Research On IMU/Wi-Fi Indoor Location Method Based On Pedestrian Gait And Turn Analysis

With the continuous diversification of user lifestyles,people’s requirements for location information services are also increasing.Today GNSS outdoor positioning technology is very mature,while indoor positioning technology is still in the development stage.So far,no general indoor positioning method like GNSS has been found.Although there are high-precision positioning methods such as UWB positioning,ultrasonic positioning,and visual navigation in the mainstream positioning technology,they are either expensive or require a large number of base stations,which is not conducive to widespread promotion.Since Wi-Fi has spread all around our lives,there is no need to establish additional base stations,and now Wi-Fi is embedded in smart phones,so Wi-Fi is widely used in indoor navigation and positioning systems,but Wi-Fi positioning However,due to multi-path effects or environmental fluidity,the positioning accuracy is poor;as a relatively new device,the inertial measurement unit(IMU)has been favored in navigation and positioning systems in recent years.It can obtain high accuracy in a short time.The positioning information does not need to rely on external signals at the same time,but there is a problem of cumulative error,and it is impossible to achieve long-term positioning.Therefore,it is difficult to balance positioning accuracy and stability using a single positioning technique.In view of this situation,this paper proposes an IMU/Wi-Fi indoor positioning method based on pedestrian gait recognition and turn recognition,which recognizes the advantages of high-precision relative positioning in a short time and the strong feature of Wi-Fi at indoor turns.Ability to combine.The research content mainly includes the following aspects:Aiming at the problem of IMU cumulative error,a zero-speed correction method of threshold method is proposed.Firstly,the pedestrian gait feature is acquired through the IMU sensor.By analyzing the pedestrian gait feature,the zero-speed state at the moment when the foot touches the bottom is analyzed,and then the threshold method is used to filter the lower speed of the IMU and determine that the speed is zero at that time;Based on this,the pedestrian motion parameters are corrected and updated using the extended Kalman filteralgorithm,thereby eliminating accumulated errors.Aiming at the problem of large turn error caused by yaw angle error,an improved Wi-Fi RSSI fingerprint positioning algorithm is proposed to assist in updating and correction.Use the peak detection method of the gyroscope and the indoor map structure to detect the turning time of pedestrians,and combine the magnetometer data to detect the turning direction;at the same time,use the time stamp to synchronize the time of the IMU and Wi-Fi data,and then use Kalman filtering to Data fusion to improve positioning accuracy.In order to verify the above positioning method,this paper has done two sets of experiments to test the effectiveness of the method in the horizontal and vertical directions.Through comparative analysis,it is found that the positioning accuracy of the combination is better than the single source positioning accuracy whether it is in the horizontal direction or the vertical direction,which verifies the effectiveness of the positioning method.