Behavior Planner Validation Using Declarative Traffic Situation Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing behavior planners for automated vehicles struggle to formally address the complex requirements and restrictions imposed by various elements of an operational design domain (ODD), particularly in complex and dynamic scenarios, due to limitations in scalability and comprehensive validation.
Innovation Solution
The method involves representing evaluation models of partial situations in a declarative program format, allowing for the determination of formally explorable boundary conditions for vehicle behavior. This is achieved by combining declarative program representations of selected partial situations into a combined program representation for predefined test situations, using tools from declarative problem-solving such as answer set programming.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If the SOCA method is used to analyze individual traffic situations with zone graphs, then the behavior of the ego vehicle can be determined for specific situations, but the approach is not sufficiently scalable for complex operational design domains with multiple repeating elements
Solution Approach 1:
The patent segments complex traffic situations into reusable partial situations that can be independently defined and combined. Instead of analyzing entire complex situations as monolithic units, the system breaks them down into smaller, manageable partial situations that can be stored in a database and recombined through template-based approaches.
Solution Approach 2:
The patent creates universal templates for partial situations that can be applied repeatedly across different contexts within the operational design domain. These templates define reusable structures and parameters that can instantiate multiple specific situations, enabling the system to handle complex ODDs with repeating elements efficiently.
2Reliability
If all operating conditions and deployment situations are broken down into partial situations in the database, then comprehensive coverage of the ODD is achieved, but the database requires corresponding variety and tailoring of partial situations
Solution Approach 1:
The patent performs preliminary action by pre-defining and storing partial situations in the database before actual validation needs arise. These partial situations are prepared in advance with their associated evaluation models, allowing the system to quickly assemble comprehensive situation coverage without ad-hoc analysis during validation.
Solution Approach 2:
The patent uses parameter changes to create variety in the database efficiently. By defining templates with parameterized characteristics, the system can generate multiple specific partial situations from a single template by varying parameters, thus achieving comprehensive coverage without manually creating every possible situation.
3Measurement precision
If evaluation models are tailored to take into account as many influencing factors as possible, then boundary conditions for permissible behavior are specified comprehensively, but the definition and maintenance of partial situations becomes more complex
Solution Approach 1:
The patent segments the evaluation process into modular components associated with each partial situation. Each evaluation model focuses on specific influencing factors relevant to its partial situation, rather than attempting to handle all factors in a monolithic model. This segmentation makes both the models and their definition more manageable.
Solution Approach 2:
The patent applies partial action by creating evaluation models that focus on the specific influencing factors relevant to each partial situation, rather than requiring comprehensive coverage of all possible factors in every model. This allows for precise boundary conditions where needed while maintaining ease of definition through selective focus.
Data Source
AI summary
A computer-implemented method for validating a behavior planner for an automated vehicle. The behavior planner has a database having previously defined partial situations and an evaluation model for each partial situation to break down a given situation into partial situations of the database and, based on the associated evaluation models, to determine boundary conditions for permissible behavior options of the vehicle in the given situation, as a combination of boundary conditions of the individual partial situations. The evaluation models of the partial situations are made available in a declarative program representation which enables the determination of formally explorable boundary conditions for permissible behavior options of the vehicle in the particular partial situation. A test situation is predefined as a composition of selected partial situations. The declarative program representations of the evaluation models of the selected partial situations are combined to form a combined program representation of the evaluation models.