Dynamic Obstacle Navigation for Static-Free Robot Map Generation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing robot autonomous map building technologies require a static map for planning, which is impractical due to frequent obstacle changes in real-world environments, leading to performance issues and risks.
Innovation Solution
A map generation method that navigates a terminal device to a first obstacle and determines a next navigation point based on candidate obstacles, updating them in real time to generate a map without a static map, using two scanning ranges to reduce repeated exploration.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a static map is used for map building planning, then the planning process can be completed with a known map structure, but the method becomes impractical when obstacles change frequently in the real world
Solution Approach 1:
The patent transitions from static map-based planning to dynamic real-time obstacle detection and navigation. The terminal device continuously scans the environment using sensors, detects obstacles in real-time, and dynamically adjusts the navigation trajectory based on current obstacle positions, making the system adaptable to changing environments while maintaining reliable map building performance
Solution Approach 2:
The patent performs preliminary obstacle detection by scanning the environment before finalizing the navigation trajectory. The terminal device detects obstacles in advance, predicts potential collision points, and pre-adjusts the navigation path to avoid obstacles, enabling proactive adaptation rather than reactive response
2Loss of information
If the terminal device explores all regions comprehensively, then complete map coverage is achieved, but energy consumption increases and repeated exploration occurs
Solution Approach 1:
The patent implements a feedback mechanism where the terminal device continuously monitors its navigation trajectory, compares it with the generated map data, and identifies regions that have already been explored. This feedback loop prevents repeated exploration of the same areas, reducing energy consumption while ensuring complete map coverage through efficient path planning
Solution Approach 2:
The patent divides the exploration space into scanned regions and unscanned regions based on sensor data. By segmenting the environment into explored and unexplored zones, the terminal device can focus its energy on unscanned areas while avoiding re-exploration of scanned regions, achieving complete coverage with optimized energy usage
Data Source
AI summary
This application discloses a map generation method, which includes: first obtaining a plurality of obstacles in a scanning range of a sensor; controlling a terminal device to navigate to a first obstacle, where the first obstacle is an obstacle closest to the terminal device in the plurality of obstacles; then determining a next navigation point of the terminal device, where the next navigation point is an obstacle whose distance from the first obstacle is the shortest in a plurality of candidate obstacles, or the next navigation point is an obstacle that is not selected in the plurality of obstacles and that the terminal device does not pass through; and generating a map based on a navigation trajectory of the terminal device.