Dual-Layer Subsumption Control for Sequential Behavior Planning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The traditional subsumption architecture in intelligent control systems lacks scalability and flexibility, making it difficult to execute specific sequences of behaviors, as all behaviors are emergent and unpredictable, which is undesirable in situations requiring planned actions.
Innovation Solution
A dual-layer subsumption control system is introduced, where sequential behavioral nodes generate plans and activity nodes execute these plans, allowing for both emergent and sequential behaviors, with priorities established at runtime to select and execute plans effectively.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a traditional subsumption architecture with parallel independent behaviors is used, then flexibility and adaptability to environmental changes are improved, but the ability to execute specific sequences of behaviors deteriorates
Solution Approach 1:
The control architecture is segmented into two distinct layers: a plan execution subsystem that handles sequential behaviors and an activity selection subsystem that manages emergent behaviors. This segmentation allows each layer to specialize in specific types of control, resolving the contradiction by enabling both sequential execution and environmental adaptability through structured division of functions
Solution Approach 2:
The invention introduces a hierarchical dimension to the control architecture, organizing behaviors across multiple levels (plan execution layer and activity selection layer) rather than a single flat layer. This dimensional organization enables the system to simultaneously maintain sequential behavior capabilities at the higher level and emergent behavior capabilities at the lower level
2Reliability
If fixed priority relationships between behaviors are established, then deterministic control is improved, but dynamic adaptability deteriorates
Solution Approach 1:
The plan execution subsystem dynamically adjusts the execution of sequential behaviors based on current system state and environmental feedback, while the activity selection subsystem dynamically selects among competing emergent behaviors. This dynamic operation at both layers resolves the contradiction by enabling deterministic sequential execution when needed while maintaining flexibility to adapt to changing conditions
Solution Approach 2:
The system changes the operational parameters of different behavior layers based on situational requirements. The plan execution subsystem can adjust its sequencing parameters dynamically, and the activity selection subsystem can adjust priority weights of emergent behaviors, allowing the system to shift between deterministic and adaptive control modes as needed
3Adaptability or versatility
If the number of behaviors in the system is increased, then system functionality is improved, but control cycle feasibility deteriorates
Solution Approach 1:
By segmenting the control architecture into two specialized subsystems, the invention enables the system to handle a larger number of behaviors without sacrificing control cycle efficiency. Each subsystem processes a specific type of behavior, avoiding the computational overhead of managing all behaviors in a single undifferentiated system
Solution Approach 2:
The activity selection subsystem selectively activates only the relevant emergent behaviors needed for the current situation, rather than evaluating all possible behaviors. This partial action approach maintains control cycle feasibility even as the total number of available behaviors increases, as only a subset is actively processed at any given time
Data Source
AI summary
A method for controlling operation of a system that interacts with an environment includes obtaining, by a plurality of sequential behavioral nodes, a plurality of plans for controlling the system, each plan comprising a sequence of activities to be performed by the system; selecting one of the respective plans for execution based on relative priorities of the sequential behavioral nodes; and executing the selected plan. Related subsumption-based controllers are also disclosed.