A Hybrid Real-Time System for Fault Detection and Sensor Fusion Based on Conventional Fuzzy Clustering Approach

Multiple sensor integration and fusion is essential to increasing sensor accuracy and reliability in control systems. Most popular fusion methods depend on the sensors models and signals statistics, where a previous knowledge about the sensors is required 21s in Kalman filtering based approaches. In this paper, an efficient new hybrid approach for multiple sensor fusion and fault detection is proposed, addressing the problem with multiple faults in different directions, which is based on conventional fuzzy soft clustering, and requires no prior knowledge or information about the used sensors. The proposed hybrid approach consists of two main phases. In the first phase a single fuser for the input sensor signals is generated using the fuzzy clustering c-means algorithm. In the second phase a fault detector was generated based on the artificial immune system (AIS)