Design And Implementation Of TDOA Indoor Positioning System Based On Kalman Filter Algorithm

With the development of the service industry based on location information,indoor positioning technology has been increasingly concerned and studied.Due to the centimeter-level positioning accuracy,indoor positioning technology based on UWB is playing a more and more important role in the industrial application field.In UWB indoor positioning,the signal will inevitably be blocked by various obstacles,so the non-line-of-sight environment composed of various obstacles will reduce the indoor positioning accuracy.Therefore,it is urgent to solve the problem of non-line-of-sight error caused by obstacles in UWB indoor positioning.Aiming at the problem of non-line-of-sight errors,an error compensation method based on area division is used in the TDOA positioning system based on KF.Firstly,after measuring the position and size information of obstacles,an extension line connecting the base station and the endpoint of the obstacle is used to divide the positioning area,in this way the obstacle type between any point in a sub-area and a base station is the same.Tags at various locations are tested to judge whether the positioning results fall in the correct sub-region.Next,the fitting of the error of different obstacles to the measured value is performed to obtain the error fitting functions,and the fitting functions and the area division information are saved.Finally,in the actual positioning process,the method of combining KF and 3? criterion is used to optimize the measured values between the tag and the base station,and the preliminary positioning results is obtained by solving TDOA equations using Fang algorithm.The sub-region of the preliminary positioning result and the type of obstacle between the result and each base station are obtained through the table look-up method,these obstacles' fitting functions are used to compensate the measured value in the iterative positioning process.GTRS algorithm is then used to re-solve the tag position,and a new round of compensation is carried out.When the termination threshold is met,the iteration is stopped and this tag position is used as the final result.A set of wireless communication nodes based on STM32 chip and DWM1000 RF module is designed,and software development and functional testing are carried out on the hardware platform.The test results show that the positioning system has an average accuracy of 13.59 cm in a non-line-of-sight environment of9.46 m * 8m,which can effectively suppress the NLOS error and can be used for indoor positioning in non-line-of-sight environments.