Research On The Forward Looking Sonar System Of Unmanned Underwater Vehicle

As the eyes of UUV(Unmanned Underwater Vehicle),forward-looking sonar system plays an indispensable role on the voyage.Obstacle avoidance requires forward-looking sonar to acquire the DOA(direction-of-arrival and distance)of obstacles;underwater navigation requires current sonar to acquire the velocity of the unmanned vehicle relative to the bottom of the sea so as to calculate the current position;the seafloor topography detection needs to acquire the seafloor distance information of different azimuths,and the seafloor fluctuation state can be obtained according to the distance information of a continuous area.Therefore,almost all kinds of sonar systems are based on the three basic functions: DOA estimation,ranging and current measurement.Around these three functions,forward-looking sonar system is studied.The specific contents are as follows:The shape of array and parameters of signal are designed.The beam performance of circular array and concentric double ring array was designed and analyzed.Based on the theory of phased-array current measurement,a conformal array which cover the bottom surface of unmanned vehicle is designed.The fuzzy functions of CW(Continuous Wave)pulse and LFM(linear frequency modulation)pulse are analyzed,and LFM pulse is used as detection signal.LFM pulse can give consideration to both range resolution and range,and is insensitive to Doppler shift caused by vehicle motion.Algorithmically,the spatial gain is obtained by wideband beamforming,and the time processing gain is obtained by pulse compression with matched filter.The hardware of the forward-looking sonar system is mainly composed of transmission driver module,acquisition and transmission module and signal processing hardware platform.The transmitting drive module provides the driving signal for the transmitting transducer.The collector samples the signals of each channel of the receiving array,and the transmission module sends the collected data to the signal processing platform.The signal processing process includes seven steps: bandpass filtering,wideband beamforming,distance attenuation compensation,matching filtering,envelope extraction,smooth & compression and threshold judgment.GPU based on CUDA is used to accelerate the signal processing.The results show that the forward-looking sonar system designed can complete the realtime obstacle detection function and output the elevation angle,azimuth angle and distance of the obstacle.Under static conditions,Banks of 63.7 meters can be detected.Under moving conditions,obstacles such as bridge piers of 46 meters and bottom of the river of 4-6 meters can be measured.Test results verify that the prototype system can achieve real-time underwater collision avoidance function.