Research On Positioning And Tracking Algorithms Based On Time Difference Information

In recent years,target location and tracking technology has been widely used in the military and civilian fields,the main functions include location navigation,field search and rescue,location acquisition,etc.However,the rapid development of technology has led to the demand for a high standard of location tracking technology.Target location tracking in practical applications is vulnerable to the limitations of the observation environment and equipment.The highly nonlinear state of the target state model makes it more difficult to solve,which becomes a major challenge in positioning tracking technology.The time difference information positioning tracking has the characteristics of high fault tolerance,high accuracy and small measurement environment constraints.Therefore,this paper uses the Time Difference Of Arrival(TDOA)information to make improvements for fixing target positioning and mobile target tracking,respectively,with the following main improvements:Firstly,in the process of locating a fixed target with no trajectory of motion,the parsing and iterative algorithms are prone to algorithm dispersion and ambiguity in the case of poor initial values.An improved Adaptive Crow Search Algorithm(ACSA)is proposed,which integrates the overall change of the population with increasing number of evolutionary generations,balances the global search ability and the local optimization ability in the iterative process of the algorithm,and designs an adaptive perceptual probability model,so that the algorithm retains more excellent individuals at the initial stage,ensures population diversity,avoids local optimization,and converges quickly in the later stage.Theoretical and simulation results indicate that the accuracy of the improved algorithm is better than the existing classical TDOA positioning algorithm and the original algorithm,and the convergence speed is significantly improved.Secondly,to address the problem that a single motion model cannot fully describe the actual state tracking effect of the target in the tracking of motorized targets,we propose to use the Interacting Multiple Model(IMM)for motorized target tracking,and comparison of filtering performance of Extended Kalman Filtering(EKF),Unscented Kalman Filter(UKF)and Particle Filter(PF)algorithms in nonlinear systems by simulation,and introduce UKF into IMM to solve the problem of large EKF error and poor filtering effect in models with high nonlinearity.