ICIEOM2020_FULL_0004_37436
TRACKING MANEUVER TARGET USING 3D POSITION FORECAST MODEL WITH ESTIMATED SPEED BY LEAST SQUARES METHOD
ÁREA: 04. Data Science and Stochastic Optimization
RESUMO:
Tracking targets is a complex process, which requires association algorithms capable of handling the use of linear and nonlinear filtering techniques. It is known that the two-dimensional models of air traffic control systems are insufficient for the treatment of three-dimensional maneuvers of military targets, due to considerable variations in altitude. The work was based on flight dynamics models, which describe the evolution of the state of a target, treated as a punctual object in three-dimensional trajectories, addressing the problem of its movement uncertainty. The adopted model has application in civil and military navigation and surveillance systems, allowing the tracking of targets in real time. The Kalman Filter (KF) and the Extended Kalman Filter (EKF) were adopted as state estimators with integration through the filter of hybrid systems Interacting Multiple Models (IMM). The innovation of the work is in obtaining the scalar velocities of each Cartesian axis to be part of the vector of z_k, observations, through the Method of Least Squares, resulting in greater precision than in previous works. Numerical examples illustrate the applicability and performance of the proposed method.
PALAVRAS-CHAVE:
tracking targets, mathematical models, kalman filter
DOI: 10.14488/ijcieom2020_full_0004_37436
RESUMO:
Tracking targets is a complex process, which requires association algorithms capable of handling the use of linear and nonlinear filtering techniques. It is known that the two-dimensional models of air traffic control systems are insufficient for the treatment of three-dimensional maneuvers of military targets, due to considerable variations in altitude. The work was based on flight dynamics models, which describe the evolution of the state of a target, treated as a punctual object in three-dimensional trajectories, addressing the problem of its movement uncertainty. The adopted model has application in civil and military navigation and surveillance systems, allowing the tracking of targets in real time. The Kalman Filter (KF) and the Extended Kalman Filter (EKF) were adopted as state estimators with integration through the filter of hybrid systems Interacting Multiple Models (IMM). The innovation of the work is in obtaining the scalar velocities of each Cartesian axis to be part of the vector of z_k, observations, through the Method of Least Squares, resulting in greater precision than in previous works. Numerical examples illustrate the applicability and performance of the proposed method.
PALAVRAS-CHAVE:
tracking targets, mathematical models, kalman filter
DOI: 10.14488/ijcieom2020_full_0004_37436
AUTORES:
JAIR AGNER JUNIOR
CENTRO DE ANÁLISES DE SISTEMAS NAVAIS
BRUNO SCHULZE
LABORATÓRIO NACIONAL DE COMPUTAÇÃO CIENTÍFICA - LNCC
ANTÔNIO ROBERTO MURY
LABORATÓRIO NACIONAL DE COMPUTAÇÃO CIENTÍFICA - LNCC
MARIZA FERRO
LABORATÓRIO NACIONAL DE COMPUTAÇÃO CIENTÍFICA - LNCC
MARCOS DOS SANTOS
UNIVERSIDADE FEDERAL FLUMINENSE
JAIR AGNER JUNIOR
CENTRO DE ANÁLISES DE SISTEMAS NAVAIS
BRUNO SCHULZE
LABORATÓRIO NACIONAL DE COMPUTAÇÃO CIENTÍFICA - LNCC
ANTÔNIO ROBERTO MURY
LABORATÓRIO NACIONAL DE COMPUTAÇÃO CIENTÍFICA - LNCC
MARIZA FERRO
LABORATÓRIO NACIONAL DE COMPUTAÇÃO CIENTÍFICA - LNCC
MARCOS DOS SANTOS
UNIVERSIDADE FEDERAL FLUMINENSE
|