Vehicle Controller Trajectory Planning Using Other Vehicle Behavior
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing vehicle control systems face inaccuracies in predicting the behavior of other vehicles due to variability in driver behavior regarding turn signals and lane changes, leading to potential collisions.
Innovation Solution
A vehicle controller that uses time series images from cameras to detect and track other vehicles, identify signal light states, and extract behavior characteristics, such as lane change rates and turn signal reliability, to predict trajectories and adjust cost maps for safe navigation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If turn signal information is used to predict vehicle behavior, then prediction capability is provided, but prediction accuracy is insufficient due to driver variability
Solution Approach 1:
The prediction system segments vehicle behavior into multiple independent characteristics: turn signal usage, lane change actions, time lag patterns, and frequency of lane changes. Each segment is analyzed separately to build a comprehensive behavior profile, improving overall prediction accuracy while maintaining system reliability.
Solution Approach 2:
The system transitions from two-dimensional turn signal detection to multi-dimensional behavior analysis by incorporating temporal dimensions (time lag between signal and action), frequency dimensions (how often lane changes occur), and action dimensions (actual lane change vs. no lane change). This dimensional expansion resolves the accuracy limitation.
2Device complexity
If simple turn signal detection is used, then system complexity is reduced, but prediction accuracy deteriorates due to ignoring driver behavior variability
Solution Approach 1:
The system performs preliminary tracking and data collection on vehicle behavior patterns before making predictions. By pre-analyzing time lags, frequency, and action correlations during a tracking period, the system builds behavioral profiles that improve prediction accuracy without adding complex real-time processing requirements.
Solution Approach 2:
The prediction system uses the vehicle's own behavioral data to improve its predictions. By analyzing the target vehicle's historical turn signal usage, lane change patterns, and time lag characteristics, the system self-adjusts its prediction model, reducing the need for external calibration while improving accuracy.
3Measurement precision
If tracking period is extended to gather more behavior data, then prediction accuracy improves, but response time increases
Solution Approach 1:
The tracking period is made dynamic rather than fixed. The system adjusts the tracking duration based on the amount of behavioral data collected and the confidence level achieved. When sufficient patterns are identified (turn signal correlations, time lags, frequency), the system automatically concludes the tracking period, optimizing the balance between accuracy and response time.
Solution Approach 2:
The system changes key parameters during tracking: the weighting of different behavioral characteristics, the threshold for confidence, and the duration of tracking. By dynamically adjusting these parameters based on data quality and quantity, the system achieves accurate predictions without excessive tracking time delays.
Data Source
AI summary
A vehicle controller includes a processor configured to detect an object region including another vehicle near a vehicle from each of time series images obtained by a camera mounted on the vehicle; detect a predetermined action taken by the other vehicle, based on a trajectory of the other vehicle estimated from the object region of each image; identify the state of a signal light of the other vehicle, based on characteristics obtained from pixel values of the object region of each image; extract information indicating characteristics of an action of the other vehicle or the state of a signal light at the predetermined action taken by the other vehicle, based on the predetermined action detected in a tracking period and the state of the signal light related to the predetermined action; and predict behavior of the other vehicle, using the extracted information.