| Global navigation satellite system (GNSS) has caught broad attention because of its high accuracy, global, all-weather, continuous. It has great potential in attitude measuring, either takes the place of the complex and expensive attitude measuring equipment as a new attitude measurement system, or constitutes integrated attitude measurement system with traditional attitude sensor. In recent years, attitude determination through the use of GNSS signals is a new field of applied research of GNSS. The key to the attitude determination using GNSS lies in the algorithms of fixing integer ambiguity and attitude determination. This dissertation mainly focuses on the two issues, the main work and contributions are summarized as follows:1. The three GNSS attitude measuring mathematical models were established, including the single baseline mathematical model based on baseline parameters, the multi-baseline mathematical model based on baseline parameters, and the multi-baseline mathematical model based on attitude parameter, which provided a foundation for fixing integer ambiguity and attitude determination.2. For the GNSS attitude measuring mathematical model based on baseline parameters, choosing Euler angle/quaternion as attitude parameter, two methods of attitude determination were proposed based on multiplicative error Euler angle/quaternion. For the mathematical model based on baseline parameters, choosing direction cosine matrix as attitude parameter, the method of attitude determination was proposed based on direction cosine matrix. As simulation results, the accuracy and efficiency of the two methods based on multiplicative error Euler angle/quaternion are improved than those based on additive error Euler angle/quaternion, and the accuracy and efficiency of those three methods of attitude determination based on multiplicative error Euler angle/quaternion, direction cosine matrix respectively are comparable.3. For the GNSS attitude measuring mathematical model based on attitude parameter, selecting direction cosine matrix as attitude parameter, a method of attitude determination were proposed. Compared with the algorithms of attitude determination for GNSS attitude measuring mathematical model based on baseline parameters, the estimation accuracy and efficiency of are both improved.4. According to the prior constraint information in the three GNSS attitude measuring mathematical models, the corresponding virtual observational equations were provided. By increasing the virtual observational equation into the GNSS attitude measurement model, the efficiency of fixing ambiguity is improved.5. A GNSS multi-baseline mathematical model based on error quaternion was established, the float solution of integer ambiguity can be solved based on iterative least squares estimation, while the fixed solution of integer ambiguity can be also obtained by using the LAMBDA algorithm. The efficiency of fixing ambiguity is improved.6. By using the constraint information of direction cosine matrix, a new integer ambiguity test method was proposed, and a reasonable threshold was provided for reference. The new method can verify the solution of integer ambiguity is correct or wrong in one epoch or a few epochs, which is high efficiency than the past test methods.7. Denoting the attitude parameters by direction cosine matrix to build GPS/Gyro Attitude Determination System, and a matrix Kalman filter method was proposed. As simulation results, with small gyro noise, the accuracy of those two methods of attitude determination are comparable, with big gyro noise, the accuracy of direction cosine matrix Kalman filter is higher than multiplicative error quaternion Extended Kalman filter.8. Selecting the vector form of direction cosine matrix quaternion as the attitude parameter, GPS/Gyro Attitude Determination System model in vector form was established, and the Kalman filter can be used. Aiming at these problems of initializer, fixing integer ambiguity and other parameters, relative resolving methods were put forward. |
PDF Full Text Request