Multi-sensor Fusion Approach with Fault Detection and Isolation Based on two-state Probability Ratio

Aiming at the problem of fault detection in the process of multi-sensor data fusion, a two-state probability ratio fault detection and isolation method is proposed in this paper, which can realize the independent detection of single sensor fault. Firstly, the state space of the current state under a certain confidence condition is selected according to the process equation, and then the probability of occurrence of the likelihood distribution of each sensor in this space is calculated, which is regarded as a test statistic. The threshold of the test statistics is designed according to the false alarm probability requirement of the system. The data fusion will be performed after isolating the sensor, when the test statistic is less than the threshold. The simulation shows that this method can detect the faults of each sensor independently in a multi-sensor system.