Method used for positioning indoor moving targets by improved unscented Kalman filtering

Abstract

The invention discloses a method used for positioning indoor moving targets by improved unscented Kalman filtering. In order to solve problems of traditional unscented Kalman filtering method that positioning precision is low and real-time performance is poor, the traditional unscented Kalman filtering method is improved by an iteration algorithm. Wherein according to the iteration algorithm, a noise covariance matrix is regulated dynamically by using scale factors, i.e. in filtering processes, filtering values of last moments are continuously substituted into a time updating equation; prior estimation of current states is established; estimated values of state variables of current states and estimated values of error covariance are calculated; and observation noise and systematic noise are updated according to the estimated values so as to ensure convergence and stability of the algorithm.

Description

technical field [0001] The invention belongs to the field of real-time tracking of irregular moving targets in an indoor positioning system, and in particular relates to an improved lossless Kalman filter indoor moving target positioning method. Background technique [0002] The traditional filtering technology is proposed based on the filtering problem of dealing with deterministic signals, but in dealing with the state estimation of stochastic systems, useful signals and noise interference are both multi-dimensional non-stationary stochastic processes, in order to solve problems such as real-time filtering and non-stationary signals , Rudolph E.Kalman proposed the Kalman filter theory in 1960. [0003] The original Kalman filter algorithm is only suitable for linear systems with linear observations. Some scholars have proposed an extended Kalman filter algorithm for nonlinear conditions. The algorithm performs Taylor expansion on the system equation or observation equatio...

AI Technical Summary

Problems solved by technology

[0005] When the lossless Kalman filter algorithm is used for indoor moving target positioning, the filter with poor anti-interference ability is easily affected by the drift of system parameters, the change of the performance of the noise estimator, and the jump of the system state. Loss of accuracy and stability, and even cause the filter to diverge; at the same time, since the mathematical model of the dynamic system and the statistical characteristics of the noise are only an approximation to the actual physical model, in practical applications, filtering errors always exist

Images