Attractor based performance characterization and reliability evolution for electromechanical systems

Electromechanical systems (EMS) possess multi-type, fast-iteration, and customized characteristics in modern industry and are usually required for high reliability. To meet the reliability requirements of EMSs, it is essential to identify the physical relationship between the design parameters and reliability of EMSs, i.e., the physical reduction and holistic ability of reliability. However, current reliability models based on physical principles are costly and highly specific, making them lose the transferability to other EMSs and resulting in high computational and time costs when applied to multi-type EMSs. Nevertheless, current reliability models with transferability often lack clear physical reduction and holistic ability, which limits their ability to elucidate the reliability mechanisms in EMSs. Motivated by the above issue, an attractor based performance characterization and reliability evolution modeling method for EMSs is proposed. Firstly, we construct operational state equations of EMSs based on physical principles, discover the attractor as a universal characteristic in EMS operational states, and conduct attractor modeling to characterize EMS performance under constant and periodic external forces. Then, the attractor based reliability evolution modeling is conducted, including attractor based margin, degradation, and measurement equations of EMSs. Two typical EMSs including BLDCMs and resonant circuits are used for the case study, and their attractor based reliability equations physically related to the physical properties, functional thresholds, degradation, and uncertainties are established, revealing the combination of the physical reduction and holistic ability with the transferability of reliability. This work also provides a universal framework for EMS reliability modeling and is conducive to exploring the universality of EMS reliability.