RFID Indoor Positioning Research And Application Based On Grasshopper Optimization Algorithm And Extreme Learning Machine

Radio Frequency Identification(RFID)has the advantages of non-contact,long identification distance,high anti-interference capability,strong environmental adaptability and high confidentiality.RFID indoor positioning technology continues to develop not only has a broad market prospects and has far-reaching practical significance.The existing RFID indoor positioning algorithm,when dealing with complex indoor environment,still has some urgent problems to be solved on such as positioning accuracy,positioning time and adaptability.With the rise of machine learning,the localization fingerprint model built by artificial neural network adaptive noise data training and its strong nonlinear mapping ability and good data fitting ability can reduce the update cost and improve the ability of adaptive environment change.This paper further proposes an indoor positioning Algorithm based on RFID and Extreme Learning Machine(ELM).The location-based model was optimized with Grasshopper Optimization Algorithm(GOA),and thecorresponding campus safety supervision platform system was developed.The main research contents and innovations of this paper are as follows:1.Optimization of the Extreme Learning Machine by Grasshopper Optimization Algorithm: For the traditional Extreme Learning Machine algorithm,the input layer weight and hidden layer threshold are randomly selected during the model construction,so the former ELM algorithm will not be corrected in the training process.For Grasshopper Optimization Algorithm,all targets will participate in the position update of each target through continuous update and iteration of position vector.In this way,not only can we avoid falling into the local optimal solution and fast convergence to the greatest extent,but also we can get better weights and thresholds for ELM to find the global optimal solution and improve the performance of the algorithm.2.Apply the optimized ELM algorithm of Grasshopper Optimization Algorithm(GOA)to RFID indoor positioning system: The optimized positioning system can obtain extremely fast learning speed by virtue of its random feature mapping and tight network structure,so as to reduce the offline learning time and effectively overcome the impact of environmental change and RSSI time-varying on the positioning accuracy.Compared with LANDMARC algorithm and ELM algorithm,the average positioning error is reduced by 21.67% and 11.72% respectively.Simulation and experimental results show that the proposed algorithm can obtain more accurate positioningresults,reduce the time cost,and has better adaptability to environmental changes.3.Based on the above algorithm research,this paper takes R200 reader as the hardware platform,with Spring Boot and My Batis as the project architecture and My Sql,Redis and Elastic Search as the database.The platform of campus security supervision is designed and implemented with Java and Python language.The main functions include student information management,student indoor location,and historical track management,etc.Through real environment experiments,the effectiveness of student location function and other functions in specific complex indoor scenes was tested.