Research On RSSI Indoor Localization Algorithm Based On Set-membership Filtering

Indoor positioning technology is one of the key technologies in service robot,indoor navigation,shopping guide in shopping mall,mine safety and other related fields.Due to the building block,multi-path effect and other factors,positioning accuracy of GPS and other global satellite positioning technology can't meet the performance requirements of the positioning in the indoor environment,so a variety of indoor positioning technology emerged.Among them,the indoor positioning technology based on RSSI is favored by researchers because of its low cost and easy deployment.However,because radio signal affected by the multi-path effect,non-line-of-sight and other environmental factors in the propagation process,its accuracy cannot meet the needs of industrialization process.In order to improve the positioning accuracy,set-membership filtering algorithm is adopted in this paper to study the indoor positioning problem based on RSSI.Compared with the traditional Kalman filter and other algorithms,set-membership filtering algorithm does not need to obtain the statistical characteristics of noise,so it is of more practical significance.In this paper,the set-membership filtering algorithm is used to study indoor positioning from the following two aspects:(1)Aiming at the problem that the statistical characteristics of indoor location noise are not easy to obtain and the communication bandwidth is limited,this paper proposes a set-membership filtering algorithm based on event-trigger mechanism.Firstly,based on the algorithm of finding the intersection and union of ellipsoids,an elliptical trilateral localization algorithm is proposed to solve the problem that the traditional trilateral localization algorithm can not obtain a feasible solution in the case that the distance measurement error is large.Secondly,an anchor node switching algorithm is proposed to solve the problem that error increases caused by moving target.Finally,the set-membership filtering algorithm based on event-trigger is used to reduce the influence by the noise of indoor positioning in the condition of limited bandwidth.The experimental results show that the set-membership filtering algorithm based on event-trigger has higher positioning accuracy when the statistical characteristics of noise are unknown,and it can effectively save communication resources.(2)The traditional set-membership filtering algorithm need a large amount of computation and cannot be used in situations where real-time requirements are high.In view of this problem,the sequential fusion algorithm is used in set-membership filtering in this paper,which can effectively improve the computational efficiency.The whole filtering process is divided into prediction step and measurement update step.In the prediction step,a prediction ellipsoid is obtained on the cloud filter according to the system model.In the measurement update step,the measured values are gradually fused in a sequential way,and finally an estimation ellipsoid is obtained.Simulation results show that the proposed algorithm can maintain good positioning performance even when the initial estimation error is large.