| The thesis has investigated two key technologies involved in strapdown inertial navigation system (SINS). One is the moving-based alignment through an inertial frame; the other is the depressing of navigation errors by rotation. The content mainly comprises of the following five aspects.1. A systematic study is performed on the inertial frame based alignment (IFBA) method, which uses the gravity vector and its derivative. By virtue of a cascade of low-pass FIR filters, IFBA attenuates the disturbing acceleration and maintains the gravity. According to the geometric relations among space vectors, the analytical expression relating the navigational errors to the instrumental errors is derived. In order to attenuate the disturbing of vehicles, an FIR filter group consisting of two decimations and three sub-filters is designed. Rapidness of the IFBA method is discussed, and the ability to conquer large misalignment is confirmed. Comparisons with other traditional algorithms prove that proposed method converges much faster than the conventional methods at no cost of precision and also works well under any large initial misalignment.2. According to the basic principle of the IFBA method, a maneuverable alignment algorithm for vehicle-based SINS with twice short stops is proposed. The relationship between the alignment error and the relative position error is illustrated. To satisfy the requirement of alignment during sailing, an iterative alignment method for ship-based SINS is designed and simulation verifies its feasibility.3. The performances of the IFBA method is tested by a navigation grade SINS (0.01/h, ring laser gyroscopes) and the azimuth aligning accuracy is 0.04 (1σ). For disturbing ground condition, experiments show that the IFBA method needs 200 seconds. And for swaying marine condition, the aligning time is not more than 300 seconds. Compared with the traditional algorithms, IFBA method has obvious advantage in alignment time and does not need coarse alignment stage. Vehicle experiments show that the maneuverability of the vehicle is improved by using the proposed algorithm. To deal with the aligning problem during sailing, an alignment scheme comprising a coarse step by the IFBA method and a fine step by a Kalman filter is presented. The ship-based alignment experiment proves that the proposed scheme works well for the condition with approximate constant speed. 4. After the IFBA method, the thesis investigates another key technology: the depressing of inertial navigation errors. The analytical expressions of navigation errors when rotating are derived and compared with the static results. The resonant frequency of the rotation scheme is indicated. In addition to single direction rotating, the principles of reciprocative rotating, intermittent rotating and multi-axis rotating are also investigated. The analytical error expressions of reciprocative rotating are derived. And the performance of intermittent rotating is examined through simulations. It shows that well-designed rotation can depress the influences of instrumental errors and of calibration errors remarkably. Advantages and disadvantages of the rotating schemes are analyzed, and the rules of the evaluation are presented for actual system design.5. A simulation model based on Matlab/Simulink is designed to prove the analytical results. Table experiments and simulations show that the performance of reciprocative rotation about the azimuth axis is optimal. A prototype with reciprocative rotation scheme is designed for a real naval ship. Using a SINS made up of 0.01°/h ring laser gyroscopes, mooring and sailing experiments show that the position error is no more than 3 nautical miles in 8 hours.
|
PDF Full Text Request