With wide application of multi-sensor systems

online fault seriously affect system performance

so how to make multi-sensor system with independent fault detection and diagnostic capabilities become primary problem.According to input signal

output signal and fault array of nonlinear multi-sensor system

establish a kind of multiinput multi-output processing and fault analysis model of extended Kalman filter（ EKF） with enhance selfregulating functions

on this basis

propose a kind of multi-sensor system online fault detection algorithm and its implementation framework on sensor node. Experimental results show that the proposed algorithm at low and high concurrent fault environments has highly accurate fault reporting capabilities. Also in contrast to the results of the temperature sensor to implement the proposed algorithm

the value of temperature monitoring verify the proposed algorithm has better system performance assurance capabilities than traditional algorithms.