Gyro/GPS Integrated Attitude Measurement Method Based On Trajectory Environment

The development of low-cost controlled missile is one important way to improve firing accuracy of conventional gun weapon systems. In order to improve the guided projectile flight attitude measurement precision then to achieve accurately trajectory control, this paper focuses on some of the basic theory and problems encountered in the low-cost MEMS gyroscope attitude measurement and attitude nonlinear filtering techniques, combined with ballistics theory to analyze and research.This paper firstly established a model of rigid body motion of the guided projectile. By the numerical calculation of the model to obtain a theoretical trajectory, the theoretical trajectory on the one hand provide the basis for the simulation of the onboard sensor measurements, on the other hand, also as an attitude comparison benchmark of estimated and calibration values considering measurement error, to assess the precision of the calibration and filter.Aiming at the low-speed rotary projectile ballistic characteristics, this paper presents an inflight calibration method based on the assumptions of the approximate attack angle with dynamic balance angle and zero. According to the established trajectory model, motion characteristics of low-spin tail projectile angle of attack were studied. Make full use of the ground ballistic test data and flight measurement data, the approximate angle of attack by the dynamic balance angle was used to solve the bullet attitude, and then attitude calibration mathematical model is derived under the assumption. Also based on the assumption of zero angle of attack, establish the missile simple flight attitude of the calibration model, and contrast the precision of the models through simulation. In the trajectory of gliding, the initial calibration method of attitude angles has been proposed by consider the impact of the balanced angle of attack. The article also analyzed the gyro integration algorithm for the whole trajectory and a method to suppress gyro integral accumulated error with attitude calibration.The nonlinear attitude filter model has been built with the angle of attack, sideslip angle and roll angle. With the angle of attack, the sideslip angle and roll angle as the state variables, the nonlinear attitude filter estimated state equations and measurement equations were established combined with ballistic parameters and the MEMS gyroscope and GPS measurement data. The models take full account of the variability of the trajectory parameters and gyro random measurement noise. Optimal algorithm for attitude estimation model has been obtained by comparing a variety of numerical simulation results.Researched on MEMS angular rate gyro general principle and measurement error causes, using the time series method for gyro random error modeling and compensation, removed the periodic terms and the trend term in the noise, this article has received a normal distribution random noise sequence in order to take advantage of filtering methods to compensate. Putting forward the idea of using wavelet neural network modeling and analysis of the gyro output data, a high-precision of predictive value for gyro output has been given through simulation modeling. Using of wavelet filtering and multi-resolution decomposition theory for the gyro output noise filter compensation, simulation experiments have been made to obtain the optimal wavelet basis function and scaling function.Semi physical simulation research has been conducted under laboratory conditions. Using three axis flight attitude simulation platform and MEMS angular rate gyro designed simulation test method. According to the theory of ballistic data as control parameters, the gyro output data was sampled and pretreated. The results were brought into the attitude calibration model and nonlinear filtering estimation model, and it obtained the satisfactory attitude calibration and estimation results. The semi-physical simulation test verified the validity and accuracy of the attitude calibration model and filtering the estimated model.