Road Surface Condition Control System Using Dynamic Learning Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing vehicular control systems struggle to accurately determine road surface conditions, particularly on surfaces like muddy, sandy, and rocky roads, as they rely on differences in angular acceleration which are affected by vehicle aging and fuel type, leading to inaccuracies beyond paved and sandy roads.
Innovation Solution
A control system that uses a learned model based on supervised learning from traveling data, including lateral and longitudinal acceleration, rotational speed, and operational data, to estimate and determine road surface conditions, with separate models for low and high modes to account for varying road characteristics.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If the determination device uses the difference between presumed angular acceleration and actual angular acceleration to determine road surface condition, then the device can determine whether the vehicle travels on a paved road or sandy road, but the determination accuracy deteriorates on muddy roads, rocky roads, and other road types where road surface features change
Solution Approach 1:
The system dynamically adapts the determination period based on the detected road surface condition. When a rocky road is detected, the system extends the determination period to capture the characteristic features of rocky roads that appear later in time. This dynamic adjustment allows the system to maintain high determination accuracy across different road types by optimizing the observation window for each specific road condition.
Solution Approach 2:
The system changes the determination period parameter according to the detected road surface condition. For rocky roads, the determination period is extended beyond the standard period used for paved or sandy roads. This parameter change enables the system to capture the delayed appearance of rocky road features and improve determination accuracy for this specific road type.
2Productivity
If the system collects traveling data for a fixed period to determine road surface condition, then the processing is simple and fast, but the determination accuracy deteriorates for rocky roads where road surface features appear later
Solution Approach 1:
The system implements dynamic determination period adjustment based on road surface condition detection. When rocky road conditions are detected, the system automatically extends the data collection period to capture the characteristic features that appear later in time. This dynamic approach maintains high processing efficiency for common road types while improving accuracy for rocky roads when needed.
Solution Approach 2:
The system uses feedback from the learned model's road surface condition estimation to adjust the determination period. The learned model continuously analyzes traveling data and provides feedback about the current road condition, which then triggers appropriate adjustments to the determination period. This feedback mechanism ensures that the system collects data for the optimal duration needed to accurately determine each specific road type.
3Reliability
If the determination device uses error rate correction based on average acceleration values, then the presumed acceleration can be corrected for vehicle aging and fuel type, but the determination accuracy still deteriorates on road types where road surface features change over time
Solution Approach 1:
The system combines error rate correction with dynamic determination period adjustment. The error rate correction based on average acceleration values addresses vehicle aging and fuel type variations, while the dynamic extension of the determination period for rocky roads captures the temporal characteristics of rocky road features. This combination maintains reliability across different vehicle conditions while improving accuracy for specific road types.
Solution Approach 2:
The system performs preliminary error rate correction using average acceleration values to account for vehicle aging and fuel type before proceeding with road surface condition determination. This preliminary action establishes a baseline correction that improves the accuracy of subsequent road type classification, especially when combined with extended determination periods for rocky roads.
Data Source
AI summary
A control system configured to accurately determine a condition of a road surface on which a vehicle travels. A learned model estimates the road surface condition based on the travelling data collected during propulsion of the vehicle, and a controller determines the road surface condition based on the road surface condition estimated by the learned model.