Vehicle Path Prediction Using Segmented Trajectory Phases
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current vehicle safety systems struggle to accurately predict the future path of a vehicle, leading to potential unnecessary activations that can be distracting and hazardous, as they often rely on simplistic or incomplete trajectory calculations.
Innovation Solution
A method that calculates three distinct phases of vehicle trajectory - an initial unavoidable phase, a phase based on current steering action, and a phase where the driver attempts to follow the driving lane - and combines these with varying contribution weights over time to formulate a comprehensive predicted path, using sensors for speed, direction, yaw rate, and lane detection to enhance accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a single-path prediction method is used, then the system complexity is low, but the prediction accuracy is insufficient leading to unnecessary safety system activations
Solution Approach 1:
The patent segments the future path prediction into three distinct time periods with different prediction models: first period uses current trajectory continuation, second period incorporates steering action effects, and third period accounts for driver lane-following behavior. This segmentation allows each period to be modeled with appropriate complexity, improving overall prediction accuracy while managing system complexity through structured division of the prediction task.
2Reliability
If multiple path predictions are combined, then the prediction reliability improves, but the calculation complexity increases
Solution Approach 1:
The patent performs preliminary calculations for each of the three path predictions separately before combining them. Each prediction is calculated independently based on its specific time period and model, allowing the system to prepare multiple reliable predictions in advance and then integrate them systematically, thereby improving reliability while managing calculation complexity through structured preliminary processing.
3Reliability
If safety systems are activated frequently, then the safety coverage is high, but the driver experience deteriorates due to unnecessary activations
Solution Approach 1:
The patent changes the parameter of prediction time horizon by considering three different time periods with varying assumptions about driver behavior and vehicle dynamics. This allows the system to distinguish between imminent hazards requiring immediate safety activation and longer-term trajectory variations that are normal driving behavior, thereby improving safety reliability while reducing unnecessary activations that cause driver distraction.
Data Source
AI summary
A method of predicting a future path of a vehicle, comprising the steps of: sensing a speed, and direction, and yaw rate of the vehicle; sensing a steering angle of the vehicle; sensing a driving lane near the vehicle, or along which the vehicle is being driven; calculating a first path prediction, for a first period of time following the current time, the first path prediction comprising a trajectory predicted based on the sensed speed and the direction and the yaw rate; calculating a second path prediction, for a second period of time, at least some of which is later than the first period of time, which assumes that a steering action arising from changes in the steering angle will take effect on the vehicle; calculating a third path prediction, for a third period of time, at least some of which is later than the second period of time, which assumes that the driver of the vehicle will control the trajectory of the vehicle to attempt to follow at least substantially the driving lane; and formulating a combined predicted path for the first, second and third periods of time, wherein the first, second and third path predictions each contribute to the combined predicted path.