Drive Assistance System with Prediction Locus Estimation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing drive assistance systems face challenges in effectively avoiding collisions with multiple objects around a vehicle, as they often increase the risk of interference with other objects while attempting to avoid one, such as when braking for a pedestrian, potentially leading to collisions with oncoming vehicles.
Innovation Solution
A novel drive assistance system that designates a primary and secondary target based on the probability of interference, using object detection and prediction loci to execute avoidance controls, ensuring the vehicle can navigate without colliding with multiple objects by estimating and adjusting its path to avoid both primary and secondary targets.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If collision avoidance control is executed by braking or steering to avoid a detected object, then the own vehicle can avoid collision with the primary target, but the risk of collision with other objects increases
Solution Approach 1:
The system performs preliminary estimation of prediction loci and identification of both primary and secondary targets before executing avoidance control. By predicting where the vehicle will move when avoiding the primary target, the system proactively identifies secondary targets that may become problematic, allowing for coordinated avoidance planning that prevents chain reactions of collisions
Solution Approach 2:
The prediction locus estimation unit acts as an intermediary between object detection and avoidance control execution. It simulates the vehicle's movement path when avoiding the primary target and uses this intermediate information to identify secondary targets, enabling the system to mediate between avoiding one object and potentially colliding with another
2Reliability
If the drive assistance system focuses on avoiding a single detected object, then collision with that specific object can be prevented, but interference with multiple objects existing around the own vehicle cannot be avoided
Solution Approach 1:
The system segments the avoidance problem into primary target avoidance and secondary target avoidance. By dividing the complex multi-object avoidance task into hierarchical levels (primary target first, then secondary targets based on prediction locus), the system can systematically handle multiple objects without being overwhelmed by the complexity of simultaneous multi-object avoidance
Solution Approach 2:
The system adds the dimension of prediction locus estimation to the traditional single-target avoidance approach. Instead of only considering current object positions, it incorporates the predicted future position of the vehicle when avoiding the primary target, creating a two-dimensional assessment (current position + predicted position) that enables multi-object avoidance capability
Data Source
AI summary
A drive assistance system includes a primary target setter to designate an object determined based on a current movement locus of the own vehicle to be a primary target probably interfering with the own vehicle; a prediction locus estimation unit to estimate a prediction locus along which the own vehicle moves when primary avoidance control is executed in the own vehicle to avoid the interference with the primary target; and a secondary target setter to designate another object determined based on the prediction locus as a secondary target probably interfering with the own vehicle. The drive assistance system also includes a drive assistance controller to execute drive assistance control in the own vehicle based on a determination of whether the interference by the secondary target can be avoided by executing a secondary avoidance control in the own vehicle to avoid the interference by the secondary target.