Research On Acoustic Localization Algorithm In AUV Guided Docking

In recent years,autonomous underwater submersibles play an important role in the exploration and development of seabed mineral resources,hydrological information observation and collection,and underwater equipment salvage.With the gradual application of underwater submersibles in actual engineering,the problem of insufficient self-sustaining capacity has become increasingly prominent.Therefore,the method of docking the submersible with the underwater docking station is generally adopted to realize the energy supply and data interaction of the submersible.In order to ensure the successful docking between the submersible and the platform,this paper analyzes and studies the guided positioning method and the trajectory correction method in the docking process.Aiming at the problem that the existing docking mode cannot meet the three requirements of real-time,short-term and attitude-like,this paper proposes a single-primary receiving docking mode,which transmits multiple signals through the docking platform and receives a single primitive on the submersible.The way to obtain the relative position between the submersible and the platform in real time provides a reliable guarantee for the navigation control system.Based on this positioning method,the orthogonality requirement should be adopted between the multiplexed signals,and the COSTAS frequency hopping signal is used as the system acoustic transmission signal to complete the positioning.The two working modes under the docking mode are simulated and analyzed.The advantages and disadvantages of the two working modes and the positioning performance under different environments are compared.The time-sharing working mode is selected to complete the guiding and positioning.Aiming at the problem that the fault-tolerance rate of the docking platform to the submersible motion trajectory is small,this paper builds a trajectory correction model based on the extended Kalman filter and the interactive multi-model algorithm combined with the position,velocity and slant range measurement information,which can adapt to the possible docking process.The absence of delay information and the trajectory of the maneuver.For the typical weak maneuvering trajectory in the docking process,‘straight-turning-straight',the algorithm design uses a combination of basic kinematics model and maneuvering variation model to construct the model set.At the same time,the model probability is modified by dynamically updating the Markov transfer matrix to optimize the filtering performance of the algorithm.The algorithm can effectively locate when only one delay information is detected.When the sub-model parameters match,the trajectory positioning accuracy is improved from 3.15 m to 1.06 m.When the sub-model parameters are not matched,the trajectory positioning accuracy is improved from 3.15 m to 1.13 m.Finally,on the basis of the guiding and positioning method and the trajectory correction method,the test data processing on the pool and the lake is performed.The results of the pool fixed-point test show that the positioning deviation results and the positioning accuracy results in the time-sharing transmission mode are better than the simultaneous transmission mode.The experimental results on the lake show that the trajectory correction method can estimate the target position only when one axis position is obtained,and improve the efficiency of algorithm positioning.At the same time,this algorithm can effectively adapt to the variability of the actual motion trajectory of the target,and improve the target positioning accuracy,and correct the horizontal positioning accuracy to 0.87 m.