Vehicle Dynamics Model Library for Real-Time Control Domains
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current vehicle dynamic models are computationally heavy, making them inefficient for real-time processing in autonomous vehicles, and require complex simulations that are not suitable for on-board controllers.
Innovation Solution
A computer-implemented method to derive a library of optimized vehicle dynamic implementation models, each correlated to specific operational design domains, by defining operational design domains, choosing an abstract model, parameterizing and simulating it, optimizing based on reference trajectories, and validating the models to create simplified, efficient models for different use cases.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional vehicle dynamic models are used, then model accuracy and reliability are improved, but computational complexity and processing time increase
Solution Approach 1:
The patent segments the vehicle dynamic model into multiple simplified implementation models, each optimized for specific operational design domains (e.g., low-speed maneuvering, high-speed cruising, transient conditions). This segmentation allows the system to use computationally lightweight models appropriate for each domain rather than a single complex model, resolving the contradiction between accuracy and computational complexity.
Solution Approach 2:
The patent implements dynamic model selection where the system automatically chooses or switches between different simplified implementation models based on current operational conditions. This dynamic adaptation maintains model accuracy by selecting the most appropriate simplified model for each situation, while keeping computational complexity low by avoiding unnecessary use of complex models.
2Adaptability or versatility
If comprehensive vehicle dynamic models are used, then coverage of all use cases is improved, but processing speed decreases
Solution Approach 1:
The patent applies local quality by creating specialized simplified implementation models tailored to specific operational design domains and use cases. Each model is optimized locally for its intended domain (e.g., parking maneuvers, highway cruising, obstacle avoidance), providing high adaptability within each domain while maintaining fast processing speeds by avoiding the overhead of comprehensive models.
Solution Approach 2:
The patent changes model parameters and structural complexity based on operational conditions. By adjusting model parameters and selecting different simplified models according to the current operational design domain, the system achieves comprehensive use case coverage while maintaining high processing speeds through parameter-driven model adaptation.
3Productivity
If simplified models are used, then computational efficiency is improved, but model accuracy for regression testing may be insufficient
Solution Approach 1:
The patent performs preliminary validation and calibration of simplified implementation models against reference data from comprehensive models or experimental measurements. This preliminary action ensures that even though the models are simplified for computational efficiency, they maintain sufficient accuracy for regression testing by pre-adjusting parameters and validating performance before deployment.
Data Source
AI summary
A computer-implemented method of deriving a vehicle dynamic library including at least two optimized vehicle dynamic implementation models is provided. The method comprises defining at least two operational design domains, choosing one vehicle dynamic abstract model, choosing one vehicle dynamic implementation model to be optimized, simulating the vehicle dynamic abstract model for each operational design domain, wherein the simulation generates reference trajectories, optimizing the vehicle dynamic implementation model for each operational design domain so as to create at least two optimized vehicle dynamic implementation models, wherein each optimization is based on at least one of said generated reference trajectories, validating each optimized vehicle dynamic implementation model to achieve a validation level for each optimized vehicle dynamic implementation model, and saving each optimized vehicle dynamic model together with its validation level.