Driving Model Generation for Accurate Skill Evaluation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing driving evaluation systems fail to provide accurate assessments of driving skills as they primarily focus on variability in deceleration and fuel consumption, neglecting specific driving situations and environments, leading to incorrect evaluations.
Innovation Solution
A driving model generation device and method that detects vehicle state quantities during specific driving operations, generating models based on initial and final states to reflect appropriate driving behaviors in various situations, allowing for evaluation and support without relying on environmental conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If evaluation is made based on variability in deceleration correlation, then measurement precision is improved, but adaptability deteriorates because it cannot account for different driving situations
Solution Approach 1:
The patent applies dynamics by making the evaluation system adaptable to different driving situations through dynamic classification. Instead of using a fixed evaluation criterion, the system dynamically selects appropriate evaluation models based on the detected driving situation (e.g., traffic jam, curve, slope), allowing the evaluation to adapt to varying conditions while maintaining measurement precision for each specific situation.
Solution Approach 2:
The patent changes the evaluation parameters based on driving situations. Different evaluation models with different parameters are selected and applied according to the situation classification. For example, different deceleration thresholds, time constants, or evaluation criteria are used for traffic jams versus curves, allowing accurate measurement for each situation while improving overall adaptability.
2Reliability
If abrupt braking operation is performed to stop before traffic light, then stopping reliability is improved, but energy consumption increases
Solution Approach 1:
The patent applies parameter changes by adjusting the evaluation criteria for braking operations based on the driving situation. For traffic light scenarios, the system recognizes the situation and applies appropriate evaluation parameters that account for the necessity of abrupt braking, thereby evaluating the reliability improvement while understanding the energy trade-off, rather than uniformly penalizing all abrupt braking.
Solution Approach 2:
The patent inverts the traditional evaluation approach by not simply penalizing abrupt braking but rather evaluating whether the braking was appropriate for the situation. Instead of assuming abrupt braking is always inefficient, the system inverts the logic to recognize that situation-appropriate abrupt braking may be the correct behavior, thus evaluating reliability separately from energy consumption in a context-dependent manner.
Data Source
AI summary
A driving model generation device is provided that enables driving evaluation according to a traveling situation of a vehicle. A vehicle is provided with a vehicle state detector, which detects a vehicle state quantity that changes according to a driving operation by a driver. A model generator generates a driving model, which is a reference related to a specific driving operation, based on the vehicle state quantity at the time the driver starts the specific driving operation and the vehicle state quantity at the time the specific driving operation is finished. A driving evaluator evaluates the driving skill of the driver by comparing the driving operation of the driver presented by a detection result of the vehicle state detector to the driving model generated by the model generator.