Study On Data Fusion Method And Application Technology Of Multi-information Measurement For AUV

Autonomous Underwater Vehicle (AUV) play a important part in the areas of investigation and exploration of the marine environment, and the role is more important in the areas where the conventional marine research vessel can't enter. AUV is multidisciplinary research which is an integration of senor technology, artificial intelligence, automatic control, pattern recognition, information fusion, and systems engineering, and is a fast development discipline. This paper focuses on data fusion method and application technology of multi-information measurement for AUV. Research contents and results are as follows:1. Describes the importance of multi-information measurement for AUV of the marine environment; discusses the development status of the technical aspects of marine surveying for AUV; finally analysis the background and significance of the data fusion theory in AUV multi-information measurement areas.2. Describe the both internal and external sensing system of AUV according to the sources of information; establish an accurate model of environmental information by refinement perception model.3. Research the application of the wavelet transform denoising method in AUV sea trial measurement data treatment. Wavelet transform in signal denoising can retain the high-frequency details of signal. According to AUV sea trial measurement data denoising, prove that wavelet transform method meet the AUV sea trial measurement denoising requirements.4. Research the AUV sea trial multi-information data fusion method. AUV single-information on the marine environment is isolated; this article established a set of asynchronous multi-information data fusion method. Rationality of the data fusion method is proved by integrating the current profile information and navigation information.5. Propose an auxiliary method of AUV navigation when the DVL data in case of failure. An "AUV speed-current profile information" relevancy is established when the DVL senor data is normal, and when the DVL senor data is failure, AUV speed is estimated from the relevancy and the current profile information. From the simulation results, the calculate AUV speed have high accuracy.