Functional dependability of the display unit software of the BLOK system
https://doi.org/10.21683/1729-2646-2017-17-2-36-40
Abstract
Aim. The article is dedicated to the challenges of evaluating the functional dependability of the display unit software (SW) that is part of the BLOK vital integrated onboard system as attributed to program errors within a 24-hour target time. One of the key tasks is the calculation of the values of such SW functional dependability characteristics as accuracy, correctness, security, controllability, reliability, fault tolerance and availability, which are the primary indicators for evaluating the health of safety devices. With all this taken into account, it is to be evaluated whether the checking of the software of the display unit before each trip with a departure test is required.
Method. The reference conditions do not contain statistical data of program executions over the course of its maintenance. There is also no information on the structural characteristics of the program (number of operators, operands, cycles, etc.) which prevents the use of statistical models of dependability, such as the Halstead metrics, IBM model or similar ones. That is why the Schumann model was chosen as the initial data definition apparatus. The method of evaluation of the display unit’s functional dependability is based on the findings of [1].
Results. At the first stage, the following initial data values were defined: initial number of defects in the program, program failure rate and probability of correct run. At the subsequent stage, the identified values were used to define such dependability parameters as probability of no-error as the result of program run within a given time, probability of no-failure of display unit as the result of program run within a given time and mean time to program failure. After the probability РSW (t) of no-error as the result of program run within a given time was calculated, such SW dependability attributes as accuracy, correctness, security and controllability were evaluated. After the probability of no-failure of the display unit РR (t) as the result of program run within a given time was calculated, an evaluation was given to such attributes as SW reliability and fault tolerance, while after the mean time to program failure TavSW was calculated, knowing the mean downtime due to elimination of the program error τpdt, the display unit availability for faultless execution of an information process at an arbitrary point in time Cfa was defined. The calculated partial functional availability coefficients for the display unit have shown that pre-trip checking of the unit and immediate elimination of errors, should such be identified, will enable a significant improvement of user performance of the onboard display unit (BIL) in terms of timely notification of the driver on the current operational situation to enable timely train control decision-making.
About the Authors
E. N. RozenbergRussian Federation
Professor, Doctor of Engineering, First Deputy Director General,
27, bldg 1 Nizhegorodskaya St., 109029 Moscow
N. G. Penkova
Russian Federation
Deputy Head of Safety and Algorithmic Support,
27, bldg 1 Nizhegorodskaya St., 109029 Moscow
A. S. Korovin
Russian Federation
Chief Specialist of Computer-Based Devices Development,
27, bldg 1 Nizhegorodskaya St., 109029 Moscow
References
1. Shubinsky IB. Funktsionalnaia nadiozhnost informatsionnykh system. Metody analiza [Functional reliability of information systems. Analysis methods]. Dependability Journal LLC 2012 [in Russian].
2. Shukhina EE, Astrakhan VI. Bezopasni lokomotivni obiedinenni kompleks BLOK [BLOK Vital Integrated Onboard System]. Moscow; 2013 [in Russian].
3. GOST R IEC 61508–7–2012. Functional safety of electrical, electronic, programmable electronic safetyrelated systems. Part 7: Overview of techniques and measures.
4. Explanatory note with the display unit dependability calculation.
Review
For citations:
Rozenberg E.N., Penkova N.G., Korovin A.S. Functional dependability of the display unit software of the BLOK system. Dependability. 2017;17(2):36-40. https://doi.org/10.21683/1729-2646-2017-17-2-36-40