Research On Tightly Integrated SINS/DVL Navigation System Methed

With the development of modern science and technology, the activities of human exploration on marine resources accumulated continually. Various underwater vehicles will be applied widely, and then the requirements of accuracy and reliability of underwater navigation system will get higher. SINS/DVL integrated navigation system is very important in surface or underwater navigation due to its superior performance, and has been widely used in underwater vehicles. In order to improve the accuracy of existing SINS/DVL integrated navigation system,a tightly integrated SINS/DVL navigation system is designed in this paper, which makes the data from DVL and SINS consistent in space and reduces the errors of DVL assisting SINS integrated system.Firstly, the method of reducing velocity measure error of DVL and the guidelines of selecting measurement parameters are explained in this paper. Then the factors affecting the errors of integrated navigation system are analyzed. Moreover, in order to estimate DVL installing errors,filtering model of installing misalignment angles is established,then the differential GPS is utilized as reference speed under the condition of ship sailing, and then Kalman filter algorithm is used to estimate the installing misalignment angles. The effectiveness of this method is verified by computer simulation. As for the DVL velocity measure errors caused by ship rocking, the strapdown matrix outputted by SINS is used to project the velocity on IMU coordinates, obtained by utilizing the 4-beam projection method,onto navigation coordinates. That achieves the purpose of DVL and SINS assisting each other and data's consistence in space, thus the measurement accuracy is enhanced.Secondly,the design of a tightly integrated SINS/DVL navigation system is proposed.CKF and UKF are utilized for data fusion in the proposed integrated navigation system. The nonlinear error equation of SINS is derived in this paper, based on which the filter model of SINS/DVL integrated alignment is established. Performance of CKF and UKF are tested in the proposed integrated navigation system under both static base and moving base.Meanwhile, the accuraty of tightly integrated method and non-tightly integrated method are also compared under the moving base. Simulation results confirm the good performance of tightly integrated method.Furthermore, in order to avoid the position parameters from diverging in the integrated navigation system, external device is utilized to provide position measurement information,apart from SINS and DVL. Then the nonlinear filter model of SINS/DVL integrated navigation system is established based on the measurements. Simulation results under the conditions of long endurance and curved movement show that the proposed tightly integrated method can effectively improve the accuracy of SINS/DVL integrated navigation system,outperforming the non-tightly integrated method. Moreover, simulations also demonstrate that CKF is better than UKF in filtering estimation of the proposed SINS/DVL integrated navigation system, when estimate accuracy of the navigation parameter errors and computing execution time are compared of the two filters.Finally, the fault detection techniques of the tightly integrated SINS/DVL navigation system are studied. Performance of state and residual test methods in the proposed integrated navigation system are compared in the simulations and the results show that the residual x2 test method outperforms the state x2 test method in sensitivity. Moreover, a tightly integrated SINS/DVL navigation simulating platform is designed based on the MFC technology,which realizes the simulation, calculation and real-time visual display of navigation data.