| With the rapid development of satellite Internet,the miniaturization of satellites has gradually become a development trend.The application prospects of microsatellites are becoming wider and wider,and they have become the current research central issue of satellite technology.For the actual networking applications of microsatellites,due to the large number of satellites,how to reduce the cost of satellites and how to improve the integration of satellites are all key issues.In the satellite inertial navigation system,three sensor fusion complementary methods of gyroscope,accelerometer and magnetometer are usually used to calculate and obtain the optimal attitude estimation of the satellite.Among them,the gyroscope is the key sensor for measuring the satellite angular rate.At present,fiber optic gyroscopes are widely used in satellite systems.Fiber optic gyroscopes are expensive.There is an increasing demand for low-cost sensor solutions.Due to the rapid development of MEMS technology,more possibilities are provided for the technological upgrade of micro-satellites.MEMS devices have many advantages,such as small size,low power consumption,and low cost,and are more suitable for the design requirements of micro-satellite integration,modularization,and short development cycle.However,compared with fiber optic gyroscopes,MEMS gyroscopes have high noise and low accuracy,and cannot meet the accuracy requirements under specific conditions.This paper designs a set of multiple MEMS gyroscopes combined system to improve the accuracy of MEMS gyroscopes.The main controller of the system uses STM32,and the MEMS gyroscope uses six identical MPU6050 s.The voltage regulator circuit and other functional circuits are designed and manufactured.The PCB board is tested.The STM32 main controller uses SPI serial communication to receive data collected by six MEMS gyroscopes in real time at high speed.Uses C language to write algorithms for online real-time filtering calculations and passes the best estimated posture data through the CAN bus in the computer system transmitted to the satellite.The function of the virtual gyroscope is realized through the data fusion of multiple gyroscopes.In the filtering algorithm of the system,in order to improve the performance indicators of the virtual gyroscope,the output of a single gyroscope is selected for noise analysis,based on Allan variance analysis and classical signal analysis methods.Assuming that there is a good correlation between the noise processes of the gyroscope signal,a simple random model of the random bias component in the gyroscope output signal is established and implemented it in MATLAB.In order to reduce the complexity of the algorithm and improve the dynamic performance,the first-order autoregressive model was used to establish the prediction model of the system.Six gyroscopes are used for zero point calibration of initial conditions,and Kalman filter is used to process data fusion algorithm for multiple sets of data.When designing a Kalman filter,choosing different parameters will change the performance of the filter.In order to further improve the static and dynamic performance of the gyroscope,an adaptive Kalman filter algorithm is proposed.Through the real-time feedback update of the Q array,the dynamic parameters of the system can be changed,which can slightly improve the static performance of the system and greatly improve the system’s ability to adapt to dynamic changes.Use a high-precision turntable for actual testing.First,perform static data test,and then perform dynamic data test.Validate the effectiveness of the algorithm in different situations by given angular rate input with different characteristics.After real-time data processing,the output result of the filtering system is obtained.By calculating the variance of the output error of a single MEMS gyroscope and comparing the variance of the virtual gyroscope,the experimental results show that the algorithm can better improve the accuracy of the gyroscope. |
PDF Full Text Request