EPE Association: EPE 1991 - 26 - Session 3.4: FAULT DETECTION AND DIAGNOSTIC

EPE 1991 - 26 - Session 3.4: FAULT DETECTION AND DIAGNOSTIC
You are here: EPE Documents > 01 - EPE & EPE ECCE Conference Proceedings > EPE 1991 - Conference > EPE 1991 - 26 - Session 3.4: FAULT DETECTION AND DIAGNOSTIC

   [return to parent folder]

   REAL-TIME IMPLEMENTATION OF FAULT DETECTION AND CONTROL ALGORITHMS ON ELECTRICAL DRIVES
By Giovanni Bucci; Carlo Cecati; Da Qing Zhang [View]
[Download]

Abstract: In this paper an electrlcal drive capable of real-time adaptive control, fault detection and diagnosis is presented. The system performs the required functions concurrently, starting from parameter estimation, using an on-line identiflcation method based on input-output measurements. It is mainly constituted by a multi-transputer board which drives the power converter by means of an on purpose developed I/O board. The software has been developed using Occam2 and the Occam Toolset. First experimental results show the effectiveness of the system and encourage to further development.

   IDENTIFICATION OF SIGNIFICANT PARAMETERS RELATIVE TO FAILURE DURING OPERATION OF A SWITCHMODE POWER SUPPLY APPLICATION TO PREVENTIVE DIAGNOSTIC
By P. Venet; G. Grellet; G. Rojat [View]
[Download]

Abstract: 30 % of failures in electronic systems being accountable to power supply, it seems necessary today to foresee the failures by analysing the different probable faults. With the secondary-resonant Forward structure as example, we compare the different modes of running with and without failure. After classification of the different components with respect to fiability, we study the alterations of the characteristics of the converter for the probable faults occuring on each element, the analysis of the results allowing us to restrict the number of informations necessary to detect and to discriminate the main failures.

   FAILURE DIAGNOSIS IN MEDIUM POWER SEMICONDUCTORS
By P. A. Aloïsi [View]
[Download]

Abstract: When a power electronic system fails during the development phase or in the field, it is very difficult to know why in order to avoid such event in the future. By the observation of the silicon die, by experience and with the collaboration of the user it is, most of the time, possible to know why the power semiconductor has failed. The main reasons of the failure could be either from the system itself: Overvoltage, overcurrent, thermal runaway, thermal fatigue, infririgement of safe operating areas, electrostatic discharge, or from the power semiconductor device itself: early failure due to a wrong assembly procedure. By the knowledge of the physical phenomena inside of the silicon crystal and the main overstresses given by each type of application and by a long experience and simulations, it is possible to determine the main reason of the failure. A picture of a failed die by each type of stress will be given to help the electrical engineer in his research of a more reliable system.

   A NON-CONTACT, REAL-TIME MONITORING AND CONTROL SYSTEM FOR ELECTRONIC CIRCUITS USING INFRARED THERMAL IMAGING
By S. A. Merryman; M. F. Rose [View]
[Download]

Abstract: A real-time diagnostic and control system has been developed for monitoring the temperature of components in an electronic circuit. The temperature is measured using a non-contact method utilizing an infrared, thermal imaging camera. Component temperature data is transferred to the control computer via a GPIB interface and compared against preset threshold values (both upper and lower limits). lf the measured temperature is out of the range for a particular component, the computer enters an automatic troubleshooting mode based upon a preprogrammed logic tree. Based upon the component temperature and the system operating conditions, an option is selected to modify the operating conditions of the circuit such that the component temperature will return to a normal and safe value. Among the control options are adjusting input parameters, load redundancy, system redundancy, total system protection, and alerting a human operator. The control of the circuit is accomplished using both analog and digital outputs from a multifunction board installed in the computer.

   KNOWLEDGE-BASED MONITORING AND DIAGNOSIS OF RAIL-GUIDED VEHICLES
By Eric Nantier; Markus Brotz [View]
[Download]

Abstract: The recent advance in the field of process computers and bus systems for railway vehicles has opened up a new area of information processing applicatlons. A knowledge-based monitoring and diagnostic system has been developed for real-time driver support. In contrast to rule-based approaches, the diagnostic knowledge concerning operation guide and fault repair is being formalized using a causal model and applied to the analysis of process data. The system includes a set of tools for editing, translating and evaluating rules for on-line use. The paper describes the realized system and discusses its influence on railway system applicatlons.

   DESIGN OF DIAGNOSTIC CONTROL EQUlPMENT DEVELOPED WITH AN EXPERT SYSTEM, APPLIED TO HIGH-SPEED ITALIAN RAILWAYS LOCOMOTIVE ( ETR 500 )
By W. Felloni; V. Lovati [View]
[Download]

Abstract: Following paper describes how basic concepts and techniques of an expert system are used in order to obtain a very friendly diagnostic system applied to three-phase drive controls of the high-speed train ETR 500. The preliminary tests have provided positive results confirming the design validity. In order to obtain an ''Intelligence Diagnostic System" a particular architecture has been developed. Every diagnostic fatal or warning message about failures happened on train are sent out by the utilised Expert System. In this way, talking about expert system means to treat a great part of the applications already developed in A.I. (Artificial Intelligence) laboratories, so it can be thought as the "heart" of this software tool application.

