Gyroscope Bias Modeling Based On Multi-source Fusion

In the development of navigation and positioning,since the inertial sensor itself has many advantages over satellite positioning,its demand is getting higher.However,inertial sensors have an unavoidable disadvantage,the output noise has a large influence on the results.Although its own error as a low-cost sensor is not large,its error will accumulate during the solution process,so that when only the inertial navigation is used for positioning,the positioning result will gradually deteriorate until it collapses.Therefore,whether the noise effect of the sensor can be effectively solved is a key issue for improving the accuracy of inertial navigation.In this paper,a method of combining auxiliary information source and inertial navigation system to reduce the influence of inertial sensor output noise on position ing results is designed.And for the auxiliary information source,satellite navigation is used to provide position assistance information,and the map and the magnetic field provide attitude assistance information.Different schemes for reducing the influence of inertial sensor noise are designed for different auxiliary information sources.Firstly,in the scheme of the fusion of satellite navigation and inertial navigation system,the problem of unifying the space-time reference between satellite data and inertial navigation data is solved.On this basis,the initial large-angle uncertain heading of the inertial navigation system is estimated,and the positional auxiliary information is used to correct the cumulative error of the inertial sensor.Secondly,in the scheme of map information and inertial navigation fusion,the long straight corridor heading information in the map is used to estimate the deviation of the gyroscope.Finally,in the scheme of fusion of magnetic field and inertial navigation,the geomagnetic field information is used to estimate the navigation direction,and the deviation of the gyroscope is estimated.The experimental results show that the satellite navigation and inertial navigation fusion system can effectively estimate the inertial initial attitude and correct the error accumulation and trajectory drift of the inertial navigation system.It also verifies the robustness of the system for long-term operation at low satellite navigation observation frequencies.Secondly,based on the scheme of map information and inertial navigation fusion,the deviation of the gyroscope can be accurately estimated.Under fixed bias,the gyroscope's z-axis bias accuracy can reach 84%.For linear deviations,60%accuracy can also be achieved.Finally,based on the fusion scheme of magnetic field information and inertial navigation,the carrier heading can be accurately estimated,and the estimation is accurate in the gyroscope bias estimation.In this paper,based on the method of multi-source fusion to reduce the influence of sensor noise,several schemes are designed to have certain applicability in long-term positioning,indoor positioning and heading correction.