Fault Diagnosis while Monitoring Environmental Variables by a Sensor Network

This paper develops limit and trend checking methods for fault detection in a wireless sensor network, dployed for monitoring environmental variables (temperature and relative humidity) inside a closed space container. Based on the data of the practical experiments, it classifies many types of faults and defines some rules to diagnose corresponding faults. It also investigates fault detection based upon residuals, performed by a method attained of a new modeling approach.