Evidence system reliability

All of the areas of proof described above provide evidence concerning the current operations of the computer systems in place, but can those same controls be expected to continue to function over time Certainly a trend of control provides some presumption, and annual SOP review procedures provide a degree of assurance, but the most significant evidence of system reliability lies internal to the software and is documented only through a review of that source code itself. 

Operational and maintenance plans should be prepared for the computer system and its associated measurement and control instrumentation. Operational plan review will focus on system reliability, performance, diagnostic records, instrument and system I/O calibration, and the provision of critical data to support the batch record. Procedures for controlling the system (e.g., system management, security, and process operations) should be reviewed to verify that they are current, in place, and being followed. For each procedure required for the system there should be documented evidence that the relevant operatives have been trained in its use. All procedures must be written and approved according to the site procedures for writing and approving SOPs. 

The FAT provides evidence that the hardware and software are fully integrated, that they operate as indicated in the computer system specification deliverable, and meet the expectations of the user as defined in the requirements specification deliverable. This final formal integration test should be completed in an environment very similar to the operational environment. The system can be subjected to a real-world environment by using emulators and/or simulators which mimic system interfaces. The user s representative should evaluate the supporting documents, the operation, system functionality, and system reliability. 

The data summarized in Tables 2.4 and 2.5 indicate that spacecraft power system reliability improves with an increase in redundant units, regardless of the type of regulator integrated in the power system. Furthermore, it is evident from the reliability data that the number of redundant units (2, 2) for the dissipative regulator offers higher reliability compared with the PWM regulator. 

Preliminary studies performed by the author indicate that implementation of redundant components in the spacecraft or satellite power systems does improve the system reliability, but it does so at the expense of higher weight and component costs as a function of mission length or duration. An estimated increase in total power system weight and cost as a function of mission duration is evident from the published data summarized in Tables 2.10 and 2.11, respectively. 

Zhao, S. 2010. Power distribution system reliability evaluation by D-S evidence inference and Bayesian network method. IEEE 11th International Conference on Probability Methods Applied to Power System (PMAPS), Singapore, 14-17 June, 2010. 

An increase in system reliability is evident from comparing Equations 1 and 3. Figure 3, which depicts the system reliability with and without FACTS devices, confirms this assertion. The average increase in reliability was found to be about 0.18%. The financial savings that result from the prevention of cascading failures magnify the impact of even the smallest improvements to grid reliability. 

An examination of all calculated cases presented here discloses that, whatever the nature of the system or the shape of the initial distribution, b" (x) nearly always drops steeply in the rai e of large x-values. Evidently, a reliable preparative method must make use of this general feature. Hence, repeated extraction of the concentrated phase riiould yield the best results obtainable. By its very nature the procedure is highty suited for application in large-scale fractionation work. 

Another area where formalization intersects with probability is in assurance for fault-tolerant systems. Many system failures are due to flaws in fault tolerance the very mechanisms that are intended to prevent failure become the dominant source of failure Formal verification of these mechanisms produces two very valuable results first, it requires precise specification of assumed component failure modes, the number of these to be tolerated, and their assumed probabilities second, it provides convincing evidence (i.e., a proof) that the mechanisms work, provided the number and modes of component failure are consistent with those specified. This bipartite division separates assurance for the correctness of the mechanisms from calculation of system reliability. 

One way in which we can carry over confidence into a new product from past experience with earlier ones is via reuse of intellectual property. Thus a software module that has been in use in earlier systems will have built up a history of use that should allow its contribution to the unreliability of a new system to be computed. There already exist probability models that allow the system reliability to be computed from the reliabilities of the component modules [Littlewood 1979]. However, there has been little operational vindication of this approach, and there must be some concern that the assumptions are a little unrealistic for example, it is not clear that the reliability of a module in a novel application (i.e. in a new architecture) will be the same as it was in the environment in which the failure data used in the estimation of its reliability was collated. This is an area where further research is needed in the event that this assumption is violated, it may still be possible to use past evidence to estimate the reliability of a module in a novel application so long as we have sufficient information about the differing natures of the... 

With BBNs, it is possible to articulate expert beliefs about the dependencies between different variables and to propagate consistently the impact of evidence on the probabilities of uncertain outcomes, such as future system reliability (Falla, 1997, Ch 4). 

Variations in a produet s material properties, serviee loads, environment and use typieally lead to random failures over the most protraeted period of the produet s expeeted life-eyele. During the eonditions of use, environmental and serviee variations give rise to temporary overloads or transients eausing failures, although some failures are also eaused by human related events sueh as installation and operation errors rather than by any intrinsie property of the produet s eomponents (Klit et al., 1993). Variability, therefore, is also the souree of unreliability in a produet (Carter, 1997). However, it is evident that if produet reliability is determined during the design proeess, subsequent manufaeturing, assembly and delivery of the system will eertainly not improve upon this inherent reliability level (Kapur and Lamberson, 1977). 

In addition, the protocol should specify a sufficient number of process runs to prove consistency of the process, and provide an accurate measure of variability among successive runs. The number of batches should depend on the extent of validation and complexity of the process or importance of any process changes. Furthermore, the protocol should address the quality of materials used in the process from starting materials to new and recovered solvents, and evidence of the performance and reliability of equipment and systems. 

Because the rules or procedures in expert systems are heuristic they are often not well-defined in a logical sense. Nevertheless, they are used to draw conclusions. A conclusion can be uncertain because the truth of the rules deriving it cannot be established with 100% certainty or because the facts or evidence on which the rule is based are uncertain. Some measure of reliability of the obtained conclusions is therefore useful. There are different approaches used in expert systems to model uncertainty. They can be divided into methods that are based on... 

---