| In recent years,with the rapid development of economic society,people's demand for location information is growing and expanding from the outside to the indoors.The application of global navigation and positioning technology in outdoor navigation and positioning has been quite mature after decades of development and successfully applied in the location service software such as Baidu Map and Amap.However,satellite signals cannot be received indoors due to the occlusion of buildings,and mature outdoor positioning technology cannot be applied indoors.Fortunately,various indoor positioning technologies are constantly developed.Pedestrian Dead Reckoning(PDR)technology is one of them.It uses the sensor data such as acceleration and gyroscope collected during pedestrian movement to judge the current state of pedestrians.The current position coordinates of the pedestrian is determined by calculating the step size and heading of each step of the pedestrian movementSensor data acquisition software based on Android platform for experimental needs was developed in this paper.At the same time,the program developed by JAVA is used to combine the accelerometer and barometer data in the mobile phone to realize the step estimation and combine the gravimeter and the magnetometer and the gyroscope data to realize the attitude calculation of the mobile phone,the barometer data is used to realize the height measurement.Finally,the collected multi-sensor data is processed and analyzed by the indoor three-dimensional positioning program.The research content mainly includes following parts:1?Android software was developed to collect multi-sensor data,and the low-pass filtering was used to preprocess the data.The development status and theoretical research basis of indoor positioning are introduced in detail.The existing indoor positioning algorithm is studied theoretically and the formula derivation,program design and result analysis are completed.2?On the basis of learning the existing step frequency detection method,the barometer data is introduced to identify the pedestrians going up and down the building and riding the elevator state,and the gait detection method of "Stepping Monitoring + Threshold Judgment + State Recognition" was proposed.The relationship between the characteristic quantities of the acceleration data and the step size was analyzed,the bivariate step size estimation model was calculated by experiments.3?Aiming at the advantages and disadvantages of various heading estimation algorithms such as gyroscope integral and complementary filtering,the heading angle was solved by fusing multi-sensor data with Kalman filter,and other algorithms were compared through experiments.4?The existing indoor air pressure altimetry model was studied.According to the principle of relative air pressure altimetry,the indoor high algae experiment was carried out by using the improved international standard atmospheric pressure high formula.5?The pedestrian dead reckoning principle was taken as the experimental basis,the school office building was taken as the experimental site and a variety of experimental walking routes were designed,the overall test of the indoor positioning algorithm was conducted by using the step size estimation,heading estimation and indoor altimetry method described in this paper.The results show that the error of 44 m short-distance straight-line walking is within(0.12 m,0.52m),and the average accuracy of 147 m long-distance rectangular walking is 2.5m,and the height accuracy is 0.6m. |
PDF Full Text Request