Health Assessment of Complex Systems Based on the Evidential Reasoning Rule With Nonequidistant Data
- 31 March 2025
- journal article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 74 (00189456) , 1-13
Abstract
Health assessment of complex systems is critical for Prognostics and Health Management (PHM), enhancing operational reliability and preventing catastrophic failures. However, in engineering systems, different testing intervals for subsystems complicate the evaluation of the system’s overall health state. Excluding data points with brief testing intervals could lead to a significant loss of critical information and compromise the accuracy of the health assessment. Towards this end, we develop a novel information fusion mechanism for non-equidistant data, introducing a federal evidential reasoning (FER) rule-based health assessment method. Firstly, we divide non-equidistant indicators into corresponding subsystems and define the calculation period and fusion period. On this basis, a corresponding non-equidistant information processing method is proposed. Secondly, to reduce the effect of non-equidistant data on the assessment results, the non-equidistant information discount coefficient is introduced to quantify the impact on indicator weight and reliability. Moreover, based on the calculation period and fusion period, the predicted health state and original health state can be obtained by the ER rule-based information fusion mechanism, respectively. Finally, a health assessment experiment of the Laser Inertial Measurement Unit (LIMU) is conducted. Comparative experiments demonstrate that the proposed FER rule-based model significantly outperforms the traditional ER rule-based model and other similar health assessment methods.Keywords
Funding Information
- National Natural Science Foundation of China (62403474)
- National Natural Science Foundation of China (62203365)
- National Natural Science Foundation of China (62227814)
- Shaanxi Province Science and Technology Innovation Team (2020TD-24)
This publication has 33 references indexed in Scilit:
- An evidential combination method with multi-color spaces for remote sensing image scene classificationInformation Fusion, 2023
- BRN: A belief rule network model for the health evaluation of complex systemsExpert Systems with Applications, 2022
- A new hidden behavior prediction model of complex systems under perturbationsKnowledge-Based Systems, 2022
- Belief Combination of Classifiers for Incomplete DataIEEE/CAA Journal of Automatica Sinica, 2022
- Improving the Navigation Performance of the MEMS IMU Array by Precise CalibrationIEEE Sensors Journal, 2021
- A New Evidential Reasoning Rule With Continuous Probability Distribution of ReliabilityIEEE Transactions on Cybernetics, 2021
- A New Evidential Reasoning Rule-Based Safety Assessment Method With Sensor Reliability for Complex SystemsIEEE Transactions on Cybernetics, 2020
- Monocular Visual–Inertial State Estimation With Online Initialization and Camera–IMU Extrinsic CalibrationIEEE Transactions on Automation Science and Engineering, 2016
- Evidential reasoning rule for evidence combinationArtificial Intelligence, 2013
- Rule and utility based evidential reasoning approach for multiattribute decision analysis under uncertaintiesEuropean Journal of Operational Research, 2001