Design And Implementation Of Multi-floor Positioning System Based On LoRa

With the increasing demand for indoor positioning in people's daily life,indoor positioning technology has attracted increasing attention from domestic and foreign scholars.Most of the existing indoor positioning platforms focus on positioning on a two-dimensional plane.At present,the bottleneck in the development of indoor 3D positioning technology is mainly reflected in vertical positioning.Therefore,on the basis of mature 2D positioning,improving the shortcomings of floor positioning technology and improving the service quality of indoor positioning have become the top priority for breakthroughs in indoor 3D positioning technology.The current existing floor positioning methods are affected by factors such as indoor environment and technical means,and have many problems such as low floor determination accuracy,high algorithm complexity,high node density,and inability to identify trajectories.In response to the above problems,this thesis proposes a LoRa-based multi-floor positioning method,and designs the floor positioning system from the aspects of system design,data processing and networking,to achieve high accuracy of judgment,low node density,trajectory recognition and press Multi-floor positioning to be deployed..The main research work of this thesis includes:(1)In view of the low accuracy of floor determination and high algorithm complexity in existing floor positioning methods,this paper proposes a LoRa-based multi-floor positioning method.This method deploys LoRa longitudinal positioning equipment on demand,establishes an offline fingerprint database and a threshold fingerprint database,combines the RSSI floor determination algorithm with a ranging-based signal source autonomous switching mechanism to achieve multi-floor determination,and uses the fingerprint database positioning method to achieve trajectory recognition.Through the combination of floor determination and trajectory recognition technical means,more comprehensive vertical direction information can be obtained.(2)In view of the jitter problem of LoRa signal propagation under dynamic activities,this paper uses Kalman filter algorithm to preprocess the LoRa RSSI value.The results show that the Kalman filter algorithm combines the measured value with the predicted value through the Kalman equation to optimize Optimize the estimation,obtain a smooth curve that tends to the true value after correcting the data,reduce the influence of LoRa signal jitter on floor positioning,and obtain a more accurate RSSI value.(3)It is difficult to coordinate the indoor multi-node communication.In order to solve the problem,this paper proposes a Self-Organizing Networks solution based on time division multiplexing,and designs the network node and coordinator.The test results show that the communication distance of the node can reach about 200 m under non-line-of-sight,and the packet loss rate within 150 m is 0.In this networking mode,the indoor LoRa multi-node communication has high stability and meets the needs of multi-floor positioning in the experimental environment.(4)This paper is based on LoRa's multi-floor positioning method,combined with data preprocessing and LoRa self-organizing network solution,designed and implemented a multi-floor positioning system with Arduino as the client and Raspberry Pi as the server.The development model combined with Arduino and Raspberry Pi can reduce the complexity of setting up a development environment,improve the system's diversified task management and coordination capabilities,and provide a reference for the development of related systems.After experimental verification,the system has a floor accuracy rate of 97.9%,and can identify trajectories,achieving a high accuracy rate and more comprehensive floor positioning.