Research On Self-Calibration Technology Of Platform System Based On Optimal Roll

Improving the calibration accuracy of the platform inertial navigation is of great significance to improving the accuracy of missile guidance.This paper studies the platform system self-calibration technology.This technology uses the characteristics of the three-ring three-axis platform system and only depends on the rotation of the platform body relative to the base,and assists when necessary.With external information,the key parameters of the platform system are calibrated with high accuracy under the optimal rollover situation.Although it works efficiently,the model establishment and experimental design are more complicated.In this paper,self-calibration test scheme and error analysis,continuous roll self-calibration error model,observable optimal roll path scheme,platform self-calibration test parameter identification and accuracy verification and other related issues are studied.First of all,according to the working principle of the inertial navigation system and the desktop inertial navigation system,two important self-calibration schemes are compared and discussed.The more efficient continuous roll self-calibration method is selected to design the test,and the type and source of errors in the continuous roll test are analyzed.Combined with the static error model of the inertial device,a more comprehensive error scheme is designed.Secondly,in order to establish the continuous roll self-calibration error model,based on the completion of the platform’s own error model,with the aid of the attitude error equation,the dynamic equation and the output equation to complete the continuous roll model,the principle of this model is clear and apply.Set reasonable simulation conditions and roll paths for the established model,and perform simulations,and verify the correctness of the model by using the accelerometer simulation method.Then,in order to analyze the observability of the system,the observation matrix is calculated by using piecewise linear constant theory(PWCS).The full rank of the matrix proves that the system is observable under a certain rolling path.In order to analyze the identifiability of each error parameter,spectral conditions are used.The numerical method calculates the observability of each state parameter at this time,and theoretically shows that each parameter can be identified;in order to design the optimal roll path,the determinant of the system information matrix is taken as the objective function,the angular velocity is added as the control variable,and the design is designed with constraints In order to improve the efficiency of solving the optimal problem,the particle swarm optimization algorithm(PSO)is used,and a variety of constraint methods are tested.By comparing multiple sets of test results,it is proved that the absorbing boundary method is more suitable for PSO;to ensure the convergence of the algorithm,by calculating the convergence area and reasonable configuration parameters,the optimal solution of the roll path is obtained.Finally,in order to complete the filter identification of the error parameters in the optimal roll path,the Kalman filter identification method suitable for linear time-varying systems is adopted.By analyzing the simulation diagram and the identification results,it is found that the platform self-calibration test designed in this subject can stimulate all error parameters.It can complete the identification task,not only has better identification accuracy,but also has better convergence time than the multi-position self-calibration method;in order to further verify the calibration accuracy,the modulus error method is adopted,and the calibration accuracy is found to be excellent by analyzing the standard deviation of the modulus error.