Pedestrion Dead Reckoning Technology Research Based On Portable Electronic Devices

With the rapid development of smart mobile devices technology, it makes a big difference to people’s daily life; as a part of it, smart phone and tablet PC become indispensable parts of daily life. In order to have novel applications and visual effects, MEMS sensors, such as MEMS accelerometer, compass and gyroscope, are widely used in smart phone and tablet PC.This project relies on UESTC-NOKIA international cooperative project. The main goal is to explore a novel pedestrian navigation system based on MEMS inertial sensors and suitable for consumer electronics equipment.MEMS inertial sensors used in consumer electronics equipment have some disadvantages, such as low accuracy, larger random error et al. in order to control the cost. This article detailedly analysis LIS344 ALH acceleration’s and Ex3500A4962 A gyroscope’s error, then built error model, use six-position method to calibrate sensors errors. Low-pass filter and extended Kalman filter are designed for random errors of sensors, which effectively reduce the random error and improve the accuracy of entire system.The MEMS IMU is attached to the pedestrian’s waist. Accelerometer sensor measures the pedestrian’s acceleration signal. Features extracted from pedestrian kinetic model are used to recognize the step event, divided the step signal into two phases: the stance phase and the swing phase, estimate step frequency, step counting, stride length, attitude angles. MEMS gyroscope measures the angular rates of pedestrian, the signal are used to calculate the orientation matrixes and attitude angles of pedestrian directly. The attitude calculation algorithm basing on data fusion technology is used to reduce the impact of the attitude angles calculation accuracy. It improves the calculation accuracy and makes the system more suitable for low accuracy inertial sensors. The trajectory of pedestrian is reconstruction by DR theory.Finally, this paper completed the entire system algorithm programming in Linux system. The system tests were performed by experimenters in indoor environment and in a normal walking speed. This attitude projection algorithm accuracy was comparative analyzed and assessed. Compared experimental results with the actual trajectory, analyzed the accuracy and feasibility of pedestrian navigation algorithm. The results show the system can meet the real-time demand of consumer electronics field. In the end, this paper gives the suggestions for improvement, according to the experimental results and signal analysis.