A Research Of High-Precision And Low-Cost Indoor Pedestrian Positioning Algorithm Based On Inertial Navigation

Indoor pedestrian positioning technology has been widely used in the fields of commerce,transportation,and security.Compared with other technologies(such as Wi?Fi or UWB),inertial measurement units based inertial navigation system does not require external infrastructure and has lower cost.However,the main problem is that the rapid accumulation of errors will seriously affect the positioning precision.Existing inertial navigation-based indoor positioning technologies have reduced cumulative errors,but still face several problems in positioning precision and cost.Therefore,this thesis proposes a high-precision and low-cost IMU-based indoor pedestrian positioning algorithm.Based on existing algorithms,it takes advantage of pedestrian motion speed and step information to achieve high indoor pedestrian positioning precision with low Cost.The thesis first introduces the basic knowledge of inertial navigation system,including inertial sensor principle and noise analysis,inertial navigation principle and its application,etc.It focuses on how to reduce the impact of inertial sensor error on positioning precision in inertial navigation system which provides a theoretical basis on the proposed high-precision and low-cost IMU-based indoor pedestrian positioning algorithm.Then,the thesis starts with the Kalman filter algorithm commonly used in inertial navigation systems,introduces the development of indoor pedestrian positioning technology,analyzes the advantages and disadvantages of the existing technology,including problems in positioning precision and system cost,and leads to the high-precision and low-cost IMU-based indoor pedestrian positioning algorithm.Next,the thesis expatiates the proposed indoor pedestrian positioning algorithm based on motion speed detection and adaptive parameter update.By detecting the motion speed,it adaptively updates multiple parameters in the Kalman filter algorithm,including the error of the gyroscope and the threshold in the zero velocity update algorithm,and the positioning precision under different motion speed is improved.At the same time,through the artificial intelligence technology based on convolutional neural network,the accuracy of motion speed detection is improved.In the case of using only the accelerometer and the gyroscope,this method achieves high horizontal positioning precision.Afterwards,in view of the vertical positioning precision,the thesis expatiates the proposed zero height update algorithm,up/downstairs height estimation and trajectory tracking algorithm,and improves the vertical positioning precision by updating the height coordinates of pedestrians when zero velocity intervals are detected.At the same time,through the detection of the step number and the prior knowledge of the stair height,the precision of the height estimation and trajectory tracking of up/downstairs is improved.This algorithm is combined with the aforementioned indoor pedestrian positioning algorithm based on motion speed detection and adaptive parameter update to improve the overall positioning precision.The closed error is down to 0.11%,the distance error is down to 0.15%,and the height error is down to 0.32%that is superior to the existing inertial navigation-based positioning algorithms and achieves high-precision and low-cost in indoor pedestrian positioning.Finally,the content of each chapter of the thesis is summarized,and the shortcomings of the proposed algorithm are analyzed,and the future work is prospected.