Research On Gyroscope Bias Estimation Method Based On Map Information

With the rapid development of the mobile Internet,location-based services are growing explosively,and the demand for positioning is increasing rapidly.Satellite positioning occupies a dominant position in many positioning methods.However,satellite positioning seems to be powerless for indoor or some extreme environments.Strapdown inertial navigation and positioning system is an addition of them,it is by wearing on pedestrian inertial sensors to collect data and obtain the pedestrian's position.The two most important problems of strapdown inertial navigation are:heading estimation and step length estimation.Currently,the most widely used method for step length estimation is detecting the foot land time by zero zero-speed detection algorithm and update the position and velocity by kalman filter.However,there is no good solution to the heading correction.It can be said that the heading drift problem is a key issue that restricts the further improvement of the strapdown inertial navigation positioning accuracy.For the problem that gyro error drift accumulates over time,a gyro error estimation method based on map information is designed.By analyzing the regions in the indoor map that are strongly constrained by heading,the angle information that can be used for gyro error estimation and heading correction is extracted,The gyro error estimation method was established by Kalman filter and quaternion method,respectively.The recursion relationship between the current Euler angle and the angular velocity deviation is established by the Kalman filter,and combined with the map and the zero velocity detection information,the current estimation value of the gyro error can be obtained.The quaternion-based gyro error estimation method establishes the functional relationship between the current output and the gyroscope deviation and the observed value.When the observable measurement is acquired,the bias of the gyroscope can be reversed.The experimental results show that the gyro error estimation method designed in this paper can dynamically detect the zero-speed drift of the gyroscope and self-correct the gyro.After the noise was estimated,the positioning accuracy was significantly improved.In the simulation experiment with a total step length of 311.89 meters,the total error is reduced by 81.37%and the average error is reduced by 81.41%compared to the noise path.The gyro error estimation method proposed in this paper only needs to analyze the current motion state and the heading angle information of the real position,and does not need to provide additional auxiliary means such as turntables in the laboratory environment,so it has great practical significance in disaster relief,anti-terrorism operations,shopping mall navigation,and some other everyday indoor.