This paper proposes a derating design method for system reliability improvement using a probabilistic approach. The proposed design approach tries to select derated levels of stresses such as temperature and current in system level unlike recent desig...
This paper proposes a derating design method for system reliability improvement using a probabilistic approach. The proposed design approach tries to select derated levels of stresses such as temperature and current in system level unlike recent design approaches focusing on component levels. System reliability is evaluated using component reliability metrics given as functions of uncertain stresses and time, based on a series system reliability model. Variation in stress, not considered in previous derating designs, is introduced to consider uncertain both environmental and operating conditions at customers’ hands in the field. Optimization problems for system reliability improvement are formulated and solved using FORM to find the best derating design. An electrical system comprising four components shows the details of the design method and the potential of the approach.