Showing papers on "Condition monitoring published in 1993"

Parameter Estimation, Condition Monitoring, and Diagnosis of Electrical Machines

Abstract: Introduction Summary of space phasor theory On-line signal processing of space-phasor quantities The effects of time harmonics on various space-phasor loci, harmonic amplitude estimation Monitoring of the rotor speed and the rotor angle Monitoring various machine parameters Diagnosis, condition monitoring Bibliography Index

Monitoring and diagnosis of rolling element bearings using artificial neural networks

Abstract: Vibration monitoring of components in manufacturing plants involves the collection of vibration data from plant components and detailed analysis to detect features that reflect the operational state of the machinery. The analysis leads to the identification of potential failures and their causes and makes it possible to perform efficient preventive maintenance. Work on the design of a vibration monitoring methodology for rolling element bearings (REB) based on neural network technology is presented. This technology provides an attractive complement to traditional vibration analysis because of the potential of neural networks to operate in real-time mode and to handle data that may be distorted or noisy. The significance of this work relies on the fact that REB failures are responsible for a large fraction of the malfunctions in manufacturing equipment. The technique enhances traditional vibration analysis and provides a means of automating the monitoring and diagnosis of vibrating equipment. >

Sensor/Actuator Failure Detection in the VISTA F-16 by Multiple Model Adaptive Estimation

Abstract: Multiple model adaptive estimation (MMAE) is applied to the Variable Inflight Stability Test Aircraft (VISTA) F-16 flight control system at a low dynamic pressure flight condition (0.4M at 20000 ft). Single actuator and sensor failures are first, followed by dual actuator and sensor failures. The system is evaluated for complete or "hard" failures, patial or "soft" failures, and combinations of hard and soft actuator and sensor failures. Residual monitoring is discussed for single and dual failure scenarios. Performance is enhanced by the application of a modified Bayesian form of MMAE, scalar residual monitoring to reduce ambiguities, automatic dithering where advantageous, and purposeful commands.

Method of predicting deterioration and damage of structural member and prediction apparatus for use with the method

Abstract: An apparatus for predicting a deterioration and damage in a structural member of a prime mover or the like has a use condition monitoring unit for monitoring a condition of use of the structural member, an inspection unit for inspecting a deteriorated and damaged state of the structural member, a use condition setting unit for setting a use condition by being supplied with use condition data of the structural member from the use condition monitoring unit, a simulation unit for forming a model of the deterioration and damage in the structural member on the basis of the use condition of the structural member and a result of inspection of the structural member at a preceding time, and for simulating the advancement of the deterioration and damage in the structural member on the basis of the deterioration/damage model, and an inverse analysis unit for correcting a simulation model formed by the simulation unit by comparing the simulation model with a result of inspection of the structural member at the present time. The simulation unit is arranged to predict a future deterioration and damage in the structural member by using the corrected simulation model and the result of present inspection. By a deterioration/damage prediction apparatus of the invention using this apparatus, the deterioration and damage in the structural member can be accurately predicted using a simulation and an inverse-problem analysis method with respect to use conditions and results of inspection of the structural member, i.e., information actually available.

Hardware self-tuning and circuit performance monitoring

Abstract: Self-tuning is a new clocking methodology borrowing heavily from both the synchronous and self-timed disciplines. A self-tuned system has an adjustable clock and measurement logic. During the tuning process the adjustable clock is made to run faster and faster until before the system fails. After tuning and during operation each cycle is measured and, if a failure is imminent, the system is retuned. During the tuning phase test vectors-either hardware embedded or software-select near maximum speed for a particular instance of the system. As self-tuning is predicated on self-test, it is essential to build in self-test features. These same self-test features are useful in circuit level performance monitoring. Two extremes on the continuum of self-tuning are discussed: at one extreme is purely hardware self-tuning and at the other, nearly purely software. Data is given from an experimental self-tuned primary memory indicating 70 ns access time DRAM can be operated at 45 ns or less. >

Pseudo Wigner–Ville Time-Frequency Distribution and Its Application to Machinery Condition Monitoring

Abstract: Machinery operating in a nonstationary mode generates a signature that at each instant of time has a distinct frequency. A Time-frequency domain representation is needed to characterize such a signature. Pseudo Wigner–Ville distribution is ideally suited for portraying a nonstationary signal in the time-frequency domain and is carried out by adapting the fast Fourier transform algorithm. The important parameters affecting the pseudo Wigner–Ville distribution are discussed and sensitivity analyses are also performed. Practical examples of an actual transient signal are used to illustrate its dynamic features jointly in time and frequency.

