The Study Of Underwater Acoustic Active Matched Field Localization

In recent years, there is much emphasis on matched field processing technology,which is a new cross research of acoustic model and signal processing technology. Matched field processing includes three parts: the inversion for acoustic parameters,underwater sound source localization and noise suppression. Matched field source localization is a technology of reversing underwater sound source location based on acoustic model and received signals. Compared with traditional method,since it relies on the physics and acoustic characteristics of water channel,not only has the application range of the matched field source localization been extended,but also the accuracy has been greatly improved.Currently,the research of matched field localization mainly focuses on passive localization. With noise reduction technology is widely adopted in submarine,passive detection technology encounters more challenge. Also there is an increase in the demand for the detection of underwater static target. Therefore,active sonar is receiving more attention,again. Due to the complexity of ocean environment,traditional active sonar technology does not take into account the effect of ocean environment,and thus degrades the accuracy of localization. In this dissertation,we have taken in-depth study of matched field source localization theory and methods,and further proposed a method of active matched field source localization based on the features of active sonar signal.There are several creative points in this dissertation:1. Successfully applying conventional passive matched field localization method to the localization issue of active sonar system. With the scheme of active matched field localization presented, effective cancellation of localization deviation generated by the uncertainty of ocean environment can be gained, thus ameliorating the localization performance of active sonar system.2. Modifying existing acoustic model and employing the modified form in the computation of replicas. Acoustic model is the basis of matched field source localization. This dissertation introduces the acoustic modeling theory and several acoustical models in this area,among which Normal Mode model is mainly studied theoretically and experimentally. As a kind of acoustic calculating program,KRAKEN is modified to adapt to the requirement of source localization,and used to solve active acoustic field as well as demonstrate the validity of theoretical calculation.3. Active matched field processor is devised according to the feature of active sonar signals. Matched field processor is another important component of matched field source localization. This dissertation has done intensive study on linear matched field processor and adaptive matched field processor,and theoretically analyzed their performances,respectively. According to the feature of active sonar signal,we have done in-depth study on broadband matched field processor,and successfully extended it to active matched field source localization. The ocean environment is devised through computer simulation,and active replica field is calculated based on Normal Mode model. Using broadband matched field processor the active replica vector and received signal vector are matched,and derived the ambiguous surface. Simulation results demonstrate the efficiency of our method.