FORECASTING TIME-DEPENDENT FAILURE RATES OF SYSTEMS OPERATING IN SERIES AND OR IN PARALLEL

This paper discusses an unconventional but powerful approach to analysing the observed failure rates of complex systems that operate in series under changing operational and environmental conditions. This approach contrues that such failure rates are time-dependent and hence can be treated as time series. Since the time series thus generated are complex in the sense that they are either aggregates, products or both of two or more time series, especial time series techniques are required for their analyses. Fortunately, some results recently developed by the author and others have been found useful for the analyses of such time series. Forecasting procedures are discussed and illustrative examples are given. The methods and techniques used in this paper can also be used for the analysis of reliability-decay processes (see Singh, 1991), actual failure times, times between failures and interactions between failure times and maintenance times of complex systems.