Research On Indoor Positioning Method Of Smartphone Based On Acoustic Signal And PDR

With the development of Internet technology,location-based services provide many convenience for people's daily life.Under the outdoor environment,the GPS has a high positioning accuracy,however it is obstructed by the shelter in the indoor environment.People urgently need an indoor positioning technology,which has the characteristics of high precision,low cost and good robustness.However,a single indoor localization technology has various defects and it is difficult to meet the needs of users.In recent years,smartphones are becoming more and more popular.The use of smart phones built-in sensors,a variety of positioning technology integration,and then develop a high-precision,low-cost,robust indoor positioning system,has become a new research hot spot.Through reading a large number of domestic and foreign literatures and doing related preliminary experiments,we found that PDR(Pedestrian Dead Reckoning)is self-contained and not susceptible to environmental interference,but it accumulates errors.The acoustic signal localization has high positioning accuracy,but is susceptible to environmental noise.Based on the above analysis,this thesis presents an indoor positioning method of smartphone based on acoustic signal and PDR.The main work and contributions of the thesis include the following aspects(1)Robust estimation and algorithm implementation of PDR related parameters based on smart phone.This thesis presents a dynamic threshold-based algorithm for the step detection.For step length estimation,a personalized step length estimation based on acoustic signal localization and a dynamic step length estimation based on particle filter are proposed,which can effectively reduce the positioning error caused by fixed step length.(2)Particle filter fusion localization solution based on acoustic signal localization and PDR for indoor open scene.A time registration method based on the step length model is proposed to solve the influence of the asynchronous problem on the fusion location.In this thesis,we propose a non-line-of-sight identification and particle weight updating method based on PDR for acoustic signal localization.This method effectively eliminates the bad influence of the outliers of acoustic signal on fusion localization and improves the robustness of the system.(3)A low-cost particle filter fusion localization solution based on the TDOA(Time Difference of Arrival),PDR and map information for indoor corridor scene.We use the TDOA invariance in a specific geometric location to obtain the initial position of the user,so as to realize the particle initialization and reduce the dependence on the initial position.In this thesis,TDOA non-line-of-sight identification and particle weight updating method based on PDR are proposed to eliminate the negative effect of TDOA anomaly.Furthermore,an improved particle filter resampling method based on user's walking characteristics in corridor is proposed,which improves the positioning accuracy and system robustness effectively.