Research On AUV Cooperative Navigation System Of Information Fusion Technology Under Communication Constraints

From now on,more cooperation researches about AUV navigation is under the ideal communication,which assumes underwater acoustic communication between AUVs are real-time,reliable,neither delay nor loss of communication.However,in actual AUV navigation process,underwater acoustic communication becomes unavailable,that due to the internal and external factors like the uncertainty of underwater acoustic communication and the complexity of the underwater environment.Therefore,with the background of the application of AUV,this paper analysises the data fusion algorithms of multi-AUVs cooperation navigation under the limited communication.First of all,this paper proposes the joint distributed state of information filter algorithm under the ideal communication of multi-AUVs cooperation navigation,based on the kinematic model of multi-AUVs.The method preserves the historical states in the filter,which avoids the process of time updating,by using the joint distribution state;at the same time,the information parameters is sparse,thus the computational complexity of the filter is less;and then get the moments parameters which has clearer physical meaning than the information parameters,by using the Cholesky modification algorithms for its special properties make the parameters change only partial to further decreasing the computational complexity.Finally,the paper designs the distributed form to deduce the decentralized joint distributed state of information filter algorithm.After that,the paper deduces the new data fusion algorithm of multi-AUVs cooperation navigation under the limited communication based on the decentralized joint distributed state of information filter algorithm.Firstly in the adding states,regardless of the communication delay effect,the method uses the joint distribution information parameters to contain the historical states about the communication delay in the filter,and adds the latest time of state variables to the joint distribution information parameters.Then in the updating measurements,considering about the random loss of communication,the Gillbert-Elliott channel model is established to introduce the random variables that detemines the loss circumstances of the latest information,and then brings the random variables into the filter to strike the information update parameters under the communication loss;At the same time,taking into account the effect of communication delay,the method measures to obtain the latest information on real arrival time based on the communication delay time and sampling time,for geting the latest measurements information to fusion the corresponding of the the joint distribution information parameters,which to realize the error-free data fusion.At lastest in the processing of the state recovery,the method designs to take advantage of three times of the Cholesky decomposition method towards the small-scale and low-dimensional matrix,insteading of one time of the Cholesky decomposition method towards a wide-scale and highdimensional matrix,which completes the conversion of the information parameters and solves the state estimates of the lastest time.The use of the generalized Cholesky decomposition method is studied to solve the information matrix no-positive in the recovery process,completing the data fusion method of multi-AUVs cooperation navigation under the limited communication,enhancing the practicability of the algorithm to achieve the optimization algorithm.What is more,this paper designs the simulation experiment and the really AUVs experimental data of the new data fusion method of multi-AUVs cooperation navigation under the limited communication,and analysis the simulation results,in order to indicate the superiority of he data fusion method of multi-AUVs cooperation navigation under the limited communication.