Robotic Control Strategy Selection in Dynamic Object Scenarios
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing motion planning methods for mobile agents in multi-agent and dynamic object environments struggle to effectively account for intelligent behaviors and interactions, particularly when dynamic objects anticipate and collaborate with the agent, making it difficult to select an appropriate control strategy that optimizes motion trajectories while avoiding collisions and fulfilling tasks.
Innovation Solution
A computer-implemented method that provides multiple scenarios for dynamic objects with behavior specifications, benchmarks control strategy candidates using model predictive control, and selects the best strategy based on key performance indicators, considering intelligent behaviors and interactions through a cost function that weighs collaborative, neutral, or sabotage actions of dynamic objects.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If multiple control strategy candidates are benchmarked against multiple scenarios with intelligent dynamic objects, then the reliability of motion planning is improved, but the computational time and complexity increase
Solution Approach 1:
The patent pre-generates multiple control strategy candidates and pre-defines multiple scenarios with dynamic objects before actual motion planning execution. This preliminary preparation allows the system to have ready-to-evaluate strategies and scenarios, reducing computational time during real-time operation while maintaining reliability through comprehensive benchmarking.
Solution Approach 2:
The patent divides the motion planning problem into separate control strategy candidates and evaluates them independently against multiple scenarios. Each strategy is benchmarked separately, allowing parallel evaluation and reducing overall computational complexity. The segmentation enables systematic comparison of different strategies without requiring simultaneous complex simulations.
2Adaptability or versatility
If the system accounts for intelligent behaviors and interactions of dynamic objects through optimization problems, then the adaptability of motion planning is improved, but the device complexity increases
Solution Approach 1:
The patent introduces an intermediary evaluation layer that assesses control strategies against predefined scenarios containing intelligent dynamic objects. This intermediary benchmarking framework mediates between the simple control strategies and the complex intelligent behaviors, allowing adaptability without directly implementing complex interaction models in the core motion planning system.
Solution Approach 2:
The patent creates simplified copies or representations of intelligent dynamic objects within predefined scenarios for benchmarking purposes. These copied scenarios capture essential intelligent behaviors and interactions without requiring the full complexity of real intelligent agents, enabling adaptability testing with reduced system complexity.
Data Source
Figure 1~2
AI summary
The present invention relates to a computer-implemented method for determining an appropriate control strategy for a robotic device (1) in an environment with at least one dynamic object (2, 3, 4), comprising the steps of: - Providing (S1) a number of different scenarios wherein to each of the scenarios the at least one dynamic object (2, 3, 4) is associated, wherein with each of the scenarios, each of the at least one dynamic object (2, 3, 4) is associated with a start (S), a goal (G) and an individual behavior specification; - Providing a number of control strategy candidates for the robotic device (1); - Evaluating (S3) each of the control strategy candidates in each of the provided scenarios; - Selecting (S4) the control strategy for the robotic device (1) depending on the result of the evaluating of the control strategy candidates.