A Single-reader Indoor Localization Method Based on Multi-objective Evolutionary Algorithm Optimization of Deep Model

Nowadays, RFID (radio frequency identification) -based indoor positioning technology is usually combined with deep learning in artificial intelligence for positioning and identification. Since there are usually several hyperparameters in deep models, finding appropriate hyperparameters needs the support of optimization algorithms. the present optimization technology consumes plenty of resources, takes an extended time to optimize, and has room for improvement. We provides a multi-objective evolutionary algorithm (MOEA) to optimize the indoor single reader positioning model, which may not solely greatly save the resources within the optimization process, shorten the optimization time, however conjointly improve the accuracy of the time sequence model in indoor positioning. The experimental results show that the model optimized based on the multi-objective evolutionary algorithm performs well within the regression analysis index of the positioning problem, that effectively improves the accuracy of the model, and may be combined with numerous time sequence model, that is wide utilized in indoor cheap and high-precision position.