Closed-Loop Simulation Qualification for Semi-Automated Driving Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods lack a well-founded approach to computationally ascertain uncertain models for simulated control of mobile platforms in closed-loop simulations, particularly for semi-automated driving, where recreating critical vehicle behaviors is complex and requires accurate replication of operative states to validate control systems.
Innovation Solution
A computer-implemented method for comparing generated data sequences from closed-loop simulations with recorded data sequences of semi-automated mobile platform trips, using similarity metrics and determinants to categorize the control qualification, ensuring the simulation accurately represents real-world scenarios.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If closed-loop simulation is used to validate control systems, then the complexity of validation is reduced, but the accuracy of replicating real-world operative states deteriorates
Solution Approach 1:
The patent replaces physical closed-loop simulation with a computational data comparison approach. Instead of running complex simulations, the method compares recorded trip data directly with simulated data sequences, substituting mechanical simulation processes with information processing to achieve accurate validation.
Solution Approach 2:
The patent creates copies of real trip data through simulation and compares these copies with actual recorded data. By generating simulated data sequences that replicate real-world operative states and comparing them systematically, the method achieves accurate validation without requiring complex physical simulations.
2Measurement precision
If more determinants are compared between simulated and recorded data, then the qualification accuracy improves, but the computational effort increases
Solution Approach 1:
The patent segments the comparison process into distinct evaluation classes (first, second, third) based on the degree of similarity between determinants. This segmentation allows systematic comparison of multiple determinants while managing computational effort through hierarchical evaluation, where not all determinants require the same level of analysis.
Solution Approach 2:
The patent changes the state of data sequences by categorizing them into different evaluation classes based on similarity thresholds. By adjusting parameters such as similarity limits and evaluation criteria, the method optimizes the balance between qualification accuracy and computational effort required for comparing determinants.
3Reliability
If the simulation process is made more realistic, then the validation reliability improves, but the ease of operation deteriorates
Solution Approach 1:
The patent inverts the traditional simulation approach by not trying to make the simulation perfectly realistic, but rather by comparing simulated data with recorded real-world data. This inversion simplifies the setup process while maintaining validation reliability, as the method works with whatever simulation data is available rather than requiring highly realistic simulation models.
Data Source
AI summary
A computer-implemented method for comparing generated data sequences for an at least semi-automated driving of a mobile platform, which were generated with the aid of a closed-loop simulation process, and a recorded data sequence of a trip of the mobile platform, controlled in at least semi-automated fashion, for the qualification of the control. The method includes: providing the recorded data sequence, which is based on a multiplicity of determinants, of trips of the mobile platform controlled in at least semi-automated fashion; providing a multitude of generated data sequences, which are based on the multiplicity of determinants, of simulated trips, which were generated with the aid of the closed-loop simulation process; providing similarity limits and a similarity metric for the respective determinant; comparing the recorded data sequence to each individual generated data sequence of the multitude of recorded data sequences.