Hybrid enhanced intelligent trajectory prediction method and device based on hybrid grey wolf optimization SVM

Abstract

The invention discloses a hybrid enhanced intelligent trajectory prediction method and device based on a hybrid grey wolf optimization SVM. The method comprises the steps that real-time position information of a running target in front of a coke pusher is acquired through a laser radar; classifying each moving target in the coke oven area based on a human-in-loop hybrid enhanced intelligence concept; embedding the position information and the classification result of each operation target into a support vector machine SVM kernel function with a classification prediction function; the Levy flight, the grey wolf algorithm and the self-adaptive improved differential evolution algorithm are combined; sVM kernel function parameters and penalty factors are optimized, prior data are tested, trained and predicted through optimal parameters, a coke oven area on-site hybrid enhanced intelligent track prediction model is established, track prediction is conducted on the next moving positions andmoving directions of various moving targets, and after the moving trend of the moving targets is analyzed and judged, the coke pusher is controlled to run. Trajectory prediction of the moving target can be quickly performed so that the practicality is high and the prediction result is accurate and efficient.

Description

Technical field [0001] The invention relates to the field of intelligent vehicles, in particular to a hybrid enhanced intelligent trajectory prediction method and device based on a hybrid gray wolf optimized SVM. Background technique [0002] The environment in the coke oven area is harsh, and there are not only pedestrians in the operating area of ​​the coke pushing car, but also various vehicles running in the large lane. The coke pushing car has a large inertia and a relatively long braking distance. If the existing prediction method is used to predict the trajectory of a moving object, there is no guarantee in practice that normal and safe production can be carried out for large inertial motions such as the operation of a track pushing car. In the actual operation process, the operator has a large proportion of the control factors of the pushing car. The usual path tracking control method only focuses on the trajectory control of the pushing car itself, and ignores the pushin...

AI Technical Summary

Problems solved by technology

[0003] The embodiment of the present application provides a hybrid enhanced intelligent trajectory prediction method based on the hybrid gray wolf optimization SVM, which solves the problem that when the existing prediction method predicts the trajectory of a moving target, it cannot be guaranteed that the large inertial motion of the orbital coke pusher can be carried out. For the problem of normal safety production, it is possible to accurately locate the real-time position of the moving target in front of the coke pusher, and accurately identify and classify it, embed it into the SVM kernel function, and bring it in with the optimal value of its parameters and penalty coefficient. Quickly predict the trajectory of moving targets, with high practicability and accurate and efficient prediction results

Images