Application of non-stationary analysis to machinery monitoring

Abstract: The author discusses how nonstationary signal processes, such as the wavelet transform and the Wigner-Ville distribution, can be applied to machinery monitoring and diagnostics in industry. One major area of application is incipient failure detection in mechanical and electrical devices. It is argued that optimum incipient failure detection requires nonstationary analysis because failure signals: (a) are nonstationary; (b) are not repetitive in the earliest stages; (c) consist of several active frequency components; and (d) often occur over several scales. Some conventional methods used for machinery diagnostics are described, and their shortcomings are noted. These techniques include natural frequency envelope monitoring, cepstral analysis, and kurtosis. The opportunity for applying nonstationary techniques is indicated. >

Tyre condition monitoring

Abstract: A tyre condition monitoring system comprises a unit (1) having a sensor (2) for sensing pressure and temperature. The unit (1), which is mountable in a wheel of a vehicle comprises a voltage controlled oscillator (4) and a code generator (5) arranged such that a coded signal is generated in a time period related to the value of the pressure or temperature sensed by the sensor (2). The coded signal is transmitted from the unit (1) by a transmitter (8). In order to conserve power the unit (1) comprises a power supply (15) which is activated by a timer (16) from time to time causing the coded signal to be transmitted. A counter (11) monitors when the coded signal has been transmitted and, once transmitted, causes the power supply to be deactivated. A monitor circuit (18) continuously monitors for an unacceptable condition. An override circuit (19) responsive to the monitor (18) or to an external stimulus can also activate the power supply.

The use of hidden Markov models for condition monitoring electrical machines

Abstract: This contribution is concerned with the application of a statistical pattern recognition method to the diagnostic function of electric machine condition monitoring. It describes the hidden Markov modelling technique (HMM), which uses historical data as a training set against which it constructs and tests models of the processes under observation. Operating under the classification mode it fits multi-sensor inputs to appropriate models which allow simple rule based decision making to take place. The technique may also be regarded as possessing the properties of a data fusion centre, making it very applicable to process monitoring and performance mapping of systems. A description of the basic hidden Markov method is given, and experimental results, which give evidence of its utility for monitoring the condition of electrical machines, are presented.

Neural Networks For Intelligent Vehicles

Abstract: This paper is a survey of current research in applying artificial neural networks to the domain of intelligent vehicles. It describes work in three areas: video-based traffic monitoring, monitoring and control of onboard systems, and vision-based lateral control.

A new approach to motor condition monitoring in induction motor drives

Abstract: An approach for condition monitoring of an electrical machine is proposed. The method uses the power leads of the machine itself as the communication link between the sending station located within the machine and the receiving station located remotely outside the machine. The required communication circuit for realizing data transmission for both the ordinary 60-Hz power line and the special PWM (pulse-width-modulated) inverter-fed power line is constructed. The communication circuit uses an asynchronous serial communication protocol and FSK (frequency shift keying) modulation for realizing frequency multiplexing in the power line. An online winding temperature monitoring system for an inverter-fed induction machine is constructed using this power line communication link. Experimental results demonstrate satisfactory operation of the system. >

Condition monitoring of rolling-element bearings by using cone-kernel time-frequency distribution

Abstract: A method for condition monitoring of rolling element bearings is proposed in this paper. Timefrequency analysis using a cone kernel distribution applied to the envelope signal of the bearing vibration has shown an advantage over the high frequency resonance technique. In particular it is proven to be effective for multi-fault and ball fault detection of ball bearings. A comparison of the diagnostic ability of two other timefrequency distributions (spectrogram and distribution) is also presented.

