Towed seismic exploration is an important way to detect the geological structure of the seabed. The traditional application of the towing streamer in the marine seismic exploration is the liquid streamer. Because of the easy deformation of the cross section, the vibrations of the controller and mechanical towing equipment are transmitted to the hydrophone through the axial extensional wave. Solid streamer can reduce the flow noise and the influence of mechanical vibration, so it can greatly expand the operational weather window. In addition, compared with the present applied liquid fiber towed streamer, the solid fiber streamer does not have oil leakage problems, so its a more environmental friendly choice. Three kinds of fiber optic scalar hydrophone streamers have being made. They were filled with oil, gel and polyurethane. The sensitivity of fiber optic hydrophones were measured by standing wave tube method before and after filled. The acoustic properties of solid streamer are analyzed by means of comparison. The experimental results show that the solid fiber optic hydrophone streamer is feasible.
In this paper, a pressure and acceleration insensitive reference Interferometer is used to obtain laser and public noise introduced by transmission fiber and laser. By using direct subtraction and adaptive filtering, this paper attempts to eliminate and estimation the transmission noise of sensing probe. This paper compares the noise suppression effect of four methods, including the direct subtraction (DS), the least mean square error adaptive elimination (LMS), the normalized least mean square error adaptive elimination (NLMS) and the least square (RLS) adaptive filtering. The experimental results show that the noise reduction effect of RLS and NLMS are almost the same, better than LMS and DS, which can reach 8dB (@100Hz). But considering the workload, RLS is not conducive to the real-time operating system. When it comes to the same treatment effect, the practicability of NLMS is higher than RLS. The noise reduction effect of LMS is slightly worse than that of RLS and NLMS, about 6dB (@100Hz), but its computational complexity is small, which is beneficial to the real time system implementation. It can also be seen that the DS method has the least amount of computational complexity, but the noise suppression effect is worse than that of the adaptive filter due to the difference of the noise amplitude between the RI and the SI, only 4dB (@100Hz) can be reached. The adaptive filter can basically eliminate the influence of the transmission noise, and the simulation signal of the sensor is kept intact.