Location:Home > Engineering science > Information and Communication Engineering > Human Walking Analysis and Displacement Calculation Based on Inertial Sensing System

Human Walking Analysis and Displacement Calculation Based on Inertial Sensing System

Downloads: []
Tutor: LiuZuo
School: Tianjin University
Course: Information and Communication Engineering
Keywords: Indoor localization,MEMS,Gesture Recognition,Single-Steplength,SVM
CLC: TP212.9
Type: Master's thesis
Year:  2012
Facebook Google+ Email Gmail Evernote LinkedIn Twitter Addthis

not access Image Error Other errors

Indoor localization has become one of the focuses in the current informationsciences research. Lots of methods for solving the problem of indoor localization havebeen presented recently, one of which is important by using MEMS (Micro-electroMechanical Systems) to implement the positioning function. In this thesis, detailedresearch about the indoor localization based on MEMS is conducted.In order to divide the walking action, which is one of the most common dailyactions of human life, into several parts, firstly MEMS is placed on the leg and waistrespectively to collect data in the thesis. And the collected parameters are analyzed.Using the values of acceleration and angular velocity, the entire process of walkingmotion is identified. By comparing the values of acceleration and angular velocitywhich acquired from waist with the values from leg, the dividing of walking actionand parameter distribution are derived separately and crosschecked together. Bycomputing the magnetic data distribution of stationary and walking state, thedistribution can be analyzed. And the decision algorithm of the moving direction isproposed.The inertial parameters can be collected by placing MEMS on one of the shanks.Through these parameters, the continuous walking process can be divided into severalsingle-steps and the total displacement can be cumulated by these single-stepdisplacement. Firstly, the detailed calculation principle of this algorithm is given outin the thesis. A large number of experiments to validate the algorithm are conducted.Secondly, the selection of parameters in the algorithm is verified. And the experimentresults are analyzed. Finally, the accuracy of this algorithm on the standard conditionsis tested. Based on the moving displacement in experiments, the conclusion thatcomputing the vertical displacement of walking can verify the moving displacementresult is confirmed. Lots of experiments aiming at the problem about the influences oflimitation of Inverted are made. And the experiments results are analyzed. Theadvantages, disadvantages and limitations of the algorithm are given at the last.Aiming at the inconvenient of carrying MEMS on shank, a new algorithm tocalculate the walking displacement by placing MEMS on waist is proposed in the thesis. By analyzing the characteristics of waist trajectory during walking, dividingthe continuous walking process into single steps accurately, the single-stepdisplacement using the algorithm. A lot of experiments to verify the correctness andaccuracy of the algorithm are made. The results show that the algorithm can estimatethe single-step displacement and total displacement accurately. And the method canbe applied in the fields of indoors localization.To address the problem of distinguishing going upstairs, downstairs and standardwalking, Support Vector Machine (SVM) is utilized in the thesis. Data collection,motion analysis and feature extraction for the three actions are carried out. In additionto the specific characteristics of the parameters, the three actions are distinguished.The result shows that using SVM can effectively distinguish the motions of goingupstairs, downstairs and standard walking.
Related Dissertations
Last updated
Sponsored Links
Home |About Us| Contact Us| Feedback| Privacy | copyright | Back to top