Signal processing research in automatic tool wear monitoring

Abstract: A particularly important machine monitoring problem is the monitoring of tool wear in automatic metal drilling systems. The author briefly discusses the role of signal processing in tool wear monitoring and highlights several avenues of potential research in this area. These include the application and development of signal enhancement algorithms to reduce the corrupting effects of extraneous structural vibrations on the tool wear signal. Research directions in improved tool wear signal understanding are also discussed including detection techniques to classify a tool's condition using knowledge-based and statistical approaches. >

Orbit Complexity: A New Criterion for Evaluating the Dynamic Quality of Rotor Systems

Abstract: This paper introduces a new criterion called the index of orbit complexity (ICPX) to evaluate the dynamic quality of rotor systems during operation. This criterion originates from the singular spectra of the received signals and their spectra entropy. It can be used to evaluate the dynamic design quality of a rotor system. It is also useful in the condition monitoring of large-scale rotating machinery to monitor dynamic stability. Some examples from production practice are used to apply the criterion.

Partitioned simulation of hydraulic systems using transmission-line modelling

Abstract: This paper describes research into a distributed processor scheme for time-domain simulation of hydraulic circuits, ultimately for use in real-time applications, involving control and on-line condition monitoring. The approach adopted uses transmission line models of the pipelines as a means of decoupling the system components and implements the concepts developed in the authors' preliminary work. To assist this study a simulation program generator has been developed which links models of relevant components contained in a library. The program is executed on a multi-transputer platform. Using this facility, a study has been undertaken to assess the most appropriate number of processors and best circuit partitioning strategy for the case of a particular hydraulic circuit. From this investigation, partitioning guidelines are proposed

Microcomputer-based real-time bearing monitor

Abstract: A new approach to real-time monitoring of bearing operational conditions has been developed. An effective approach is to measure the real-time load and temperature fluctuations inside the bearing environment against any potential danger of overloading and/or overheating. The design, testing and implementation of the measuring electronics are presented, which forms the key part of the development of sensor-integrated smart bearings. Laboratory experiments illustrate the feasibility of the developed circuitry and verify that high fidelity wireless data transmission can be obtained using this method. >

Application of statistical signal processing for condition monitoring of rotor faults in induction motor

Abstract: The paper presents the research and development effort in finding a reliable and economical rotor fault condition monitoring system using an ordinary 486 PC together with a readily obtained inexpensive 14 bit AD converter The work presented in this paper focuses mainly on the no-load data so as to avoid the effect of cyclic variation of the load pattern upon the current signature obtained via current transformers which are readily available in large machines Results obtained from laboratory sized 22 kW cage induction motors are also presented

Helicopter Structures - A Review of Loads, Fatigue Design Techniques and Usage Monitoring

Abstract: : This report is a review of traditional practice with respect to helicopter structural integrity. Aspects covered are: significant fatigue loads for the airframe and rotor system, the methods used in the fatigue design of current and previous generation rotorcraft, fatigue test requirements, and health and usage monitoring methods.... Helicopters, Structural design, Loads (forces), Fatigue tests, Airframes, Condition monitoring, Rotors

Typical problems encountered with variable frequency drives in the industry

Abstract: Typical problems associated with voltage source inverter type VFDs (variable-frequency drives) and their surroundings are discussed. Suggestions are put forward to identify problems and to take corrective steps to reduce the overall down-time of an installation. The suggestions made in certain circumstances may not be the best in all cases. Also, for installations using VFDs, suggestions are made to improve the overall system performance by introducing passive and/or active filters and by making use of programmable logic controllers and computers to monitor and control process parameters. >

Specification of FMS real-time control based on Petri nets with objects and process failure monitoring

