Intersection Speed Model Generation for Autonomous Vehicles

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

In automatic driving scenarios, predicting the behavior of other vehicles at intersections is challenging due to complex speed changes, making it difficult for vehicles to accurately anticipate the movements of surrounding vehicles.

Innovation Solution

An information processing apparatus with a travel information acquisition unit, intersection information acquisition unit, route specification unit, control point detection unit, and model generation unit that predicts the speed of a moving object at an intersection by specifying a reference route and detecting speed control points to generate a speed model that minimizes speed changes, allowing for accurate speed prediction and trajectory calculation.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If the vehicle attempts to predict other vehicles' behavior at intersections using conventional methods, then the prediction capability is limited, but the system complexity remains low

Engineering Contradiction:
Improveprediction accuracyVSAvoidsystem complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent segments the intersection prediction problem into distinct components: detecting speed control points, generating speed models for different vehicle types, and calculating trajectories. This segmentation allows the system to handle complex predictions through modular, manageable sub-tasks, improving prediction accuracy without overwhelming system complexity

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The system changes parameters by generating different speed models for different vehicle types (e.g., vehicles with different acceleration patterns, sizes, or destinations). By adjusting speed model parameters based on vehicle classification and detected control points, the system achieves higher prediction accuracy while maintaining a structured approach to complexity

Inventive Principle:
Principle #35Parameter changes

2Loss of energy

If the vehicle reduces speed at intersections to match traffic flow, then fuel efficiency improves, but travel time increases

Engineering Contradiction:
Improvefuel efficiencyVSAvoidtravel time
Core Design Contradiction:
Loss of energyVSLoss of time

Solution Approach 1:

The system performs preliminary actions by detecting speed control points and generating speed models in advance before the vehicle reaches the intersection. This allows the vehicle to optimize its speed profile proactively, reducing speed smoothly at appropriate points to match traffic flow and improve fuel efficiency while minimizing unnecessary delays

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system applies dynamics by generating dynamic speed patterns that alternate between acceleration and inertia travel sections. The vehicle adjusts its speed dynamically based on predicted trajectories and detected control points, allowing it to cooperate with traffic flow for fuel efficiency while maintaining optimal travel time through adaptive speed control

Inventive Principle:
Principle #15Dynamics

Data Source

PatentEP3425341B1Information processing apparatus, vehicle information processing method, and computer-readable medium
Publication Date: 2022.12.21 KK TOSHIBA
  • EP3425341B1 patent drawingFigure 1~2
  • EP3425341B1 patent drawingFigure 3
  • EP3425341B1 patent drawingFigure 4

AI summary

According to an arrangement, an information processing apparatus (10) includes a travel information acquisition unit (42), an intersection information acquisition unit (44), a route specification unit (48), a control point detection unit (50), and a model generation unit (52). The travel information acquisition unit (42) acquires a dynamic state related to traveling of a moving object entering an intersection. The intersection information acquisition unit (44) acquires intersection information indicating a configuration of the intersection. The route specification unit (48) specifies a reference route along which the moving object is predicted to travel at the intersection, based on the dynamic state and the intersection information. The control point detection unit (50) detects a speed control point included in the specified reference route. The model generation unit (52) generates a speed model representing a temporal change in a predicted speed of the moving object so that the speed at the speed control point is locally minimized, based on the dynamic state and the intersection information.