Research On Manufacturing Workshop-oriented WiFi And Image Fingerprint Fusion Positioning Method

With the sustained development of mobile communication technology and the increasing demand for new services,location-based service has attracted a lot of attention.At present,due to lots of equipment and enormous space in manufacturing workshop,how to achieve the real-time positioning of workshop personnel or articles is a difficult problem.The fingerprint-based positioning methods based on wireless signal and image are widely used in indoor space because of its simple principle and low cost.In this paper,AP deployment strategy and positioning algorithm of fusing WiFi signal and image space are proposed in three-dimensional.They can solve the positioning accuracy and real-time problems in indoor positioning.(1)The fingerprint-based positioning methods based on wireless signal and image are in-depth analysis,as well as fusion positioning method.The principle of positioning based on WiFi and image is introduced.This paper focuses on the typical location algorithms:deterministic algorithm and probability algorithm,as well as three methods of image feature information extraction:SIFT algorithm,SURF algorithm and LBP algorithm.(2)A kind of six-sided cylinder grid division method for RPs in positioning space of manufacturing workshop is proposed for complexity of positioning space and the interference of wireless signals.It is verified that six-sided cylinder grid division method has lower positioning error than cube grid comparing with the distance distribution function between mobile target and RPs.A multi-objective function of three-dimensional AP deployment considering signal coverage and signal space is established and solved by artificial immune algorithm.The simulation and experimental results show that the proposed AP deployment method can achieve higher signal coverage and positioning accuracy than the center deployment method and the fingerprint diversity deployment method in four real scenarios.(3)The traditional fingerprint-based positioning technology based on WiFi determines the location of mobile target by received RSS.It has high real-time positioning but low positioning accuracy.While fingerprint-based positioning technology based on image determines the location information of mobile target by matching the features of the foreground image.It has high positioning accuracy but low real-time positioning.In this paper,a positioniong strategy of fusing WiFi signal and image space is proposed.The WiFi signal is transformed into W-image according to indoor space and correction radius.Then the W-image and RGB image are fused with LBP feature by uniform LBP algorithm.A fusion location model based on compressed sensing is established and Lasso+BPDN algorithm is used to solve the model.The experimental results show that the proposed method reduces the complexity of positioning algorithm,improves the work efficiency,achieves higher positioning accuracy under the same conditions,and solves the problem of the balance between positioning accuracy and real-time positioning in the positioning process.It has a good application prospect.