Research on the Algorithm of Multi-Sensor Data Fusion in Intelligent Ship Performance Monitoring

A new approach is presented in this paper, which is based on multisensory data fusion to enhance the precision and robust of the surveillance system. This thesis can solve the uncertainty of multisource sensor data efficiently based on Dempster-Shafer evidence theory. Furthermore, fuzzy reasoning is applied to improve the fuzzy and uncertain degree of fusion, which makes it more accurate and stable under complicated operating conditions. Firstly, the multisensory data is preprocessed, which includes de-noising and characteristic extracting. Then, the Dempster-Shafer Proof Theory is applied to merge the sensor data to assure coherence and consistency among the various data sources. Fuzzy reasoning is applied to deal with uncertainty in the process of fusion, and further enhance the decision reliability. To prove the validity of the method, the paper builds an Intelligent Vessel Performance Monitoring Model and perform several sets of experiments. Finally, it is proved that the proposed approach is superior to the conventional one in terms of precision of multisensory data fusion, and its robust performance can be improved by 8.7%.