Autonomous Parking Path Planning With Neural Node Prioritization

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing pathfinding algorithms for autonomous vehicle parking, such as Dijkstra and heuristic cost functions, suffer from exponential computational time increases as the number of nodes increases, making them inefficient in complex parking scenarios.

Innovation Solution

Integrate a neural network trained via reinforcement learning to predict optimal vehicle paths, reducing the number of nodes analyzed by prioritizing the least costly paths and using traditional algorithms iteratively to handle obstacles.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If traditional pathfinding algorithms (Dijkstra, A*) are used to determine optimal vehicle paths, then path optimality is achieved, but computational time increases exponentially as the number of nodes increases

Engineering Contradiction:
Improvepath optimalityVSAvoidcomputational time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The system performs preliminary actions by using the neural network to predict the most promising next node and prioritize path exploration directions before exhaustive search begins. This pre-computation of likely path directions reduces the search space that traditional algorithms must explore, thereby reducing computational time while maintaining path optimality through subsequent verification with Dijkstra or A* algorithms.

Inventive Principle:
Principle #10Preliminary action

2Productivity

If heuristic cost functions (A*, D*) are used to reduce computational time, then computational efficiency improves, but the ability to handle complex obstacle scenarios may be compromised

Engineering Contradiction:
Improvecomputational efficiencyVSAvoidobstacle handling capability
Core Design Contradiction:
ProductivityVSReliability

Solution Approach 1:

The neural network serves as an intermediary between the heuristic cost functions and the pathfinding process. It predicts promising path directions and prioritizes node exploration based on learned patterns from training data, while traditional algorithms (Dijkstra, A*) act as intermediaries to verify optimality and handle complex obstacle scenarios. This layered intermediary approach allows the system to benefit from both the speed of heuristics and the reliability of traditional algorithms.

Inventive Principle:
Principle #24Intermediary (Mediator)

3Reliability

If exhaustive search is performed to ensure optimal path finding, then path optimality is guaranteed, but the computational load increases significantly in complex parking scenarios

Engineering Contradiction:
Improvepath optimality guaranteeVSAvoidcomputational load
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The system performs preliminary filtering of the search space using the neural network's predictions before applying exhaustive search algorithms. By pre-identifying and prioritizing promising path directions based on learned patterns, the system reduces the number of nodes that require exhaustive exploration, thereby maintaining path optimality guarantees while significantly reducing computational load in complex parking scenarios.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system applies partial exhaustive search by using the neural network to identify and prioritize the most promising subset of nodes for exploration. Instead of uniformly exploring all possible paths, the system focuses computational resources on the most likely optimal paths predicted by the neural network, performing exhaustive search only on this reduced subset, thereby reducing overall computational load while maintaining optimality for the critical path.

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentEP4610145A1Path determination for autonomous vehicle parking
Publication Date: 2025.09.03 APTIV TECHNOLOGIES AG
  • EP4610145A1 patent drawingFigure 1A~1D
  • EP4610145A1 patent drawingFigure 2
  • EP4610145A1 patent drawingFigure 3A

AI summary

A method of determining a movement path for a vehicle includes determining a set of adjacent nodes with respect to a current node. The method includes determining, via a neural network, a subsequent path node and determining whether a path based on the subsequent path node connects a source to a goal node. The method includes, in response to a determination that the path based on the subsequent path node does not connect the source node to the goal node, determining an alternative subsequent path node, and determining, via the neural network, a second path based on the alternative subsequent path node. The method includes, in response to a determination that the path based on the subsequent path node connects the source to the goal node, selecting the path based on the subsequent path node as the movement path, and executing a set of actions associated with the movement path.