Abstract: This paper presents an approach for control and monitoring of discrete events systems. We show how it is possible to preserve the advantages of a hierarchical control structure for process monitoring purposes. The limitations of the monitoring system induced by the tools classically used for the specification of the control sequences are exposed. A new approach of knowledge representation is proposed: a database model is associated to the hierarchical control and monitoring structure. In this context, the main monitoring functions become more powerful and are able to treat most of the erroneous situations following from a process failure. >

Control and monitoring of large discrete event systems: a generic approach (to EMS)

Abstract: This paper deals with a hierarchical and modular structure for real time control and monitoring of large discrete event systems. The first aspect of these works concerns the complexity of the physical process to be controlled: the whole process is split up into more simple subproblems. This splitting up procedure is recursive and stops when the resulting subprocesses are simple enough to be solved. The resulting control structure is a tree-like structure of control subsystems in which the abstraction level of each node is decreasing from the root to the leaves. In the proposed approach, each node of the tree-like structure is not only a control system but also a control and monitoring system called a 'control and monitoring module'. In fact, each node is the root of a part of the tree-like structure and implements all the functions necessary for real time control and monitoring. >

Diagnostic reasoning technology for the on-board maintenance system

Abstract: The authors describe On-board Maintenance System (OMS) developed by Smiths Industries Aerospace and Defense Systems, Inc. which includes inflight diagnostic and prognostic analysis of engines and mechanical systems. This on-board system is designed to enhance the data collection and recording function by providing the pilot with real-time system health information, on-board system and subsystem diagnosis, system performance prediction, and to assist ground maintenance personnel in diagnosing system malfunctions, and scheduling required maintenance actions. Ground maintenance crews have access to on-board diagnosis results while the aircraft is in flight via data link so they can be prepared when the aircraft lands to quickly repair failed units. In addition to the real time system health monitoring, the on-board system allows ground crews to perform maintenance on demand rather than performing maintenance based on time in service, thus tailoring the maintenance performed to the individual needs of each aircraft. A technology demonstration of the application of a hybrid reasoning system called the Diagnostic Reasoning Demonstrator (DRD) has been built which integrates a number of different knowledge sources and reasoning methods to demonstrate the diagnostic and prognostic analysis of mechanical failures trends and events in a gas turbine engine. The diagnostic reasoning technology is centered around hypothetical reasoning. The DRD consists of an engine model and simulator, a failure scenario builder, and the hybrid diagnostic reasoner. The demonstrator is implemented on a color laptop workstation. >

Using Kohonen Feature Maps To Monitor The Condition Of Synchronous Generators

Abstract: This paper illustrates the way in which a neural network, employing unsupervised learning, can be used for the automatic surveillance of the operational condition of synchronous generators. Results show that, if the size of the network is chosen judiciously, then it is possible to cosistently, and unambiguously identifi a range of atypical conditions in such machines.

Damage Assessment from the Dynamic Response of Deteriorating Structures

Abstract: The demand for the residual life determination of a structure is a topic growing in importance, due to the high rehabilitation costs in the case of late damage detection. To this end condition monitoring techniques reveal an essential tool for planning repair or strenghtening and, as long as non destructive evaluation (NDE) methods are applicable, they are usually prefereable with respect to those requiring more or less extensive sampling. Among the NDE methods special attention seems to be paid to dynamic testing, because of its execution speed and capability to work with ambient vibrations [1],[2]. The paper subject pertains to this latter field and aims to present a monitoring technique devoted to provide a synthetic index of the structural state, whose value, if critical, must address towards more refined and oriented investigations. Within this context, it has found advantageous not to try to identify the remaining load carrying capacity directly, but to split the problem into two stages. Initially the dynamic behaviour of the structure is exploited and global damage symptoms are introduced in order to identify the damage effects at least in an overall sense; subsequently further problem dependent quantities, responsible for the damage origin and growth, are derived from the damage indicators and used for safety assessment purposes. Since this latter stage corresponds to a conventional design process, the critical point in the safety assessment problem is reduced to the identification of the dynamic properties of the structure. To this end, signal analysis procedures able to deal directly with the dynamic responses arising from almost arbitrary nonconservative, nonlinear and time — variant systems are used. A time domain approach is followed in either case and the Hilbert transform or the Wigner distribution is used depending whether the system allows for the uncoupling of the vibrating contributions or not. Accuracy and reliablity of the two approaches are discussed by the help of some numerical simulations, whereas their effectiveness is tested against an actual case study.

