Research On MEMS Inertial Sensor Array System And Data Fusion Technology

This design mainly introduces a series of related basic knowledge about MEMS inertial sensor array and data fusion technology,and solves a series of problems.Today,MEMS technology is becoming more and more widely used,and MEMS inertial sensors' low-cost,durability,small-size,easy-intergration,low-error making it occupy a very large market.MEMS inertial sensors mainly include MEMS accelerometers and MEMS gyroscopes,and sometimes also include magnetometers.As the technology matures,almost all mobile phones are now equipped with accelerometers,some with gyroscopes,or directly equipped with inertial sensors,and these inertial sensors are mostly MEMS inertial sensors.For example,Apple's iPhone6 mobile phone integrated InvenSense six-axis inertial sensor MPU-6700(integrated MEMS gyroscope and MEMS accelerometer),Bosch's three-axis accelerometer BMA280 and AKM's AK8963C magnetometer.In addition to the phone there are many other electronic devices equipped with MEMS inertial sensors,such as VR,game handles,cars,aircraft,spacecraft and so on.In the case of increasing demand,MEMS inertial sensor error problem becomes more and more important.Based on this iussue,this design developed the sensor array data fusion technology,and it can effectively reduce the MEMS inertia sensor's error.Around this technique,the author builds a multi-sensor array platform and uses MATLAB software as an analysis and modeling tool(Chapter 2).Allan's variance analysis is used to identify the noise of the sensor,and it is used as a method to determine the quality of the data fusion model(Chapter 4).By constructing the state equation and updating equation,the author establishes three Kalman filters for data fusion(Chapter 6),and tested the three models by experiment(Chapter 7).Among them,Allan's variance analysis can be used to identify the underlying stochastic process of data noise.In addition to the above works,the author also summarizes the common sources of MEMS inertial sensors' error(Chapter 3),introduces the theoretical basis of ARMA model and gives examples of application of ARMA model in MATLAB software(Chapter 5).In the final chapter of the article,the author makes a summary and prospect of what has been done.