printed circuit board design and manufacture which also ruled out a conventional knowledge-based approach. The investigations led to the development of a system which tracks changes by learning in a more restricted search space derived from the original space of reports. The system performs a diagnosis by matching a failure report with information about previously seen reports. The matching rules used are heuristic. The system uses basic circuit connectivity information in conjunction with the matching procedure to improve diagnostic performance, especially in cases where matching fails to identify a unique component.

Failure Diagnostic System for Air-Operated Control Valves by Neural Network. (2nd Report).

Our objective is to diagnose the failure of actuators of air-operated control valves that are frequently used in process-control plants. In our first report, a prototype failure diagnosis system using neural network was developed and its recognition capability was discussed. Some failures were difficult to diagnose because their characteristics were very similar to each other or to the normal state. This paper deals with the fuzzy network (the "teacher" data is determined by fuzzy clustering)used to improve recognition capability and the selection of effective sensors by means of the multivariate analysis method. The amount of square errors of the output from the fuzzy network can be considerably reduced for the failures difficult to diagnose with the network with the binary digit "teacher" data. Three effective sensors were selected out of more than 10 sensors, and a new network system utilizing these three sensor signals was developed. The recognition Capability of this network is compared with that of the network utilizing full sensor signals. Its recognition capability is good, even though the number of connections is reduced by approximately 60%.

Failure Diagnostic System on Air-Operated Control Valves by Neural Network.

A prototype failure diagnosis system has been developed using neural network technology for the actuator of air-operated valves. Because actual failure data were not easily available, the data of 30 failure patterns were experimentally obtained using more than 10 sensors. The time series of sensor signals are Fourier transformed. The data of the magnitude spectrum, phase difference and others are used as the characteristic parameters in our failure diagnosis. From the data, appropriate information for use in the failure diagnosis was extracted. Furthermore, similarities among the failure characteristics were found by fuzzy clustering and statistical analysis. The new system which we developed consists of plural subnetworks and one main network. Each subnetwork is related to one specific sensor signal, and deals with the magnitude spectra from the sensor signal. The main network makes the final decision according to the outputs from the subnetworks and other data. In our system, the number of network connections can be reduced by approximately 40% without degradation of the recognition capability in comparison with the conventional system which uses only one neural network.

Development of an Expert System for Failure Diagnosis of Primary Side Systems

In this paper a prototype expert system, the NSSS-DS, is described. It is used for the diagnosis of three main systems-the rod control system, the reactor coolant pumps (RCPs), and the pressurizer-in the primary system of the Kori-2 nuclear power plant. This system diagnoses system malfunctions quickly and offers appropriate guidance to operators. The method of inference applied to this system is rule-based deduction with certainty factor operation. The diagnostic symptoms include alarms, indication lamps, parameter values, and valve lineup that can be acquired at the main control room. The application of this system to the rod control system, mainly consisting of electrical components, and the RCPs, mainly consisting of mechanical components, is described from the point of view of diagnostic strategies.

Geometric tools for an observer-based approach to residual generation

This paper emphasizes on the basic geometric tools necessary for the design of a failure diagnosis system, where the diagnosis is based on the residual generation by using elementary observers. In order to improve the sensitivity of the residuals, various methodologies are proposed. A realization over a physical example is presented.

An efficient algorithm for parametric fault simulation of monolithic IC's

An efficient methodology for performing fault simulation experiments as part of an IC failure diagnosis system is described. The methodology uses regression models that relate IC performance directly to process disturbances inherent in all IC fabrication processes. An efficient algorithm for establishing the structure of such models based on data obtained from coupled process and circuit simulators is developed. It selects a minimum number of significant input parameters and automatically establishes an optimum order polynomial regression model. Moreover. the regression models obtained contain information about the structure of the dependence of IC performances on the process disturbances, which can be coded in a dependency tree for each IC performance and used directly for diagnostic purposes. The algorithm belongs to the class of artificial neural network approaches and has been implemented in a program called MULREG (multilayer regression). Two examples of analyzing MULREG are given. >

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.

人工知能技法による舶用ディーゼル主機関起動時の故障診断システム

This system has been developed as a subsystem of the diesel engine propulsion plant of a super-automated ship. Many conventional diagnostic expert systems involve a variety of difficult problems. A study to solve these problems has produced a failure diagnostic system applying composite/structure design, which has been used to develop traditional procedure type programming. The paper describes an approach to develop the failure diagnosis system based on a Hitachi Zosen B & W low speed marine diesel engine's startup.

A Hierarchical Expert System for Failure Diagnosis in Power Plants

A hierarchical, modular real-time failure diagnosis system for technical prosesses is presented. The symptom detecting modules are written in algorithmic form and the methods used in our application are model based, but e.g. vibration analysis could also be used. The decision-making part is realized in a knowledge-based system module. The experimental system is realized with microcomputers. The lower level algorithmic modules run on a COMPAQ® 386 microcomputer under GENESISTM, a process control software. The decision-making part is done in OPS-83 language and run on a separate microcomputer. An application to failure diagnosis in a coal power plant and some results are presented.

A Computerized Expert System for Tribological Failure Diagnosis

A micro-computer supported, rule-based expert system for failure diagnosis of tribo-components is described. Included is a brief review of AI tools and system organization methods used. The system, as implemented, encompasses: (a) observation of the machine or device containing the tribo-components; (b) identification of component failure modes by visual appearance comparison with photographic standards, and (c) reverse-engineering of the component selection for the application. Not implemented, but defined are: laboratory and advanced analytic diagnostic methods. The system consists of interactive screen-dialogs, augmented by engineering computation routines and followed by the generation of formatted diagnostic and corrective action reports.

