Research On Gyroscope Drift Modeling And Compensation Based On Particle Filter

Gyroscope is the core of Inertial Measurement System. With the continuous improvement of gyroscope's accuracy in Inertial Measurement System, it is increasingly important to research the gyro drift. Compared with traditional electromechanical and photoelectric gyroscopes, Micro-electro-mechanical gyroscope (MEMS gyro) has the advantages of the low cost, small size, light weight and high reliability. At present the application of micro-gyroscope is still constrained by its low accuracy, only for low-end navigation. In order to improve the precision of a gyroscope, one of the fast and reliable methods is analyzing the error characteristics, making models and compensating for them.Gyro drift includes systematic and random errors, and the latter one is the main error source. The estimation problem on the background of intelligent vehicle attitude determination and inertial navigation can be viewed as the nonlinear systems. The widely used method Currently is Kalman filter, which is a computer-implemented real-time recursive algorithm. Kalman filter can be used for dealing with random signal, and it has outstanding performance in real time signal processing. But the Kalman filter needs to preprocess the signal into linear, Gaussian and stable signal, which can be difficult and may lead to divergence.Particle filter realize recursive Bayesian filter via Monte Carlo simulation. The method is suitable for any non-linear system that could be represented with state model. It is more practical than conventional Kalman filter, and its precision could approach optimal estimation. Particle filter is flexible and easy to be implemented. At present, it has become a research hot of the target track, the signal processing, the inertia navigation and other fields.In this thesis, combining with the background of intelligent vehicle inertial navigation, I choose low-precision piezoelectric vibrating Micro-electro-mechanical gyro chip. Because this type of gyro can not meet the requirements of intelligent vehicles'attitude determination and navigation in practice, I propose that we can adopt particle filter technique to reduce static gyro drift and improve the accuracy of it. Firstly, I combine with circuit designation and the basic principles of MEMS gyro, and use Allan variance algorithm for identify the types of gyroscope's drift error. Secondly, establish time-series model and use Kalman filter for reduce static gyro drift. Thirdly, study in detail the principles of particle filter, and establish the state equation and measurement equation, then particle filter technique used in gyro's output data. Experimental results show that the design of hardware and software filters can effectively reduce the static drift of gyro chip, and particle filter method for nonlinear signal processing is better than Kalman filter. So this paper provides a new way of thinking on applying low-precision micro-gyroscope to high-end fields.