Data-Driven Fault Detection and Diagnosis Based on Envionment/Platform Simulation

Motivated by the fact that individual unit of complex systems is not able to operate when it is isolated from the original plant. Functional duplicated platform of the original system based on CAN-Bus and hardware simulation is addressed in this study. the platform can provide all environmental stimulation signals for offline detection and diagnosis of individual unit. Meanwhile, based on dataflow, qualitative and quantitative knowledge of integrated system, construction of fault diagnosis system with knowledge active pushing, is proposed. This approach can realize integrated detection and diagnosis of faulty units by integrating server on environment platform. the effectiveness of proposed scheme is validated by experiments that have high availability and reliability, which shows its potential usage for fault detection and diagnosis of complex systems.