Research On Underwater Cooperative Navigation State Estimation Method Based On Acoustic Ranging

Although the progress of inertial sensor technology has greatly reduced the size,power consumption and cost of AUV(Autonomous Underwater Vehicle)sensor,the position error of AUV will still accumulate over time only depending on the sensors for dead reckoning.The latest advances in acoustic communication methods provide external measurement information for AUV.By fusing the relative position relationships between AUVs,it is possible to prolong underwater mission time and maintain effective XY position errors.Multi-AUV cooperative navigation system uses acoustic communication to measure the different points of task area at the same time.It can provide functional sampling,spatial sampling and time sampling that a single AUV can not provide.Thus Multi-AUV cooperative navigation system has more advantages than a single AUV system in task complexity and fault tolerance.Therefore,the state estimation problem of multi-AUV cooperative navigation,as the basis of underwater cooperative system,has also received widespread attention.Due to the ubiquitous problems of packet loss,poor robustness,multi-channel effect and unknown noise parameters in underwater acoustic communication channels,this paper focuses on the multi-AUV cooperative navigation state estimation method for underwater acoustic communication.The main work of this paper includes:1.Researchs on underwater cooperative navigation system based on acoustic ranging.Firstly,several common reference coordinate systems of underwater vehicle are introduced,and how to realize the transformation of position between coordinate systems is described.Secondly,for the multi-AUV cooperative navigation system based on acoustic ranging,various sensors related to it are introduced.Finally,in order to study the motion of underwater vehicle,the underwater kinematics model is established,the motion parameters are introduced in detail,and the measurement model based on acoustic ranging is established,which provides a model basis for designing a reasonable state estimation method of underwater cooperative navigation.2.Research on the state estimation method of underwater cooperative navigation with heavy-tailed measurement noise.Underwater communication channel is unstable,and the multi-path effect leads to outlier,which makes the relative distance measurement noise present a heavy-tailed non-Gaussian distribution.To solve this problem,a method of cooperative navigation state estimation based on maximum correntropy criterion is proposed.Firstly,the principle of maximum correntropy criterion is introduced in detail,and compared with another robust method of the M estimation.Secondly,for the non-linear system of underwater cooperative navigation,the divided difference filter based on Sterling polynomial interpolation formula is introduced,and the robust algorithm of AUV cooperative navigation based on acoustic ranging is designed.Finally,the simulation results of underwater cooperative navigation with heavy-tailed measurement noise show that the proposed method has better estimation accuracy than the existing methods,and the effectiveness and robustness of the proposed method are futher verified by the cooperative navigation experiment.3.Research on state estimation method of underwater cooperative navigation with unknown system noise and measurement noise.In underwater cooperative navigation,the ontology sensor is affected by many factors such as temperature,salinity,depth,water flow,interface reflection and refraction,which leads to unknown statistical characteristics of system noise;and underwater acoustic channel is also a high noise channel with strong reverberation,narrow channel bandwidth and strong multipath effect,which leads to inaccurate statistical characteristics of measurement noise in cooperative navigation and even changes with time.To solve this problem,an adaptive cooperative navigation state estimation method based on variational Bayesian is proposed.Firstly,the general form of Gaussian filter is introduced.Secondly,two state estimation methods of unknown noise,Sage-Husa method and variational Bayesian estimation method,are deduced in detail.The advantages of variational Bayesian method are analyzed in principle,and integrated into the underwater cooperative navigation state estimation algorithm based on acoustic ranging.Finally,the validity and feasibility of the algorithm proposed in this chapter are verified by algorithm simulation and the cooperative navigation experimental analysis.4.Research on cooperative navigation information fusion estimation method of multi-AUV.Firstly,a robust information filtering method based on Wishart distribution is proposed for the information fusion of multiple relative distance measurements received at the sampling time.Secondly,a model of communication packet loss is established to solve the problem of measurement information packet loss caused by the complexity of underwater acoustic environment,and a new robust information filtering cooperative navigation algorithm is proposed for communication packet loss.The simulation results show that the proposed algorithm is robust in dealing with the problem of communication packet loss,and has a wide application prospect in underwater multi-sensor information fusion technology.