Autonomous Vehicle Traffic Planning with Gap-Balancing Control

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Centralized vehicle control systems often lead to vehicles blocking each other, resulting in queue formation and deadlock states, which reduce system productivity and require external intervention.

Innovation Solution

A traffic planning method that optimizes a state-action value function with a command-independent term penalizing small inter-vehicle gaps, using a gap-balancing indicator to promote even gap distribution, reducing the likelihood of blocking states and improving system efficiency with reduced computational resources.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If centralized control is used to maximize productivity, then system output increases, but vehicles may block each other forming queues and deadlocks

Engineering Contradiction:
Improvesystem productivityVSAvoidblocking state occurrence
Core Design Contradiction:
ProductivityVSReliability

Solution Approach 1:

The patent applies preliminary action by introducing a command-independent term in the objective function that penalizes small inter-vehicle gaps before deadlocks occur. This predictive mechanism proactively prevents blocking states by maintaining adequate spacing between vehicles, rather than reacting after conflicts arise. The gap-balancing indicator calculates and adjusts commands to ensure minimum safe distances are maintained throughout the planning horizon.

Inventive Principle:
Principle #10Preliminary action

2Productivity

If traditional optimization methods are used to determine motion commands, then productivity is maximized, but computational resources are heavily consumed

Engineering Contradiction:
Improveoptimization efficiencyVSAvoidcomputational resources
Core Design Contradiction:
ProductivityVSUse of energy by moving object

Solution Approach 1:

The patent segments the objective function into two distinct parts: a command-independent term that penalizes small inter-vehicle gaps, and a command-dependent term that optimizes productivity. This segmentation allows the system to handle gap balancing separately from productivity optimization, reducing the computational complexity of the overall optimization problem while maintaining both safety and efficiency goals.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent changes the parameter structure by introducing a gap-balancing indicator that transforms the optimization landscape. By reformulating the objective function to include this indicator, the system achieves better convergence properties and reduces computational resource consumption while still maximizing productivity through the command-dependent term.

Inventive Principle:
Principle #35Parameter changes

3Speed

If vehicles are allowed to move freely to maximize speed, then individual vehicle efficiency increases, but queue formation and deadlocks occur

Engineering Contradiction:
Improvevehicle speedVSAvoidsystem efficiency
Core Design Contradiction:
SpeedVSProductivity

Solution Approach 1:

The patent implements feedback by continuously monitoring inter-vehicle gaps and using the gap-balancing indicator to adjust motion commands. The objective function provides real-time feedback on spacing conditions, allowing the system to dynamically modify vehicle speeds and trajectories to prevent queue formation while maintaining overall system efficiency and avoiding deadlocks.

Inventive Principle:
Principle #23Feedback

Data Source

PatentEP4113241B1Method and system for traffic control of a plurality of vehicles, in particular autonomous vehicles
Publication Date: 2023.08.16 VOLVO AUTONOMOUS SOLUTIONS AB
  • EP4113241B1 patent drawingFigure 1
  • EP4113241B1 patent drawingFigure 2
  • EP4113241B1 patent drawingFigure 3

AI summary

A traffic planning method for controlling a plurality of vehicles (v1, v2, v3, v4), wherein each vehicle occupies one node in a shared set of planning nodes (wp1, wp2, wp3, wp4, wp5, wp6, wp7, wp8) and is movable to other nodes along predefined edges between pairs of the nodes in accordance with a finite set of motion commands. In the method, initial node occupancies of the vehicles are obtained, and a sequence of motion commands is determined by optimizing a state-action value function Q(s, a) = Qs (s, a) + QL (s) which depends on node occupancies s and the motion commands a to be given. The objective function includes a command-dependent term Qs (s, a), which may be updated in each iteration based on a reward function, and a command-independent term QL (s), which penalizes node occupancies with too small inter-vehicle gaps.