Research Of Indoor Positioning And Navigation System Based On Multi-information Fusion

Since the rapid development of the Internet,handheld smart devices get a widely popularized,people's lifestyles have a sharp changing.With lifestyle changes,the demand for indoor navigation and positioning system(INPS)is also grown.Because of the market request,indoor positioning technology has been widely researched.Currently,in the indoor positioning systems,techniques such as RF signal,ultra-broadband,Micro Electro Mechanical System(MEMS)inertial sensor,Wi Fi and Bluetooth are frequently used^[1].At present,the main research direction of the indoor positioning system is how to reduce time cost and economy cost while ensuring positioning performance and accuracy of the system.Because each technology has both advantages and disadvantages;in addition,the system performance also deeply depends on the application circumstance.Consequently,it is hard to provide an indoor positioning system with a single technology that has low-cost but high-precision,low-volume but high robustness.Therefore,the combination of several technologies into a variety of indoor positioning technology is a considerable solution,and multi-sensor fusion technology is the core method.This thesis creatively proposes a design project of indoor positioning and navigation system that combines inertial sensors,Bluetooth sensors,and environmental information.The based method of this positioning and navigation system is indoor positioning method based on inertial sensor.And the core positioning technology of indoor positioning method based on inertial sensor is pedestrian dead reckoning(PDR)algorithm.It mainly includes three core contents:one,using peak detection method with multi-conditional constraint detects the pace;two,using step detection method based linear model to detects length of the step;three,using coordinate conversion method detects the pedestrian heading.In order to improve the positioning accuracy and stability of this positioning method based on small amount of cost,this thesis mainly proposes the indoor environment information correction method based on process fusion and the Bluetooth sensor assisted positioning method based on result fusion.To solve the problem of the positioning error caused by inertial sensor's poor accuracy and temperature sensitivity,this thesis proposes an indoor environment information correction method based on process fusion.Specifically,in response to the requirement of low-cost.The landmark Bluetooth information is used to draw the indoor road network map to collect and save indoor environmental information.In order to further improve the accuracy of heading detection,the sliding-window mean filtering method is used to process the accuracy of heading.Meanwhile,a heading correction algorithm based on environmental information is designed to further correct the heading.To solve the problem of positioning drift error by long-term navigation in positioning and navigation system based inertial sensor.This thesis proposes Bluetooth sensor assisted positioning method based on result fusion.Specifically,for low-cost requirements,a Bluetooth distribution scheme based on road network maps is designed to reduce costs by reducing the requirement of landmark Bluetooth.To solve the problem of positioning drift error caused by navigation and positioning,firstly,a road detection method based on the change of signal strength is designed to determine the road where pedestrians are located.Then the range-based positioning method is used to solve the specific position of pedestrians.Finally,this specific position of pedestrians and the specific positioning results solved by the indoor positioning technology based on the inertial sensor are fused through the Kalman filter.And therefore it improves the positioning accuracy of the positioning system effectively.