Reliability evaluation method for multi task profile hybrid system based on dynamic Bayesian network

DOI:	10.11809/bqzbgcxb2024.05.022
Keywords:	hybrid system; reliability assessment; system reliability; dynamic Bayesian network; multiple task profiles
Abstract:	When evaluating the reliability of hybrid systems, which have the characteristics of numerous tasks and complex and ever changing environments, it is often necessary to consider different task profiles and working conditions. Considering the problems such as the change of logic structure, the different number of components, the multi state failure of the components to be worked and the continuation of the fault, a reliability modeling and comprehensive evaluation method of the multi task profile of the hybrid system is proposed. Taking a hybrid electronic system with a voting model as an example, a multi task profile reliability model is established. Based on dynamic Bayesian analysis of the fault status of the working component, the impact of the continuity of the working component fault between task profiles on the system reliability evaluation results is obtained. At the same time, Monte Carlo sampling method is used to obtain the component life value under the influence of variable parameters in the dynamic system. Taking the radar antenna electronic system as an example, a comprehensive reliability model was tested. The model was compared with a model that did not consider the failure of the working component. The results showed that considering the failure of the working component can more accurately understand the trend of system reliability changes. This case study validates the effectiveness of this method for reliability assessments of hybrid systems under multiple task profiles.
Issue:	Vol. 45 No. 5 (2024)
Published:	2024-05-31
PDF HTML