Pedestrian Dead Reckoning System Based On MiMU

In the city of skyscrapers,the indoor environment is increasing,the traditional navigation based on satellite has been facing many problems,such as signal lost or not stable,and it is hard to meet the needs of people orientation.The indoor positioning system based on wireless technology belongs to the active positioning system,which should installed the signal source equipment in advance,and the accuracy of wireless orentation depends on the signal source's characteristics so that it is difficult to orientate in unknown environment.Pedestrian navigation system could orientate indepently when in the closed environment without satellite signals.It can be effectively used in emergency rescue,jungle adventure and military fields,and in recent years it has become a research hotspot in the navigation field.The goal of Pedestrian navigation system is to provide the real-time positioning information of people in different environment,according to this,this paper proposed a Pedestrian Dead Reckoning system(PDR)which based on Micro Inertial Measurement Unit(MIMU),the system integrates the MIMU and foot pressure sensor to become mutual support,and combinate the Dead Reckoning algorithm(DR)to realize autonomous localization.The main work of this paper includes the following parts:First of all,introducing the theoretical basis of system,including the basic algorithm of strapdown inertial navigation system,the basic principle of DR algorithm and Kalman algorithm;Aiming at the problem of gait detection,the step detection algorithm based on acceleration peak and the zero velocity detection algorithm based on inertial data are studied,and both of the two detection algorithm is verified,compared and analyzed by experiment.Secondly,with the pedestrian height and step frequency,construct a linear model to estimate step length;and obtain the model parameters by fitting calibration with experimental data;It is show that the error of step length by estimated is only a few centimeters;To solve the heading drift problem,a heading updating algorithm by the combination of gyroscope and magnetometer is proposed.Make the heading error as the observation of EKF to update and modify the heading.Finally,combined with the new algorithms,construct the overall framework of the system,and calculate the positioning error and accuracy of the system.With the three experimental data,it can be seen that the difference between calculated distance and real distance is little.Through the experiment,it is shown that the largest error of indoor positioning is less than 1.5 meters,the outdoor positioning error is increasing with the distance,the largest distance error is 12.496 meters,the total route is 6.599%.On the whole,the construction of PDR systems meet the positioning requirements for pedestrian navigation within a short period of time,it can realize real-time autonomous orientation.