Mobile Robot Path Planning With Pedestrian-Aware Sampling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current path planning algorithms for mobile service robots in man-machine environments ignore environmental uncertainties, leading to unsafe and unreliable paths due to unconsidered influences of errors in environmental modeling and pedestrian movements.
Innovation Solution
A sampling method and system that calculates distances to obstacles, detects pedestrians, and uses a search tree to randomly select growth nodes based on cumulative cost calculations, incorporating collision detection to ensure path reliability and safety by considering environmental uncertainties.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If random sampling-based path planning algorithm is used, then planning time and memory cost are reduced, but the influence of environmental uncertainties on tree growth is ignored leading to unsafe paths
Solution Approach 1:
The patent pre-calculates and stores distance maps showing the distance from each grid point to the nearest obstacle before path planning begins. This preliminary action allows the algorithm to quickly assess environmental uncertainties during tree growth without performing complex real-time calculations, thus maintaining fast planning while improving path safety by incorporating uncertainty information.
Solution Approach 2:
The patent introduces a cost function as an intermediary that integrates multiple factors including distance to obstacles, pedestrian positions, and environmental uncertainties. This cost function mediates between the simplicity of random sampling and the need for safety by providing a unified metric that guides tree growth toward safer paths while maintaining computational efficiency.
2Device complexity
If only relationships between internal nodes in tree are considered, then algorithm complexity is reduced, but environmental uncertainties such as positioning errors and moving pedestrians are ignored
Solution Approach 1:
The patent pre-processes environmental data including pedestrian positions and obstacle locations before path planning, storing them in accessible data structures. This preliminary action allows the algorithm to efficiently query environmental uncertainties during tree growth without adding significant computational complexity, thus maintaining algorithm simplicity while improving path reliability.
Solution Approach 2:
The patent modifies the cost function parameters to include weights for different uncertainty factors such as positioning errors and pedestrian movements. By adjusting these parameters, the algorithm can account for environmental uncertainties without fundamentally changing its structure, thus maintaining low complexity while improving reliability through parameter optimization.
3Adaptability or versatility
If distance calculation to closest obstacle is performed online, then path adaptability is improved, but calculation time and processing load increase
Solution Approach 1:
The patent calculates and stores distance maps showing the distance from each grid point to the nearest obstacle in advance, before the path planning process begins. This preliminary calculation allows the robot to quickly adapt to different paths by simply querying pre-computed distance values rather than performing complex real-time distance calculations, thus improving both adaptability and reducing calculation time.
Solution Approach 2:
The patent pre-calculates distance information for all possible positions in the environment, which is more than what is strictly needed for any single path. This excessive pre-computation enables extremely fast online querying and path adaptation, as the robot only needs to look up pre-computed values rather than calculate distances during path planning.
Data Source
AI summary
The invention relates to a sampling method and system for path planning of a mobile robot in a man-machine environment, including: calculating a distance between a node and the closest obstacle on a map, simultaneously detecting a pedestrian in an environment and marking a position of the pedestrian; selecting a starting point position as a root node, and initializing a search tree; choosing a candidate node in a passable region, calculating a cumulative cost from a node to the candidate node, and choosing a node with the lowest cost as a growth node; performing collision detection on a connecting line between the growth and the candidate nodes, and determining whether collision detection succeeds; connecting the candidate node to the growth node, and determining whether the candidate node is a goal; and acquiring a node connecting line set from the root node to the goal to form a final path.