Key Issues In SINS/GPS Integrated System For High-Precision Measurement Application In High Dynamic Environments

Nowadays, the socio-economic development of our country is rapid and its international competitiveness is gradually being enhanced. Meanwhile, demands are also increasing for the parameter measurements of carrier motion in the fields of engineering and technology. These demands mainly include measurement accuracy, long-term stability and dynamic range, etc., and the corresponding engineering applications involve the motion error compensation of high-resolution earth observation system, the real-time measurement of large ship hull dynamic deformation, and so on. Currently, considering for the accuracy performance, it is very difficult for a single measurement system to meet the requirements in practical engineering applications. Therefore, it is of great practical significance for the following methods without increased hardware cost. In which advanced theories and algorithms are fully used to integrate Strapdown Inertial Navigation System (SINS) and Global Positioning System (GPS) into a high-precision integrated measurement system, where SINS is the major part.For actual carrier motions, the properties often occur mainly include dramatic changes during angular and linear dynamics, strong burstiness, random uncertainties, and so on, especially in high-precision applications. So the primary problem for SINS/GPS integrated measurements is that how to ensure the accuracy of a complete system in high dynamic environments. In this dissertation, the object of study is the SINS/GPS integrated system in high dynamic environments, and the goal of study is to improve the overall accuracy of integrated measurement. Overall, an in-depth study is carried out in this dissertation, which consists of two aspects:the applicability of SINS algorithm and the data fusion of integrated system, and the following are the four main points.(1) Optimization design for strapdown screw algorithm based on dual quaternion. First of all, a differential equation about screw vector is derived after the elaboration of the concepts and properties of the basic tool-dual quaternion. Then, a framework for the corresponding numerical integration algorithm is given using the multiplication chain rule of dual quaternion. Combined with the understanding of the classic coning and sculling motions, a special screw motion is obtained from the expression of the classic one through the use of simplification and analysis. The corresponding optimization design is done for the core part of the numerical integration algorithm-screw algorithm. During the optimization process, two types of sensor signals are involved (incremental and instantaneous signals). Furthermore, the phenomenon is analyzed that the digital simulation results are not entirely consistent with the residual errors calculated in the optimization, and a possible cause is pointed out.(2) Deep research on the disagreement problem between the theoretical analysis and the simulation results presented in (1). In this study, three traditional optimal coning algorithms are used as object for brevity, which are constituted with angular incremental signals, hardware-enhanced angular rate signals, and angular rate signals, respectively. Moreover, the ideal value of the change of rotation vector during attitude update period is derived using inverse deduction under classic coning motion. For a particular condition of carrier motion, the degrees of nonperiodic and periodic components deviated from their ideal values are analyzed in three types of algorithms.Accordingly, comparison and classification are applied to the concrete mutil-sample algorithms. Some guidelines are given through the analysis of algorithm verification results, which can be referenced for practical selection.(3) The improvement of three types of traditional optimal coning algorithms for periodic components. Combined with the forms of traditional algorithms, we give the concrete structures of the improved ones. Then, the overall optimization goal of the improved algorithm is determined using the ideal value of the change of rotation vector obtained above. Based on the proposed error criterion for the second optimization, we solve the compensation coefficients in algorithm expressions. Considering the inertial sensors with different precision levels, simulations are designed under a variety of conditions to testify the algorithm performance under high dynamic angular motions. Test results confirm that the designed algorithms are rational and veracious.(4) Data fusion method for SINS/GPS integrated system in high dynamic environments. Noticing the high-dimension feature of the states in SINS/GPS integrated system, a nonlinear filter is introduced named Cubature Kalman Filter (CKF).Then, considering the problem that "high dynamic environments likely lead to the appearance of the uncertainty in system model", further improvement is made to the structure of the strong tracking CKF already existed. Based on the SINS nonlinear error model for large azimuth misalignment angle, we derive the state equation and measurement equation of a loosely-coupled SINS/GSP system. Fulfilled with the existed results about the observability degree of state variables, SINS/GPS nonlinear integrated filter strategy is constructed based on the observability degree of state variables and the improved strong tracking CKF. Finally, simulations under the relevant high dynamic carrier motion and semi-physical flight test data are used to verify the proposed method.