Autonomous Driving Control Using Advance Road Data and Experience
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional autonomous driving systems face challenges in quickly adjusting routes, performing complex driving actions like lane changing and branching, and accurately identifying obstacles, leading to potential safety issues due to delayed route recalculation and reliance on detailed maps, which are costly to construct and time-consuming to process.
Innovation Solution
A method and apparatus that utilize preset local information, including road attributes and user experience data, to determine optimal routes and driving speeds, allowing for real-time obstacle detection and smooth driving by generating driving control commands based on current vehicle states and target speeds.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If conventional autonomous driving systems recalculate routes in real-time when road conditions change, then the vehicle can adapt to current conditions, but the calculation takes too much time causing the vehicle to stop or move to unplanned roads
Solution Approach 1:
The system pre-calculates and stores multiple alternative routes in advance before the vehicle starts driving. When the vehicle needs to adapt to changing road conditions, it can immediately select from the pre-computed routes without time-consuming recalculation, thus resolving the contradiction between route adaptability and calculation time
Solution Approach 2:
The system dynamically selects from multiple pre-calculated routes based on real-time road conditions. The route selection is flexible and adaptive, allowing the vehicle to switch between pre-computed routes as conditions change, maintaining adaptability without requiring time-consuming recalculation
2Manufacturing precision
If conventional autonomous driving systems use detailed lane-level maps for navigation, then routing precision is improved, but constructing the map is expensive and calculating initial routes takes a great deal of time
Solution Approach 1:
The system pre-calculates multiple alternative routes based on lane-level map data before the vehicle starts driving. This preliminary route calculation stores multiple options in advance, allowing the vehicle to quickly select an appropriate route without time-consuming calculations during actual driving, thus resolving the contradiction between routing precision and calculation time
Solution Approach 2:
The system applies detailed lane-level map information locally to specific route segments that require high precision, rather than processing the entire route with maximum detail. This selective application of detailed mapping reduces overall calculation time while maintaining necessary routing precision
3Reliability
If conventional autonomous driving systems detect and avoid obstacles through sensor detection, then obstacle avoidance is achieved, but the system has difficulty clearly identifying obstacle characteristics and reflects sensing results with delay
Solution Approach 1:
The system pre-identifies potential obstacle zones and prepares avoidance routes in advance based on map data and historical information. When sensors detect actual obstacles, the vehicle can immediately execute pre-planned avoidance maneuvers, reducing the delay in reflecting sensing results while maintaining reliable obstacle avoidance
Data Source
AI summary
Provided is a method of controlling driving of a vehicle using advance information, the method including acquiring preset local information including a road name, a road section, a road attribute, a location of a building, a lane, a traffic signal, and obstacle information for a predetermined region, acquiring a driving experience value resulting from a previous drive using the local information, and setting a target speed corresponding to the road attribute using the driving experience value, determining a driving state and a driving speed of the vehicle on the basis of the local information, and the target speed, of a current position of the vehicle, and generating a driving control command corresponding to the driving state and the driving speed.