Research On The Reliability Evaluation Of The DCS And Software Development

As the basic industries of national industrial production, the power industry plays an important role in promoting the quickly development of national economy, therefore, the safe operation of the power plant is crucial. Power plant is a large, complex, fast and quite precise control object, Distributed Control System(DCS) is widely used in the process control of the power plant. As the important component of the power plant, the reliability of DCS is paied more and more attention, so the evaluation of the reliability research on DCS is of great significance.This thesis describes the reliability indicators that are commonly used in the reliability engineering, include the Failure Rate, Reliability, Availability and Mean Time Between Failures(MTBF) etc. And then establish the reliability evaluation indication system for DCS respectively from the module-level, component-level, unit-level, plant-level and system-level by applying these indicators to the DCS reliability evaluation. This thesis proposes a comprehensive evaluation indicator on the basis of the regular evaluation indicators. In the cauculation of MTBF, the Failure Rates of the same type modules are regarded as the same, and it does not take into account the influence degree on the system of each module. The comprehensive evaluation indicator improve the problem.Take the DCS of siemens in the type of TXP as the research object, make model for it in Markov and Fault Tree Analysis methods separately, and compare the results of calculating. By the comparison, it is concluded that the Fault Tree modeling method is more suitable for DCS modeling of reliability. Make the model for DCS of NO.1 unit of a certain power plant in Fault Tree method, and conduct a series of quantitative analysis.A software is developed by the research of reliability evaluation of DCS, this software is used for modeling of reliability and quantitative analysis based on Fault Tree Analysis method.