Intelligent vehicle threat estimation system and method based on variable-structure Bayesian network

Abstract

The invention, which relates to the field of the intelligent vehicle cognitive technology, claims for protection of an intelligent vehicle threat estimation system and method based on a variable-structure Bayesian network thereby evaluating a threat degree of a moving target to a vehicle. The system is composed of a threat modeling module, a data collection module and a threat estimation module. At a threat modeling stage, factors affecting intelligent vehicle threat estimation are determined, wherein the factors include an external environmental factor, a target characteristic factor and a driver factor; a topology of a Bayesian network model is constructed; and then a local condition probability table of the model is determined. During the driving process, the data acquisition module uses sensors to collect real-time data of various influence factors; the threat estimation module reconstructs corresponding variable nodes only for quickly changing factors according to changing rates of all factors to obtain a variable-structure Bayesian network model and then carries out reasoning calculation to obtain a target threat index. Therefore, the performance of intelligent vehicle threatestimation can be improved effectively.

Description

technical field [0001] The invention belongs to the technical field of intelligent vehicles, in particular to the technical field of situation estimation of intelligent vehicles, and in particular relates to an intelligent vehicle threat estimation method based on a variable structure Bayesian network. Background technique [0002] Threat estimation for various targets in the external environment is one of the key technologies for intelligent vehicle environment cognition. In the assisted driving system, threat estimation can effectively identify dangerous targets and remind drivers to avoid collisions. In autonomous driving systems, threat estimation is the basis for safe path planning. [0003] In the existing threat estimation method, the Chinese patent application: A Voice Broadcasting Intelligent Vehicle Path Planning Device and Implementation Method (Application No.: 2014100504696) only considers the distance and angle of the obstacle, and does not consider its motion...

AI Technical Summary

Problems solved by technology

Chinese patent application: Intelligent vehicle path planning method based on threat estimation (application number: CN201610050880.2), constructing a Bayesian network model for vehicle threat estimation according to the speed, distance, environment, weather and other threat factors of the target, and through the inference algorithm The threat index is obtained by reasoning, but the influence of the driver factor on the threat is not considered

Images