Research On Key Techniques Of Wireless Sensor Networks Data Fusion In The Ship Monitoring System

With the development of shipping informatization,ship monitoring plays a more and more important role in navigation safety of ships.In the cabin of the ship,the number of equipment is large,the parameter type is diverse and the environment is complicated and changeable,which result in the real-time monitoring data not only has huge quantities and numerous types,but also may exist abnormality.Therefore,this paper is based on the ship monitoring system and focuses on the research of data fusion techniques in wireless sensor networks.The main contents of this thesis are as follows:(1)Combined with the system requirements,the design of overall framework and functional module of the ship monitoring system are given,and the key techniques involved in the system are described.(2)A trust and weight based data fusion model in wireless sensor networks is designed.The model consists of two modules: In the node module,each sensor node builds a trust table by monitoring transmission process of its neighbor nodes.When the cluster head election launches,sensor node elects reliable cluster head based on the built trust table.In the cluster head module,cluster head collects the weight of each node within cluster,distinguishes abnormal nodes and credible nodes and selects the data of credible nodes to fuse,which effectively improves the security and accuracy of data fusion,and reduces the network energy consumption.(3)A fuzzy logic based multi-sensor data fusion method in wireless sensor networks is designed.The method introduces the DENCLUE based outlier detection to improve the WDFM model,eliminating the abnormal data caused by weight lag updating,implements the weighted data aggregation to reduce data redundancy and uses the fuzzy logic system to fuse the multi-sensor data,which not only improves the accuracy of data fusion,but also reduces the data transmission and energy consumption.(4)The main function modules of the ship monitoring system are finished,and the realization of data fusion techniques and the main running examples are given.