Research And Implementation Of Fusion Algorithm Of UWB Positioning And Shoe-type Personal Navigation

With the continuous development of communication technology,the traditional single positioning method can no longer meet the needs of people for positioning in the complex indoor environment,the fusion of various positioning methods has become the current research trend.Ultra Wideband(UWB)has quickly become a research focus in recent years due to its characteristics of high accuracy,large capacity,and strong antiinterference ability.At the same time,UWB signals have a wide frequency domain and corresponding time-domain signals are very short.Therefore,UWB is suitable for positioning methods based on the measurement of signal flight time,such as Time Difference of Arrival(TDOA)/ Time of Arrival(TOA)positioning.However,positioning accuracy is significantly affected by non-line-of-sight and multipath effects,especially in complex environments.The Strap-down Inertial Navigation System is an autonomous positioning system based on inertial sensors.It does not require the assistance of external equipment.It only needs to use the inertial sensors to collect the acceleration and angular velocity during the movement.In combination with its own attitude solution algorithm,navigation and positioning can be completed.This feature makes inertial navigation technology well applicable to complex positioning environments,so it has also received widespread attention in recent years.However,the inertial sensor designed by MicroElectro-Mechanical System(MEMS)technology has low accuracy,and the internal gyroscope and accelerometer have non-negligible drift.During the process,due to the attitude solution is a continuous integration mode,the error will continuously accumulate,which will easily lead to the divergence of the algorithm and finally make the positioning failure.This paper will focus on the advantages and disadvantages of UWB positioning and inertial navigation,and carry out research on the fusion positioning algorithm based on ARM Cortex M4 single chip microcomputer.First,the theory of UWB positioning and inertial navigation is studied,and then a fusion algorithm based on extended Kalman filter is proposed.The inertial navigation data is used as the input value of the system prediction process,and the UWB data is used as the observation value of the system update process.Because the hardware platform has a limited amount of calculations,the extended Kalman filter needs to be simplified.Then,the gait detection technology and zero-speed correction algorithm for constrained inertial navigation are studied.A sliding window acceleration variance determination method is added to the traditional single-threshold gait detection technology to form a three-threshold gait detection technology.Improve the accuracy of gait detection.At the same time,designed a zero-speed correction algorithm based on the Kalman filter,which is used to correct the inertial navigation error during the stationary period of the foot to ensure the accuracy of the algorithm.Finally,the paper designed separately hardware of the UWB system and the inertial navigation system,and introduced the main part of the hardware.In addition,the paper analyzed the problems of theoretical algorithms applied to practical engineering,such as the UWB antenna delay and sensor error calibration,proposed solution and verified by experiments.Then the TOA fusion inertial navigation and the TDOA fusion inertial navigation are respectively verified by fixed-point experiments and motion experiments to ensure the effectiveness of the algorithm.