Neural computing for numeric-to-symbolic conversion in control systems

A type of neural network, the multilayer perceptron, is used to classify numeric data and assign appropriate symbols to various classes. This numeric-to-symbolic conversion results in a type of information extraction, which is similar to what is called data reduction in pattern recognition. The use of the neural network as a numeric-to-symbolic converter is introduced, its application in autonomous control is discussed, and several applications are studied. The perceptron is used as a numeric-to-symbolic converter for a discrete-event system controller supervising a continuous variable dynamic system. It is also shown how the perceptron can implement fault trees, which provide useful information (alarms) in a biological system and information for failure diagnosis and control purposes in an aircraft example.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

Electrophrenic respiration in patients with high quadriplegia.

After determining that 15 patients with high spinal cord injuries who were permanently apneic had viable phrenic nerves, electrophrenic respiration units were implanted. Thirteen of the patients (86%) achieved full-time respiration and two more achieved half-time respiration. Despite the loss of 8 patients to unrelated problems, 7 now use electrophrenic respiration continuously, one having done so for 16 years. The patient selection criteria, neurophysiological evaluation method, surgical procedure, postoperative care, and methods for diagnosis of system failures are presented. A comparison of the cervical and thoracic procedures is made. The cervical approach is preferred. Complications consisted primarily of equipment failures. For the external components there were several cases of antenna connection and battery connection failures. The implanted receivers failed in 6 cases with an average lifetime of 48 months, ranging from 24 to 108 months. In one case fibrosis around the electrode resulted in failure to stimulate the phrenic nerve effectively. In another case, infection required removal of the system which was reimplanted later and has continued to provide successful ventilation.

Electrophrenic respiration in patients with high quadriplegia

After determining that 15 patients with high spinal cord injuries who were permanently apneic had viable phrenic nerves, electrophrenic respiration units were implanted. Thirteen of the patients (86%) achieved full-time respiration and two more achieved half-time respiration. Despite the loss of 8 patients to unrelated problems, 7 now use electrophrenic respiration continuously, one having done so for 16 years. The patient selection criteria, neurophysiological evaluation method, surgical procedure, postoperative care, and methods for diagnosis of system failures are presented. A comparison of the cervical and thoracic procedures is made. The cervical approach is preferred. Complications consisted primarily of equipment failures. For the external components there were several cases of antenna connection and battery connection failures. The implanted receivers failed in 6 cases with an average lifetime of 48 months, ranging from 24 to 108 months. In one case fibrosis around the electrode resulted in failure to stimulate the phrenic nerve effectively. In another case, infection required removal of the system which was reimplanted later and has continued to provide successful ventilation. (Neurosurgery 24:529-535, 1989)

Application of Robust Observer to Failure Diagnosis for D.C. Speed-controlled System

The early failure diagnosis of a steel rolling plant has great economic benefit. Invited by a certain rolling mill, we have done the research of on-line supervision for steel rolling plants. Ihis paper is the report on the research of failure diagnosis for a D.C. speed-controlled system of a steel rolling. The D.C. speed-controlled system is a too fast system for many failure diagnosis methods, etc., to apply. Ve used detection observer method to monitor the changes of the system parameters. Simulation results show that detection observers may detect very small changes of system parameters in a few milliseconds. Because the failure modes of the D.C. speed-controlled system could not be output decoupled, we used robust detection observers to provide failure isolation information and used logic reasoning to make diagnosis decisions. The diagnosis system presented in this report would be a powerful subsystem of a large expert system for failure diagnosis of the steel rolling plant.

Functions required for advanced flexible manufacturing systems

This paper describes the concept and functions of multilevel supervisory systems aimed at a more efficient operation of advanced flexible manufacturing systems. A supervisory system consists of a control system, an adaptation system, a process monitoring system and a machine failure diagnosis system. Thus supervisory system assure the control of machining accuracy, the detection of human errors and tool breakage, and the identification of machine failures. This paper presents some successful results of the implementation of these subsystems.

New Trends in Systems Approach by the Interactive Use of Microcomputers

Placting stress on 1) the use of heuristics or experiences of an expert and 2) the interactive use of microcomputers or personal computers, some of the new trends are reviewed and discussed in systems approach especially in the field of interactive modeling of large scale systems, system failure diagnosis by use of knowledge engineering techniques, multiple criteria decision making and gaming system for complex problems

Industrial diagnostic research: an approach

Abstract We refer to the process of uncovering the causes of failure in service of a manufactured item as ‘industrial diagnostic’. Based on our experience with such a problem over the past two years, we concluded that traditional approaches are inadequate, while the more modern approaches of ‘system failure diagnosis’ are tailored more towards the detection of failure in a complex piece of equipment than towards the detection of failure over the ‘life cycle’ of a product. In this paper we briefly review the traditional as well as the more modern theories of diagnosis, and expound on our approach to the problem. This approach contained both ‘field’ as well as ‘production plant’ research. The objective of the former was to determine the influence of the various factors on product degradation in service; while the objective of the latter was to determine the ‘significant plant parameters’ and to construct the ‘quality function’ (QF) of the producing plant. It is our contention that industrial diagnostics has...