Autonomous Lane Change Offset Prediction Under Uncertain Yielding
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing autonomous vehicle technologies face inefficiencies in lane changes due to reliance on environmental vehicles yielding, which is not consistently honored, leading to manual intervention or waiting for environmental vehicles to yield, rather than proactive right-of-way assertion.
Innovation Solution
A vehicle-based data processing method that determines predicted offsets and payoffs for lane changes, considering the traveling states of both the ego vehicle and reference vehicles, to calculate a target predicted offset for optimal lane change decision-making, thereby improving lane change efficiency by actively occupying the reference vehicle's space.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If autonomous vehicles rely on environmental vehicles to yield for lane changes, then safety is improved by respecting right of way, but lane change efficiency deteriorates due to waiting and manual intervention
Solution Approach 1:
The system performs preliminary actions by predicting the yielding probability of environmental vehicles before executing lane changes. It calculates predicted offsets and determines whether environmental vehicles will yield in advance, allowing the autonomous vehicle to proactively assert right of way when yielding probability is low, rather than passively waiting
Solution Approach 2:
The system changes the decision-making parameter from binary (wait or manual intervention) to probabilistic (yielding probability threshold). By introducing a configurable threshold parameter, the system can adjust the balance between safety and efficiency, automatically deciding whether to proceed with lane change based on predicted yielding probability
2Productivity
If autonomous vehicles proactively assert right of way by occupying reference vehicle space, then lane change efficiency is improved, but safety may worsen if environmental vehicles do not yield
Solution Approach 1:
The system performs preliminary risk assessment by predicting the yielding probability of environmental vehicles before executing proactive lane changes. It calculates predicted offsets and determines the likelihood of conflict in advance, only asserting right of way when the predicted probability of environmental vehicle yielding is below the threshold, thus avoiding unsafe maneuvers
Solution Approach 2:
The system uses feedback from predicted yielding probability to adjust lane change decisions. By continuously monitoring the calculated probability and comparing it against the threshold, the system provides feedback to the decision-making process, ensuring that proactive lane changes are only executed when safety conditions are met
Data Source
Figure 1
Figure 2a~2b
Figure 2c
AI summary
A vehicle-based data processing method and apparatus, a computer, and a readable storage medium, relating to the automatic driving technology. The method comprises: determining at least two prediction offsets of a first vehicle, a first traveling state of the first vehicle, and a second traveling state of a second vehicle (S401); determining, according to the first traveling state and the second traveling state, a first lane change benefit of each prediction offset when the second vehicle is in a yielding prediction state and a second lane change benefit of each prediction offset when the second vehicle is in a non-yielding prediction state (S402); and determining a prediction yielding probability of the second vehicle, generating a target lane change benefit of each prediction offset according to the prediction yielding probability and the first and second lane change benefits of each prediction offset, and determining the prediction offset having a maximum target lane change benefit as a target prediction offset (S403).