1 |
Statistical modeling and assessment of software reliabilityCamara, Louis Richard 01 June 2006 (has links)
The present study is concerned with developing some statistical models to evaluate and analyze software reliability. We have developed the analytical structure of the logistic model to be used for testing and evaluating the reliability of a software package. The proposed model has been shown to be useful in the testing and debugging stages of the developmental process of a software package. It is important that prior to releasing a software package to marketing that we have achieved a target reliability with an acceptable degree of confidence. The proposed model has been evaluated and compared with several existing statistical models that are commonly used. Real software failure data was used for the comparison of the proposed logistic model with the others. The proposed model gives better results or it is equally effective. The logistic model was also used to model the mean time between failure of software packages. Real failure data was used to illustrate the usefulness of the proposed statistical procedures. Using the logistic model to characterize software failures we proceed to develop Bayesian analysis of the subject model. This modeling was based on two different difference equations whose parameters were estimated with Bayesian regressions subject to specific prior and mean square loss function.
|
2 |
Improving asset care plans in mining : applying developments from aviation maintenanceAl Shaalane, Amir 12 1900 (has links)
Thesis (MScEng)--Stellenbosch University, 2012. / ENGLISH ABSTRACT: The aim of this thesis is to compare the aviation derived reliability metric
known as the Maintenance Free Operating Period (MFOP), with the traditionally
used, and commonly found, reliability metric Mean Time Between
Failure (MTBF), which has over the years shown some innate disadvantages
in the field of maintenance. It will be shown that this is mainly due to MTBF’s
inherent acceptance of failure and the unscheduled maintenance therewith directly
connected. Moreover, MFOP is successfully applied to a mining specific
case study, as to date, no other application of the MFOP concept to the mining
sector is known.
An extensive literature study is presented, which covers concepts relevant
to the overall study and which helps to contextualise the problem, revealing the
major shortcomings of the commonly accepted MTBF metric. A methodology
to analyse systems MFOP performance, making use of failure statistics to
analyse both repairable and non-repairable systems, is presented. Validation
makes use of a case study which applies the MFOP methodology to a system,
specifically in the mining sector.
It was shown that MFOP could be applied to the data obtained from the
mining sector, producing estimates which were accurate representations of reality.
These findings provide an exciting basis on which to begin to facilitate a paradigm shift in the mind set of maintenance personnel, setting reliability
targets and dealing with unscheduled maintenance stops.
KEYWORDS: Maintenance Free Operating Period, Mean Time Between Failure,
Maintenance, Mining / AFRIKAANSE OPSOMMING: Die doel van hierdie tesis is om die Onderhoudvrye Bedryf Tydperk (OBT),
’n betroubaarheidsmaatstaf afkomstig van die lugvaart industrie, te vergelyk
met die Gemiddelde Tyd Tussen Falings (GTTF) maatstaf wat tradisioneel in
algemene gebruik is, maar wat oor die jare inherente nadele met betrekking tot
instandhouding geopenbaar het. Dit sal bewys word dat hierdie nadele hoofsaaklik
ontstaan as gevolg van die GTTF se inherente aanvaarding van failure
en die ongeskeduleerde instandhouding wat daarmee gepaard gaan. OBT word
ook suksesvol aangewend in ’n mynwese-spesifieke gevallestudie, wat aaangegaan
is aangesien geen ander sooortgelyke aanwending in die mynwese sektor
tot datum bekend is nie. ’n Breedvoerige literatuurstudie word voorgelê wat
relevante konsepte dek en die probleem binne konteks plaas, en daardeur die
hoof tekortkominge van die algemeen aanvaarde GTTF metriek ontbloot.
’n Metodologie waardeur analise van die stelsel werkverrigting van die OBT
uitgevoer kan word met gebruik van onderbrekings statistiek om herstelbaar
sowel as onherstelbare stelsels te analiseer, word voorgestel. Geldigheid word
getoets deur ’n gevallestudie wat die OBT metodologie aangewend word spesifiek
vir ’n stelsel in die mynwese Dit is bewys dat OBT toegepas kan word op data afkomstig van die mynwese
sector, en skattings lewer wat akkurate voorstellings is van die werklikheid.
