A peripheral vehicle behavior identification method based on an HMM-SVM double-layer improved model under a complex road condition

Abstract

The invention discloses a peripheral vehicle behavior identification method based on an HMM-SVM double-layer improved model under complex road conditions. The method comprises the following steps of 1) offline training, dividing typical surrounding vehicle behaviors, extracting state feature information of surrounding vehicles, transmitting the state feature information to a vehicle through the Internet of Vehicles, converting the vehicle information into state feature information under a road joint coordinate system, generating a feature vector X, and respectively inputting the X into an HMMand SVM parameter learning; 2) carrying out the model improvement which is characterized in that a threshold processor is arranged between the HMM and the SVM, an NSGA-II algorithm is used for optimization to obtain an optimal difference factor, and an HMM-SVM double-layer improved model is obtained; and 3) online testing characterized in that the vehicle utilizes an HMM-SVM double-layer improvedmodel to identify the behavior mode of the tracked vehicle. According to the invention, the application occasion of a surrounding vehicle behavior system is expanded by using a road joint coordinate system constructed by a high-precision map; the excellent time sequence modeling capability of the HMM and the extremely high binary classification capability of the SVM are organically combined, and the double-layer model is improved, so that the accuracy rate and the recognition speed of vehicle behavior recognition are improved.

Description

technical field [0001] The invention belongs to the technical field of vehicle intelligent driving, and in particular relates to a behavior recognition method for surrounding vehicles under complex road conditions based on an improved HMM-SVM double-layer model. Background technique [0002] Road scene understanding is one of the important components of vehicle driver assistance systems, and it is also the basic requirement for future vehicle autonomous driving. The real traffic environment is a complex system in which multiple traffic participants interact with each other and change dynamically. In this complex system, smart cars must not only have visual perception, detection and tracking capabilities, but also need to be able to identify and even Predict the behavior of surrounding vehicles, thereby increasing the depth of perception. [0003] When performing behavior recognition on surrounding vehicles, it is essential to obtain accurate and practical state characterist...

AI Technical Summary

Problems solved by technology

Among them, the Hidden Markov (HMM) model is the most widely used because of its powerful time series modeling ability, but it ignores the influence of negative samples, and only uses the maximum likelihood value for classification, especially when the output multidimensional probability When the difference between the two is small, it is easy to cause a high misrecognition rate

Images