Measurement method for evaluating performance of continuous time track tracking algorithm

Abstract

The invention discloses a measurement method for evaluating the performance of a continuous time track tracking algorithm, which is used for measuring the difference between a real track and a measured track based on a time track function (T-FOT) model through a distance. The method aims to reflect the difference between advantages and disadvantages of measured tracks generated by different tracking algorithms in real time, quickly and reasonably, so that the optimal tracking algorithm is selected to realize real-time tracking of a target. The method comprises the following steps: dividing a measured track into a consistent maintenance track and an inconsistent deviation track according to the difference of track maintenance time, and on the premise of conforming to mathematical and kinematic laws, superposing a penalty term of an inconsistent deviation track part through integral operation on the consistent maintenance track part, thereby obtaining a consistent maintenance track and an inconsistent deviation track; and obtaining a difference value between each measured track and the real track. By means of the mode, real-time performance evaluation of the continuous time track tracking algorithm can be achieved, the design is simple and convenient, the calculation complexity is small, and the calculation efficiency is high.

Description

technical field [0001] The invention belongs to the technical field of target tracking, and in particular relates to a method for measuring the performance of a track tracking algorithm. Background technique [0002] Track estimation is the core issue of target tracking research. At present, the mainstream algorithm at home and abroad implements discrete-time track estimation. Because it is easily affected by factors such as maneuvering targets and unknown inputs, an alternative—continuous-time track estimation has become a Research hotspots of scholars in related fields. The key to evaluating the performance of the target tracking algorithm is to measure the "distance" between the estimated track and the real track. The widely recognized comprehensive measurement methods include optimal sub-pattern assignment (OSPA), OSPA2, generalized Optimal subpattern assignment (generalized optimal subpattern assignment, GOSPA), etc., but the above methods are only applicable to the me...

AI Technical Summary

Problems solved by technology

However, the method described in the literature is a measurement method based on two finite point sets, which is only useful in the field of evaluating discrete-time tracks, and is not suitable for evaluating the performance of continuous-time track tracking algorithms.

If the measurement track is a smooth piecewise polynomial fitting function, this method is useless, so the relevant evaluation algorithms in this field need to be further expanded

Images