Application Research On Initial Alignment Filter Technique Of Underground SINS And Positioning System

Coal energy occupies an im portant position in our country, but the safety condition of coal m ine is relatively weak. Existing underground positioning system does not provide real-time accurate location information and running track to rescuers, causing rescuers miss the best rescue time in the event of a m ajor accident. So it is necessary to improve the underground positioning system.Strapdown inertial navigati on systems which has the function of autonom ous navigation can not be inte rfered with the com plex underground environment, improving the accuracy of location data of personnel or equipment. Thus, this pap er will introduced SINS into under ground positioning system to im prove the lo w positioning accuracy. In the strapdo wn inertial navigation s ystem, initial alignment accuracy level has a s ignificant impact on the positioning accuracy. Therefore, this paper focuses on designing a m ore reasonable initial alignment filtering algorithm in the complex underground environment which can m aximize the accu racy of initial alignment in the case of reducing the am ount of calculation and filtering divergence, thereby increasing the efficiency of the initial alignment.Combined the specific application environm ent in coal m ine, based on the research, analysis and comparison of the initial alignment filtering algorithm such as traditional linear Kalman filter algorithm and EKF(Extended Kalm an Filter) algorithm in inertial navigation system, knowing that using adaptable nonlinear filtering UKF(Unscented Kalman Filter) algorithm can make higher filtering accuracy and faster conver gence. For the uncertain reas ons of non-Gaussian noise, random location errors and external dis turbances in coal m ine, we improve UKF algorithm and introduce the clos ed-loop filtering system to reducing the filterin g divergence while reducing the am ount of calculation al gorithm to improve the real-time of the system. In this paper, by optimizing and merging the filtering algorithms and inertial navigation noise error matrix equation to complete correcting data accurately in SINS initial alignment. What’s more, through c ontrasting the traditional linear Kalm an filtering, EKF and improved UKF to verify the availability and accuracy of improved optimization algorithm UKF under nonlinear conditions.