Research On Pedestrian Dynamic Step Length Measurement System And Positioning Application Based On Ultra Wideband Radar

With the rapid development of navigation and positioning technology,people have more and more demand for location services in their daily lives.Inertial positioning technology has attracted much attention because it does not rely on external signal sources and can work completely independently.It is often used in pedestrian autonomous positioning applications in indoor,tunnel and other environments.Pedestrian Dead Reckoning(PDR)algorithm is a basic algorithm used in pedestrian inertial navigation and positioning,in which heading and step length calculation are its core technologies.However,the conventional PDR algorithm estimates the step length through the model,which has poor accuracy controllability and is easy to diverge.The long-term battery life will cause the positioning accuracy of the system to diverge due to the accumulation of errors.The reason is that the PDR algorithm calculates the step length model-based estimation,not single-step actuarial calculations.In order to fundamentally solve the insufficiency of the principle of the PDR algorithm,and adapt to the walking pattern of complex,highly dynamic,multi-behavior characteristics,improve the step length accuracy,and reduce the cumulative error,this thesis designs a pedestrian dynamic step length based on ultra-wideband signals.The measurement system accurately measures the step length and realizes the step length actuarial calculation under any gait and any step change.While reducing the accumulation of errors in the PDR algorithm over time,it achieves precise positioning in any walking state.The research involves precise step measurement,time synchronization and high dynamic gait discrimination.The specific research work is as follows.1.Combining the characteristics of the displacement trajectory of pedestrians,a footprint model is constructed;based on the footprint model,a dynamic step measurement scheme based on ultra-wideband is designed,and a pedestrian dynamic step measurement system is built.The six-position test method is used for error calibration in order to reduce the deterministic and random errors of the inertial sensor.The algorithm is modified to compensate the singnal delay of the ultra-wideband sensor.Low-pass filtering is used to suppress the high-frequency noise that exists in the step length data.2.According to the foot motion characteristics of pedestrians,a multi-threshold detection algorithm for feet support and an adaptive multi-step speed detection method for feet support are proposed.Corresponding inter-foot step length models are proposed for different walking modes,and the corresponding step length formulas are deduced;combined with support moment detection and step length formulas,a pedestrian dynamic step length measurement algorithm based on ultra-wideband is proposed.Experimental results show that the accuracy of the adaptive two-foot support detection algorithm is 99.4%,the step error of the combined algorithm when going straight and turning is 1.86%,and the step error of turning is 1.73%.3.Based on the dynamic step measurement system,the long step and short step experiments are carried out under different walking speeds.Compared with the traditional step length algorithm,the errors calculated based on the combined algorithm under slow/medium/fast(short step,long step)conditions are reduced by 50.52%,31.29%,45.45%,31.34%,60.96% and 38.68% respectively.The average positioning error of the inertial positioning system has been reduced by 68%.