Distributed Sensor Networks for Multisensor Data Fusion in Intelligent Maintenance

摘要

A new distributed sensor networks (DSN)for data fusion approach is proposed for smart maintenance in mechatronics environment. The approach leads to the concept of distributed sensor networks, which are basically systems composed of connected, co-operating information sources organized into highly structured, message-exchanging networks. The proposed DSN architecture has been designed for intelligent maintenance and diagnosis of machine performance, the detection of a normal/dangerous situations. The data fusion module is organized into three different sub-modules, each one characterized by appropriate fusion strategies. A neural tree-based approach is applied to fuse information at different levels of the DSN. Data fusion benefits from multiple attributes of neural networks. A framework of distributed sensor fusion is developed for information fusion of machine fault recognition to make the maintenance systems smarter. Neural tree methods are developed for data fusion for measuring machine intelligent maintenance. This distributed sensor networks approach for data fusion provide the better platform for real time intelligent maintenance.