Closed-Loop Driving System Evaluation for Error Propagation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Complex driving systems with multiple subsystems, such as perception, determination, and control systems, face challenges in optimizing their performance due to limitations in evaluating the validity of these interconnected components effectively.
Innovation Solution
The evaluation method models the interaction between subsystems and the real world as a closed loop structure, allowing for the representation and evaluation of errors propagating through these systems, thereby confirming the composite factors and validity of the driving system.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If the driving system includes multiple interconnected subsystems (perception, determination, control), then the system functionality and automation level are improved, but the complexity of evaluating system validity and error propagation increases
Solution Approach 1:
The patent segments the driving system into distinct subsystems (perception system, determination system, control system) and evaluates each subsystem's error characteristics separately. By dividing the complex system into manageable segments, the evaluation process becomes more tractable while still capturing the interconnected nature of the subsystems through closed-loop error propagation analysis.
Solution Approach 2:
The patent implements feedback by modeling closed loops that represent error propagation paths between subsystems. The evaluation method traces how errors from one subsystem feed into and affect other subsystems, creating a feedback mechanism that captures the dynamic interactions within the driving system. This allows for comprehensive validity evaluation by considering both individual subsystem performance and their combined behavior.
2Reliability
If traditional evaluation methods are used for driving systems, then the evaluation process is simpler, but the ability to confirm validity of interconnected subsystems is insufficient
Solution Approach 1:
The patent introduces an intermediary evaluation framework that mediates between individual subsystem evaluations and overall system validity confirmation. This framework uses error propagation modeling as an intermediary mechanism to connect subsystem-level assessments with system-level validity judgments, enabling comprehensive reliability evaluation without requiring direct testing of the entire complex system.
Solution Approach 2:
The patent changes evaluation parameters by shifting from static subsystem performance metrics to dynamic error propagation characteristics. By evaluating how errors change and propagate through the system under different operating conditions, the method provides a more nuanced and reliable validity confirmation that accounts for the interconnected nature of subsystems.
Data Source
AI summary
A driving system includes a perception system, a determination system, and a control system as subsystems. An evaluation method for the driving system includes: specifying a closed loop by modeling, as a loop structure, an interaction between each of the subsystems and a real world; specifying an error occurring in each of the subsystems; and evaluating the error propagating in accordance with the closed loop.