A New Safety Assessment Method Based on Evidential Reasoning Rule With a Prewarning Function

Abstract

Safety assessment and early warning, as essential prognostics and health management elements, are of great significance for better understanding and predicting complex system states. To assess the safety of the system, this paper proposes a new safety assessment method based on evidential reasoning rule with a prewarning function. However, in practical engineering, two challenges still need to be addressed when assessing the system’s safety by fusing the monitoring in formation of safety indicators. First, the safety monitoring data may be affected by unexpected factors such as temperature, vibration, and noises caused by poor sensor quality. These disturbances may reduce the accuracy of the monitoring data used in assessing the safety of a system. For this challenge, these disturbance factors are divided into two parts: dynamic disturbances and static disturbances, namely, dynamic reliability and static reliability, respectively, and a new method is constructed to calculate the reliability. Second, the weight coefficient of the safety indicator cannot adjust to various conditions and track the characteristics of the system when assessing. For this challenge, a weight coefficient calculation approach is proposed based on the maximum deviation to assign an adaptive weight to every indicator. An experiment involving an oil pipeline is conducted. Compared with current methods, the proposed method could assess the safety of the pipeline more accurately.