Expected function safety analysis method for misoperation of automatic driving vehicle

Abstract

The invention discloses an expected function safety analysis method for misoperation of an automatic driving vehicle. The expected function safety analysis method comprises the following steps: S1, constructing a virtual dangerous scene terrain for misoperation of the automatic driving vehicle in simulation test software; S2, creating an automatic driving vehicle misoperation test scene; S3, performing a misoperation scene simulation test experiment on the simulated automatic driving vehicle; S4, exporting simulation test data; S5, performing comparative analysis on the simulation test data and the real vehicle safety test data; S6, judging whether the simulation test data is abnormal or not: if the simulation test data is abnormal, analyzing reasons causing misoperation, debugging and correcting parameters of a corresponding vehicle decision algorithm, and repeating the step S3S6; if the simulation test data is not abnormal, ending the virtual simulation test. By means of the expected function safety analysis method for misoperation of the automatic driving vehicle, safety problems caused by intelligent driving can be reduced, and the safety of the automatic driving vehicle is improved.

Description

technical field [0001] The invention relates to the technical field of unmanned driving testing, in particular to a method for analyzing the expected functional safety of an automatic driving vehicle misoperation. Background technique [0002] According to research data, nearly 94% of fatal car accidents are directly related to drivers, such as fatigue, speeding, drunk driving or other illegal behaviors, and intelligent driving is considered to significantly reduce the accident rate. With the increase in the complexity of vehicle systems, self-driving vehicles mainly rely on sensors to perceive the external environment, intelligent algorithms for control decisions, etc., and a large number of new technologies are used in vehicles, thus introducing new safety risks. [0003] In March 2018, an Uber self-driving car accidentally hit and killed a pedestrian in the United States. In the four years from 2016 to 2020, Tesla crashed into a white truck three times due to camera recog...

AI Technical Summary

Problems solved by technology

With the increase in the complexity of vehicle systems, self-driving vehicles mainly rely on sensors to perceive the external environment and intelligent algorithms to make control decisions.

[0003] In March 2018, an Uber self-driving car accidentally hit and killed a pedestrian in the United States. In the four years from 2016 to 2020, Tesla crashed into a white truck three times due to camera recognition limitations. In March 2020, Volvo issued a large-scale recall to the global market. The number reached 700,000, involving 9 models on sale. The reason for the recall was that during a safety test on the XC60 conducted by Volvo in Denmark, it was found that the automatic emergency braking system did not stop the vehicle in time in the event of a collision as expected.

There are more and more safety problems caused by intelligent driving. Whether it is a traffic accident or a recall, the reason is not all caused by the failure of the E / E system; in the automatic driving system, even if the system does not fail, it may Due to the uncertainty of complex intelligent algorithms, the deviation of functions, the performance limitations of sensors or systems, and the driver's misuse of vehicle functions cause traffic injuries

[0004] Since autonomous vehicle systems are more complex than normal vehicle systems, there is driver misuse of vehicle functions

However, self-driving vehicles cannot reduce the misoperation rate to zero under the influence of complex external environment interaction and vehicle human-computer interaction. Misoperations include the following situations: misuse of human-computer interaction inside and outside the vehicle; excessive reliance on automatic driving functions, for related early warning Ignore information or takeover instructions; intentionally take over the vehicle when it senses danger, etc., and some unforeseen misoperation events will cause risks to the expected functional safety of self-driving vehicles

Images