Hierdie bevindinge is opwindend, en dit dien as die basis vir ’n die
aanwending van ’n paradigmaskuif in die benadering van instandhoudingspersoneel
tot die daarstelling van teikens vir betroubaarheid en ook in hul
hantering van ongeskeduleerde instandhoudingsophoud.
SLEUTELWOORDE: Onderhoudvrye Bedryf Tydperk, Gemiddelde Tyd Tussen
Falings, Onderhoud, Mynbou
|
3 |
Validation of theoretical cost model for Power and Reliability : Case study of a reliable Central Direct Memory Access system / Validering av teoretisk kostnadsmodell för Power and Reliability : Fallstudie av ett tillförlitligt Central Direct Memory Access-systemShrivastava, Sonal January 2021 (has links)
Safety-critical applications employed in automotive, avionics and aerospace domains are placed under strict demands for performance, power efficiency and fault tolerance. Development of system hardware and software satisfying all criteria is challenging and time-consuming. System co-design based on specifications and desired high-performance requirements, is one solution to this problem, however, it remains a largely unexplored territory. Currently at KTH Royal Institute of Technology, a co-design framework in relation to theoretical system design models is being researched with the objective to move the embedded system design to a higher abstraction level. Presently, it focuses on correct-by-construction design of low power and reliable safety-critical systems. This thesis intends to assess the accuracy of this framework in comparison to conventional design approaches. The accuracy is evaluated empirically in terms of extra functional requirements - average total power consumption and Mean Time Between Failure (MTBF). A simple payload Central Direct Memory Access (CDMA) application is integrated with Xilinx Soft Error Mitigation (SEM) IP Core and source of system failure is a Single Event Upset (SEU) which occurs due to ionizing radiations. Measurements obtained from this reference system are compared to results determined theoretically from model related equations for the same system. Comparison of measured MTBF values with theoretical estimations shows that measured ones are higher by an average huge difference of 324.63%. Similarly for power consumption, measurements were found to be higher than estimated ones by 0.4465 Watts. In conclusion, it can be said that theoretical model design framework is not accurate and models must somehow take into account implementation dependent factors. Nevertheless, this case study provided a good insight and pathways for enhancements and optimizations to turn this reference into a dependable platform. Finally, future work required for desirable experiment system improvements are identified. / Säkerhetskritiska applikationer som används inom fordons-, flyg- och flygindustrin ställs strikta krav på prestanda, energieffektivitet och feltolerans. Utveckling av systemhårdvara och mjukvara som uppfyller alla kriterier är utmanande och tidskrävande. Systemdesign baserad på specifikationer och önskade högpresterande krav är en lösning på detta problem, men det är fortfarande ett i stort sett outforskat område. För närvarande vid KTH Royal Institute of Technology undersöks ett ramverk för samdesign i relation till teoretiska systemdesignmodeller undersöks med målet att flytta den inbyggda systemdesignen till en högre abstraktionsnivå. Nuvarande, den fokuserar på korrekt konstruktion av låg effekt och pålitliga säkerhetskritiska system. Denna avhandling avser att bedöma riktigheten i detta ramverk i jämförelse med konventionella designmetoder. Noggrannheten utvärderas empiriskt när det gäller extra funktionskrav - genomsnittlig total strömförbrukning och medeltid mellan misslyckande (MTBF). En enkel nyttolast central direktminnesåtkomst (CDMA) applikation är integrerad med Xilinx begränsning av en händelse (SEM) IP kärnan och källan till systemfel är en singelhändelse upprörd (SEU) som uppstår på grund av joniserande strålning. Mätningar erhållna från detta referenssystem jämförs med resultat som bestämts teoretiskt från modellrelaterade ekvationer för samma system. Jämförelse av uppmätta MTBF -värden med teoretiska uppskattningar visar att uppmätta värden är högre med en genomsnittlig enorm skillnad på 324,63%. På samma sätt för strömförbrukning befanns mätningarna vara högre än beräknade med 0,4465 Watt. Sammanfattningsvis kan man säga att det teoretiska ramverket för modelldesign inte är korrekt och att modellerna på något sätt måste ta hänsyn till implementeringsberoende faktorer. Ändå gav denna fallstudie en bra insikt och vägar för förbättringar och optimeringar för att göra denna referens till en pålitlig plattform. Slutligen identifieras framtida arbete som krävs för önskvärda experimentsystemförbättringar.
|
Page generated in 0.0767 seconds