Abstract
Reliability of a system may differ greatly when operating under different environments. However, the existing works have either neglected the environment factor in system reliability analysis or considered this factor for binary systems or systems subject to a single environment (parameter). In this paper, we make contributions by modeling a multi-state system operating under hybrid dynamic environments affected by multiple environmental parameters. Different Markov chains with finite states are used to represent the random system behavior and dynamic environments, leading to an aggregated Markov process that models the overall system behavior. An effective approach based on state partitions and aggregations is suggested for assessing the system reliability indexes, including reliability, availability, multi-point availability, and environment-based reliability. A high-pressure homogenizer system is analyzed to demonstrate the proposed model and show the comparison of the reliability of system under fixed and dynamic environment.