The Research On Dynamic Object Localization Based On Fusion Of RFID And Laser Information

With the development of Internet of Things technology,the demand for location-based services is growing.As the key technologies of Internet of Things and indoor positioning,RFID(Radio Frequency Identification)and 2D(two-dimensional)laser sensor have the advantages of high positioning accuracy,low energy consumption and low cost.In addition,RFID usually uses the relevant information of the wireless signal to achieve object positioning.However,the wireless signal of RFID has problems such as multipath decline in the indoor environment,which has a great influence on the positioning accuracy.Meanwhile,2D laser sensor require complicated recognition algorithms to locate object,although it can directly obtain the distance and angle information of the object.Therefore,this thesis combines RFID with 2D laser sensor information based on particle filtering and studies the problem of dynamic object localization.The specific research and design work can be divided into the following parts:Firstly,the information detected by RFID is used for object recognition,and combined the information of distance and angle obtained by 2D laser sensor.It can not only locate the position of the dynamic object,but also reduce the time required for RFID signal modeling and simultaneous localization and mapping by 2D laser sensor.At the same time,a platform for collecting RFID and laser information is established according to the requirement of the whole positioning system.Secondly,an approach of velocity matching is proposed to determine that the object detected by the RFID and 2D laser sensor is the same target.The approach uses the RFID phase difference to calculate phase-based velocity of the moving object.Meanwhile,the method estimates distance-based velocity of the moving target in the indoor environment based on the laser point detected by 2D laser sensor.Then,the dynamic object can be located based on the laser information corresponding to the matching results of the phase-based velocity and the distance-based velocity.The experimental results show that the velocity matching method can achieve rough positioning of dynamic object.Then,the traditional particle filter algorithm is improved for the problem of dynamic object positioning accuracy.Meanwhile,the point cloud clustering algorithm is used to process laser information,which improves the accuracy of velocity matching.In addition,the different prediction methods for the particle filtering are selected according to the new velocity matching results.After that,the K best clusters based on the new matching results are used to update the particle's weight in the particle filtering,and achieve the dynamic object localization.The experimental results show that the approach for localization is to improve the positioning efficiency and accuracy by complementing the RFID phase difference with laser clustering using a particle filtering.Finally,since the reading range of the RFID and 2D laser sensor selected by the system is not consistent.Therefore,the acquired data is processed separately according to whether RFID information is available.When the moving object is the range of RFID sensor,combined RFID with laser DBSCAN clustering information is used for positioning continuously.Otherwise,the laser clustering information only is used to achieve continuously positioning of the dynamic object.Combined with the continuous positioning experiment of the dynamic object based on particle filtering fusion RFID and laser information,it can be shown that the approach in this paper greatly improves the positioning accuracy of the system.