Research On High-Precision Indoor Position Technology Under Multi-Source Information Fusion

In recent years,people’s demand for location-based services has been increasing,and unprecedented requirements have been placed on the accuracy of indoor positioning.It is inevitable that a single sensor will have inaccurate positioning during the positioning process,which requires the indoor positioning system to combine the advantages of multiple sensors to maximize the strengths and avoid weaknesses to improve the positioning accuracy.Therefore,this thesis takes the positioning technology of multi-sensor fusion as the research basis and realizes high-precision indoor positioning.The specific work and innovations of this thesis are as follows:1.An ultrasonic positioning system based on the through-beam ranging and reflective cone is proposed,which can realize high-precision real-time positioning of the target.Firstly,the reflection cone structure is designed to expand the angle of sending and receiving of the ultrasonic wave.Then,a temperature compensation strategy based on the DS18B20 temperature sensor is proposed.Finally,an improved weighted multi-dimensional scale positioning algorithm is proposed,which reduces the positioning error within the effective range of ultrasonic waves.After experimental verification,the ultrasonic positioning system proposed in this thesis has high positioning accuracy and good promotion value.2.A CNN-LSTM network-based RCMS ultrasonic indoor positioning algorithm is proposed.Firstly,the BN layer is introduced into CNN to effectively prevent overfitting.Secondly,the CNN and LSTM network are combined to improve the recognition rate of NLOS signals.Finally,the CNN-LSTM network is applied to the ultrasonic Indoor Positioning System.The experimental results show that the positioning accuracy of the ultrasonic positioning system designed in this chapter is about 4 cm,which obviously improves the positioning effect.3.A multi-source fusion indoor positioning algorithm is proposed,which is based on a federated Kalman filter and can achieve higher precision indoor positioning.Firstly,the general structure and algorithm process of the federated Kalman filter are introduced.Secondly,INS is selected as the public reference system,and system state equations and system observation equations are designed for each filtering subsystem.Then,the mode of multi-sensor fusion positioning is designed,and the precondition of mode switching is given.Finally,the function verification is carried out in the experimental environment,and the experimental results show that the algorithm meets the needs of high-precision indoor positioning.