Robot Behavior Control via Trajectory Probability Overlap
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems face challenges in controlling the behavior of agents to prevent interaction or contact with moving objects, as the position of the interaction point is indeterminate, making it difficult to achieve desired behavior patterns.
Innovation Solution
A behavior control system that determines predicted object trajectories using probability density distributions, generates position trajectory candidates for counter objects to match interaction points, and adjusts behavior plans to ensure desired interaction patterns with moving objects, allowing for secure task execution while varying motion patterns.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a polynomial, Bezier or attractor technique is used to generate a reference state variable trajectory, then the agent can achieve stable autonomous motion, but the position of the interaction point becomes indeterminate when trying to prevent contact with a moving object
Solution Approach 1:
The system dynamically adjusts the behavior plan by selecting different motion patterns based on the predicted position probability distribution of the object. When contact prevention is required, the system selects motion patterns that maintain distance; when interaction is desired, it selects patterns that approach the object. This dynamic selection resolves the contradiction by making the interaction point determination adaptive rather than fixed.
Solution Approach 2:
The system changes the parameter of motion pattern selection based on the desired interaction level. By adjusting the selection criteria for motion patterns according to the probability distribution overlap with the object's predicted position, the system can transition between stable autonomous motion and precise interaction point control as needed.
2Reliability
If the agent follows a fixed behavior plan to ensure consistent interaction patterns, then task execution is reliable, but the agent cannot adapt to varying task requirements or environmental conditions
Solution Approach 1:
The behavior plan is made dynamic through the motion pattern selection mechanism. The system maintains reliability by consistently following selected motion patterns while achieving adaptability through the ability to reselect patterns based on updated probability distributions and task requirements. This resolves the contradiction between fixed execution and flexible adaptation.
Solution Approach 2:
The system performs self-adjustment by automatically selecting appropriate motion patterns based on the calculated probability distribution overlap. Rather than requiring external reprogramming for different tasks, the system serves itself by adapting its behavior plan based on the current situation and desired interaction level, maintaining both reliability and versatility.
3Adaptability or versatility
If multiple position trajectory candidates are generated for the counter object, then the system can select the most appropriate trajectory, but the computational complexity increases
Solution Approach 1:
Instead of generating all possible trajectory candidates, the system generates a sufficient number of candidates that cover the necessary range of motion patterns. By selecting from this partial set based on the probability distribution overlap criterion, the system achieves adequate adaptability without the excessive computational burden of exhaustive generation.
Data Source
AI summary
A behavior control system capable of controlling the behavior of an agent (robot) such that the agent securely applies a force to a moving object. The behavior control system calculates the degree of overlapping of a time-series probability density distribution between a predicted position trajectory of an object (ball) and a position trajectory candidate of a counter object (racket). Further, a behavior plan of the agent (robot) is generated such that the counter object is moved according to a desired position trajectory, which is a mean position trajectory or a central position trajectory of a position trajectory candidate of the counter object which has the highest degree of overlapping with the predicted position trajectory of the object among a plurality of position trajectory candidates of the counter object.