Machine component condition monitoring for automatic winders - links CAD drawing store to operating and maintenance data to help maintenance procedures and provide full service records

Abstract: The monitoring appts. for a machine comprises a store for drawings with identification for each drawing, a reader for the stored drawings and a pictorial display for the drawings read. The drawings are stored so that they can be classified and related to each other. USE/ADVANTAGE - For computer controlled condition monitoring of automatic winders. Links a CAD drawing system store to operating and maintenance data to help maintenance procedures and provide a full record. An automatic winding machine has a control unit (14) with an input section (16) for operating conditions and a display (17). The control unit (14) is linked to a central control unit (19) which in turn is linked to an evaluation unit (21). When the evaluation unit (21) diagnoses a faulty machine component from the process sensors it is shown on the display.

Abnormal condition monitoring device

Abstract: PURPOSE:To reduce a soft load in an abnormal condition monitoring device by detecting actual driving condition of a driving means, predicting driving condition, which complies to the condition of a vehicle, of the driving means, and monitoring any abnormal condition in a control means, which controls the driving means according to a difference between the actual driving condition and the predicted driving condition. CONSTITUTION:In a throttle controller provided with a constant speed traveling function, each of detection signals from a main switch 2, a control switch 3, an accelerator sensor 4, and a throttle sensor 5 is inputted to a main microcomputer 12 and a submicrocomputer 13 via an A/D converter 11 inside an ECU 1 respectively. The main microcomputer 12 and the submicrocomputer 13 control a motor driving circuit 17 and a clutch driving circuit 18 individually, and consequently, a throttle valve 10 is driven. In this case, actual driving condition of the throttle valve 10 is detected, while driving condition, which responds the condition of a vehicle, of the throttle valve 10 is predicted. Then, any abnormal condition of the ECU 1 is monitored according to a difference between the predicted driving condition and the actual driving condition of the throttle valve 10.

Intelligent maintenance support system for Syncrude mining trucks

Abstract: The large mining trucks at Syncrude are complicated integrated electro-mechanical systems. Their monitoring and maintenance involve a number of complex technical and economic issues. These ill-structured issues are becoming the bottleneck in implementing maintenance automation and reducing the maintenance cost which is an increasingly large portion of Syncrude's operating costs, presently in the tens of millions. In the paper, attempts are made to apply artificial intelligence techniques to the field of equipment maintenance. We present an Intelligent Maintenance Support System (IMSS) for truck condition monitoring, fault diagnosis and maintenance, focusing on the development and implementation of its major components, database, knowledge base, compiler, inference engine and user interface. >

Limitations of built-in current sensors (BICS) for I/sub DDQ/ testing

Abstract: Quiescent current (I/sub DDQ/) drawn by a static CMOS device is extremely small and is of the order of nanoamperes. Under many faults, (I/sub DDQ/) can increase by several orders of magnitude. Either an external or an on-chip current sensor can be used to detect enhanced static current drawn by a static CMOS device. An on-chip sensor, termed a BICS (Built-In Current Sensor) can be significantly faster. Implementation of BICS has received a lot of interest in the recent years. Some limitations posed by BICS on I/sub DDQ/ measurement caused due to increase in I/sub DDQ/ settling time as well as propagation delay is considered. Results indicate that careful attention needs to be given to circuit partitioning for implementing BICS. Some of the considerations that need to be taken into account while designing new BICS are presented. >

Validation and recovery of vibration data in electromachine systems using neural network software

Abstract: It is shown that neural network (NN) software can successfully be used for validation of vibration data and classification of distorted waveforms. This constitutes a confirmation of the feasibility of application of NNs in expert systems for the predictive maintenance diagnostics of rotating electromachine systems. Details of the developed networks, preparatory and training procedures, and experimental results are presented. >