Research On Master-slave Mode AUV Cooperative Navigation Algorithm Based On Nonlinear Filtering

Under the background of the increasingly critical development of marine resources,autonomous underwater vehicles are playing an important role in this field,and they have significant military and economic value.Multi-AUV collaboration can improve efficiency.At the same time,multi-AUV collaboration can integrate the observation information of each AUV,realize information sharing and effectively reduce the cost of the entire navigation system and improve reliability.Therefore,it is particularly important to effectively coordinate the AUV and realize its precise underwater navigation and positioning.The research is divided into the following parts:1.The commonly used coordinate systems and sensors of the collaborative navigation system are introduced.A master-slave AUV collaborative navigation mathematical model on a two-dimensional plane and a three-dimensional space for multi-AUV collaborative underwater application scenarios is established.In view of the limited observation information and weak observability of the mathematical model that only incorporates distance information,the observability analysis of the model is carried out from a quantitative and qualitative perspective.It is concluded that the system is unobservable when the master-slave AUV position is relatively fixed.A zigzag trajectory and a circular trajectory with better observability are designed for subsequent simulations.2.The distance information between the master and slave AUV is added to perform information fusion on the collaborative navigation system.The nonlinear filtering algorithm is selected to deal with the system nonlinear problems.Extended Kalman Filter is simple and easy to implement.It can achieve ideal filtering accuracy when the nonlinearity is not strong and is suitable for filter processing in three-dimensional scenes;Unscented Kalman Filter can handle systems with strong nonlinearity and the filtering accuracy is better when the dimensionality is not high.It is suitable for filtering processing in two-dimensional scenes.3.Aiming at the shortcomings of nonlinear filtering algorithm such as large nonlinear error,a master-slave AUV collaborative navigation method based on iterated Kalman filtering is proposed by introducing the idea of iteration.This method linearizes the measurement equation again based on the state quantity obtained by the measurement update until the desired accuracy is reached.Compared with the traditional nonlinear Kalman filter algorithm,this method has a certain improvement in accuracy and can effectively reduce the navigation and positioning error caused by the nonlinear error.It has validity and feasibility.