Automatic driving tracking method based on interactive multiple models

Abstract

The invention relates to an interactive multi-model-based automatic driving tracking method, which comprises the following steps of: 1) establishing a reference coordinate system, and establishing an interactive multi-model consisting of three motion models in the reference coordinate system; 2) determining a measurement equation based on the reference coordinate system; 3) taking the three motion models as a state equation of target motion to be combined with a measurement equation, and filtering to obtain a motion state predicted value of each motion model and a corresponding error covariance matrix; and 4) calculating a likelihood function of each motion model, dynamically adjusting the angular velocity of the constant-speed turning model and the probability of each motion model, and then performing motion state estimation and error covariance matrix fusion. According to the method, the motion model is reasonably selected on the basis of the interactive multi-model, so that the computing power demand is reduced; by adaptively adjusting the angular velocity of the constant-speed turning model, the problem that the angular velocity cannot be accurately estimated is solved, the error during target tracking is reduced, and the calculation result is more accurate.

Description

Technical field [0001] The present invention is a technology field for automobile autonomous driving, and it involves a method of autonomous driving tracking based on interactive multi -model. Background technique [0002] With the rapid development of automobiles, more and more cars are equipped with autonomous driving functions. Motor target tracking ahead is an essential ability to achieve autonomous driving. Tracking the motor target ahead mainly uses the motion of the target to calculate the target. In the past, most automakers used a uniform acceleration model or a uniform turn model as a motion model to calculate the target motion state. However, because the motion mode of the vehicle includes uniform straight, uniformly accelerated straight, uniform turn, and variable acceleration turns, the vehicle's motion mode is diverse and uncertain. Each motion mode corresponds to a sports model. It is difficult to use a single sports model. Accurately calculate the real movement st...

AI Technical Summary

Problems solved by technology

[0004] In view of the above-mentioned deficiencies in the prior art, the technical problem to be solved by the present invention is to provide an interactive multi-model based automatic driving tracking method, which solves the problems of complicated calculation and poor accuracy in the current automatic driving tracking method, and achieves improved accuracy. and real-time effects

Images