Scene complexity evaluation method based on gradient boosting decision tree model

Abstract

The invention provides a scene complexity evaluation method based on a gradient boosting decision tree model, and the method comprises the following steps: S1, collecting parameters, and generating asimulation driving scene; S2, performing complexity scoring on the simulated driving scene sample; S3, inputting the summarized complexity characteristic elements into a decision tree model for calculation; S4, upgrading the decision tree; S5, obtaining a characteristic parameter data set influencing the model, taking 80% of the data set as a training set and 20% of the data set as a test set, andadopting 5-fold cross validation debugging to obtain a complexity evaluation model; S6, calculating scene complexity of the to-be-evaluated data; and S7, splitting the input driving scene data into dynamic features and static features, and performing comprehensive scoring according to each influence feature to obtain scene complexity. According to the method, a clear and simple complexity estimation value of the automatic driving test scene can be given, and the requirement that a tester can select the driving scene according to the complexity of the scene is met.

Description

technical field [0001] The invention belongs to the technical field, and in particular relates to a scene complexity evaluation method based on a gradient lifting decision tree model. Background technique [0002] Most of the existing simulation test driving scene classifications are based on the functions to be tested. The driving scene categories obtained by using this type of classification method are relatively general, and testers will not be able to know the relevant test scenes concisely and clearly. For the test difficulty that can be generated by the test of automatic driving control algorithm, it is impossible to do progressive test. In the actual test process, in addition to the different types of scenarios, testers need to consider the testing of scenarios with different difficulties. Therefore, this paper designs a scene complexity evaluation model based on the gradient boosting decision tree model for the calculation of scene complexity of intelligent network ...

AI Technical Summary

Problems solved by technology

[0003] In view of this, the present invention aims to propose a scene complexity evaluation method based on a gradient-lifting decision tree model to solve the test difficulty that testers will not be able to succinctly and clearly know the relevant test scenes for the test of the automatic driving control algorithm. It is also impossible to do progressive testing

Images