Health Monitoring System for Autonomous Vehicles using Dynamic Bayesian Networks for Diagnosis and Prognosis

摘要: Autonomous Vehicles have the potential to change urban transport scenario. However, be able safely navigate autonomously they need deal with faults that its components are subject to. Therefore, Health Monitoring System is a essential component of autonomous system, since allows Fault Detection and Diagnosis. In addition, Prognosis also important, it predictive maintenance safer decisions during vehicle navigation. This paper presents Hierarchical Component-based Detection, Diagnosis using Dynamic Bayesian Network (DBN) residue generation, combination knowledge-based model-based detection, diagnosis prognosis approaches. We evaluate proposed different machine learning metrics dataset sensor readings gathered CaRINA II platform, CARLA simulator. Both simulated experimental results demonstrated positive performance DBNs even high rate missing data for some model’s variables.