Application Research Of UKF Applied In Underground SINS Initial Alignment For Large Misalignment Angles

China’s coal industry has an absolute advantage in the energy industry. Safety inproduction of coal is an important research topic. Existing coal positioning systemdoesn’t work so well that it can’t provide protection for safety production.What if somedangerous thing happens, the system can’t afford useful information for rescue.In order to overcome the disadvantages of existing coal’s positioning system,paper proposed inertial navigation system into the underground positioning navigation.Inertial navigation system effectively shield outside interference caused by the system,effectively improve the positioning accuracy in terms of personnel or equipment in theunderground environment. Strap down inertial navigation system initial posturealignment accuracy has a greater impact on the subsequent calculation. So paperfocused on the design of a rational filtering method in large misalignment, which wouldimprove initial alignment accuracy.Paper researched on the various existing filtering algorithms, which was appliedin inertial navigation system initial alignment. Considering coal undergroundsurroundings, it chosen UKF algorithm which was good at dealing with strongnonlinear system environment. Paper optimized UKF algorithm by considering theadditive relation between environment’s noise and system, as well as the specific formof the system equations. The optimization of algorithm could reduce the complexity ofthe algorithm, improve computational progress. Meanwhile, it also could reduce thecomputation time and improve the work of real-time systems. Otherwise, paperoptimized the error models, which would be fit for low speed and confined space. Thefiltering algorithms and inertial navigation error equations were optimized and mutualintegration. Then paper finished the initial posture alignment with large misalignmentangles. Finally paper made a comparisons between EKF algorithm and optimizationalUKF algorithm. By analyzing the data, it was proved that optimizational UKFalgorithm was addicted to variable nonlinear conditions with high accuracy.