EPE 1991 - 26 - Session 3.4: FAULT DETECTION AND DIAGNOSTIC
You are here: EPE Documents > 01 - EPE & EPE ECCE Conference Proceedings > EPE 1991 - Conference > EPE 1991 - 26 - Session 3.4: FAULT DETECTION AND DIAGNOSTIC
	[return to parent folder]

	REAL-TIME IMPLEMENTATION OF FAULT DETECTION AND CONTROL ALGORITHMS ON ELECTRICAL DRIVES By Giovanni Bucci; Carlo Cecati; Da Qing Zhang	[View] [Download]
Abstract: In this paper an electrlcal drive capable of real-time adaptive control, fault detection and diagnosis is presented. The system performs the required functions concurrently, starting from parameter estimation, using an on-line identiflcation method based on input-output measurements. It is mainly constituted by a multi-transputer board which drives the power converter by means of an on purpose developed I/O board. The software has been developed using Occam2 and the Occam Toolset. First experimental results show the effectiveness of the system and encourage to further development.

	IDENTIFICATION OF SIGNIFICANT PARAMETERS RELATIVE TO FAILURE DURING OPERATION OF A SWITCHMODE POWER SUPPLY APPLICATION TO PREVENTIVE DIAGNOSTIC By P. Venet; G. Grellet; G. Rojat	[View] [Download]
Abstract: 30 % of failures in electronic systems being accountable to power supply, it seems necessary today to foresee the failures by analysing the different probable faults. With the secondary-resonant Forward structure as example, we compare the different modes of running with and without failure. After classification of the different components with respect to fiability, we study the alterations of the characteristics of the converter for the probable faults occuring on each element, the analysis of the results allowing us to restrict the number of informations necessary to detect and to discriminate the main failures.

	FAILURE DIAGNOSIS IN MEDIUM POWER SEMICONDUCTORS By P. A. Aloïsi	[View] [Download]
Abstract: When a power electronic system fails during the development phase or in the field, it is very difficult to know why in order to avoid such event in the future. By the observation of the silicon die, by experience and with the collaboration of the user it is, most of the time, possible to know why the power semiconductor has failed. The main reasons of the failure could be either from the system itself: Overvoltage, overcurrent, thermal runaway, thermal fatigue, infririgement of safe operating areas, electrostatic discharge, or from the power semiconductor device itself: early failure due to a wrong assembly procedure. By the knowledge of the physical phenomena inside of the silicon crystal and the main overstresses given by each type of application and by a long experience and simulations, it is possible to determine the main reason of the failure. A picture of a failed die by each type of stress will be given to help the electrical engineer in his research of a more reliable system.

	A NON-CONTACT, REAL-TIME MONITORING AND CONTROL SYSTEM FOR ELECTRONIC CIRCUITS USING INFRARED THERMAL IMAGING By S. A. Merryman; M. F. Rose	[View] [Download]
Abstract: A real-time diagnostic and control system has been developed for monitoring the temperature of components in an electronic circuit. The temperature is measured using a non-contact method utilizing an infrared, thermal imaging camera. Component temperature data is transferred to the control computer via a GPIB interface and compared against preset threshold values (both upper and lower limits). lf the measured temperature is out of the range for a particular component, the computer enters an automatic troubleshooting mode based upon a preprogrammed logic tree. Based upon the component temperature and the system operating conditions, an option is selected to modify the operating conditions of the circuit such that the component temperature will return to a normal and safe value. Among the control options are adjusting input parameters, load redundancy, system redundancy, total system protection, and alerting a human operator. The control of the circuit is accomplished using both analog and digital outputs from a multifunction board installed in the computer.

	KNOWLEDGE-BASED MONITORING AND DIAGNOSIS OF RAIL-GUIDED VEHICLES By Eric Nantier; Markus Brotz	[View] [Download]
Abstract: The recent advance in the field of process computers and bus systems for railway vehicles has opened up a new area of information processing applicatlons. A knowledge-based monitoring and diagnostic system has been developed for real-time driver support. In contrast to rule-based approaches, the diagnostic knowledge concerning operation guide and fault repair is being formalized using a causal model and applied to the analysis of process data. The system includes a set of tools for editing, translating and evaluating rules for on-line use. The paper describes the realized system and discusses its influence on railway system applicatlons.

	DESIGN OF DIAGNOSTIC CONTROL EQUlPMENT DEVELOPED WITH AN EXPERT SYSTEM, APPLIED TO HIGH-SPEED ITALIAN RAILWAYS LOCOMOTIVE ( ETR 500 ) By W. Felloni; V. Lovati	[View] [Download]
Abstract: Following paper describes how basic concepts and techniques of an expert system are used in order to obtain a very friendly diagnostic system applied to three-phase drive controls of the high-speed train ETR 500. The preliminary tests have provided positive results confirming the design validity. In order to obtain an ''Intelligence Diagnostic System" a particular architecture has been developed. Every diagnostic fatal or warning message about failures happened on train are sent out by the utilised Expert System. In this way, talking about expert system means to treat a great part of the applications already developed in A.I. (Artificial Intelligence) laboratories, so it can be thought as the "heart" of this software tool application.