Symmetry Pruning for AI Planning Speed
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current AI planning approaches are time and resource intensive when dealing with complex planning problems, particularly PSPACE-hard problems, and lack efficient solutions for unordered top-quality planning, leading to increased CPU time and resource utilization.
Innovation Solution
The method involves identifying symmetries in planning problems to create an orbit search space, performing a two-phase search iteratively over this space to find surrogate plans, and extending these plans using symmetries to generate a set of solutions, thereby reducing CPU time and improving processing speed without increasing resources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional AI planning approaches are used to solve complex planning problems, then complete and accurate solutions can be found, but CPU time and resource utilization increase significantly
Solution Approach 1:
The patent applies asymmetry by breaking the symmetry in the planning problem through canonical representation. By identifying symmetries and selecting a canonical form (using lexicographic ordering of state representations), the algorithm transforms symmetric equivalent states into a single representative state, thereby reducing the search space while maintaining solution completeness.
Solution Approach 2:
The patent segments the planning problem by dividing the state space into equivalence classes based on symmetry relations. Each equivalence class is represented by a canonical state, effectively partitioning the original state space into manageable segments that can be searched independently, reducing overall computational complexity.
2Productivity
If the search space is reduced using symmetry-based pruning, then CPU time decreases, but the complexity of identifying and managing symmetries increases
Solution Approach 1:
The system performs self-service by automatically identifying symmetries and generating canonical representations without external intervention. The symmetry identification and canonicalization processes are integrated into the planning algorithm itself, allowing the system to autonomously reduce its search space while managing the complexity of symmetry handling.
Solution Approach 2:
The patent applies preliminary action by pre-identifying symmetries and establishing canonical representations before the main search process begins. This preliminary setup of symmetry information and canonical forms enables efficient pruning during the search phase, trading initial computational effort for significant speedups during plan generation.
Data Source
AI summary
Computer implemented methods, systems, and computer program products include program code executing on a processor(s) that obtains a planning problem. The program code obtains a bound on a number of plans (to address the planning problem). The program code identifies symmetries of the planning problem. The program code utilizes the symmetries to identify an orbit search space of the planning problem. The program code executes a two-phase search iteratively over the orbit space to identify surrogate plans in the orbit space. The program code generates new plans by utilizing the surrogate plans and the symmetries of the planning problem to map the surrogate plans to new plans. The program code extends the new plans. The extended new plans comprise the set of solutions for the planning problem.