Study On Strapdown Algorithm And Its Realization By DSP Under High Dynamic Environment

The inertial sensors of the strapdown inertial navigation system (SINS) are directly mounted to the vehicle and the vehicle's high-frequency motions will affect the accuracy of the SINS. For example, the vehicle's angular and linear vibrations will result in the coning error, the sculling error and the scolling error, so perfect algorithems should be designed to reduce the errors. Thus, it is a key point to study the high precision algorithms for SINS in highly dynamic environments.Thinking from the engineering, to use the high-precision algorithms it will result to make the quantity of compute enlarge quickly, and then the updata frequency of system's attitude will reduce, with the depressing that will directly result to increase the error of coning and sculling. It has any effect to enhance the system's precision, that make a lot of nice algorithms can not to be realized in the engineering.With the flying development of information technology and compute technology that the digital signal processing had became the key technology subject. The DSP chip take Harvard structure that the data and the program independency, it have the special multiplication, and take the advanced address/data bus and multilevel pipeling mechanism. The high-effciency instruction set make the chip's instruction cycle below to 10ns. Those traits could make suit for the requiring of real time to the strapdown algorithms. It takes the possible that the high-precision strapdown inertial algorithms could apply in the engineering. Base on the DSP to research the high-precise strapdown inertial algorithms have some theory sense and large enginer value.In the thesis, the fundamental theory of the SINS is introduced first, including the principle, the attitude updating algorithms and the navigation calculations. The following investigations have been done:1. Firstly, it is analyzed that how the coning error occurs for SINS in highly dynamic environments. Secondly, the improved rotation vector algorithms in the pure coning motion are discussed and the computing burden is compared with each other. Lastly, computer simulations are executed to test the algorithms'performance.