The Method For Inertial Platform Testing And Parameter Identification

With the development of inertial navigation, there comes increasingly strict requirement for the accuracy of inertial platforms. Limited by the processing and manufacturing level, however, the improvement on inertial navigation system accuracy merely by increasing the precision of inertial devices such as gyroscopes and accelerometers has become more and more costly. By contrast, a more economic method is the system level compensation based on inertial platform error sources identification, which is the subject of this paper. The following work is carried out.A six-position testing method for inertial platform is presented, under the gravity field, the configuration of the system and the concrete steps of test are given too. The positions are given by three-axis test-table. The inertial platform works in those positions which given by three-axis test-table, the inertial platform works in inertial stability status. Six-position scheme has been chosen that is often used in the project. The method is effective.The model of inertial platform error is derived. With the error models of gyroscopes and accelerometers available, a complete linear output equations of inertial platform error are derived, according to the platform stabilization principle and using the outputs of the accelerometers as the measurement values. The model has some points such as explicit physical meaning, liner and simple structure.The method of identifiability and correlation is presented to platform error parameters, and a new platform model is built. At first, the least-square model of platform has been established, a least square method has been used in the platform error parameters estimation problem, as follows, some error parameters are deleted according to the discussion of identifiability and correlation, output equations that includes twelve error parameters are built at last.A series of simulations and experiments are carried out to the factors that affected the definition of error parameters, such as the testing time and sampling groups. Main Changes in relations are found out, according to the discussion, that how those factors impacts the definition of error parameters. At the same time, the impact ,brought out by small-angle approximation, is discussed. According to lots of simulations and experiments, the changes in relations are found out , which of the difference between the estimated value of error parameters in the model of small-angle approximation and the true value with testing time.