Hierarchical Task Network(HTN)planning is a powerful technique in artificial intelligence for handling complex problems by decomposing them into hierarchical task structures.However,achieving optimal solutions in HTN ...Hierarchical Task Network(HTN)planning is a powerful technique in artificial intelligence for handling complex problems by decomposing them into hierarchical task structures.However,achieving optimal solutions in HTN planning remains a challenge,especially in scenarios where traditional search algorithms struggle to navigate the vast solution space efficiently.This research proposes a novel technique to enhance HTN planning by integrating the Ant Colony Optimization(ACO)algorithm into the refinement process.The Ant System algorithm,inspired by the foraging behavior of ants,is well-suited for addressing optimization problems by efficiently exploring solution spaces.By incorporating ACO into the refinement phase of HTN planning,the authors aim to leverage its adaptive nature and decentralized decision-making to improve plan generation.This paper involves the development of a hybrid strategy called ACO-HTN,which combines HTN planning with ACO-based plan selection.This technique enables the system to adaptively refine plans by guiding the search towards optimal solutions.To evaluate the effectiveness of the proposed technique,this paper conducts empirical experiments on various domains and benchmark datasets.Our results demonstrate that the ACO-HTN strategy enhances the efficiency and effectiveness of HTN planning,outperforming traditional methods in terms of solution quality and computational performance.展开更多
Action model learning has become a hot topic in knowledge engineering for automated planning.A key problem for learning action models is to analyze state changes before and after action executions from observed"p...Action model learning has become a hot topic in knowledge engineering for automated planning.A key problem for learning action models is to analyze state changes before and after action executions from observed"plan traces".To support such an analysis,a new approach is proposed to partition propositions of plan traces into states.First,vector representations of propositions and actions are obtained by training a neural network called Skip-Gram borrowed from the area of natural language processing(NLP).Then,a type of semantic distance among propositions and actions is defined based on their similarity measures in the vector space.Finally,k-means and k-nearest neighbor(kNN)algorithms are exploited to map propositions to states.This approach is called state partition by word vector(SPWV),which is implemented on top of a recent action model learning framework by Rao et al.Experimental results on the benchmark domains show that SPWV leads to a lower error rate of the learnt action model,compared to the probability based approach for state partition that was developed by Rao et al.展开更多
基金supported by the Ministry of Science and High Education of the Russian Federation by the grant 075-15-2022-1137supported by Princess Nourah bint Abdulrahman University Researchers Supporting Project number(PNURSP2025R323),Princess Nourah bint Abdulrahman University,Riyadh,Saudi Arabia.
文摘Hierarchical Task Network(HTN)planning is a powerful technique in artificial intelligence for handling complex problems by decomposing them into hierarchical task structures.However,achieving optimal solutions in HTN planning remains a challenge,especially in scenarios where traditional search algorithms struggle to navigate the vast solution space efficiently.This research proposes a novel technique to enhance HTN planning by integrating the Ant Colony Optimization(ACO)algorithm into the refinement process.The Ant System algorithm,inspired by the foraging behavior of ants,is well-suited for addressing optimization problems by efficiently exploring solution spaces.By incorporating ACO into the refinement phase of HTN planning,the authors aim to leverage its adaptive nature and decentralized decision-making to improve plan generation.This paper involves the development of a hybrid strategy called ACO-HTN,which combines HTN planning with ACO-based plan selection.This technique enables the system to adaptively refine plans by guiding the search towards optimal solutions.To evaluate the effectiveness of the proposed technique,this paper conducts empirical experiments on various domains and benchmark datasets.Our results demonstrate that the ACO-HTN strategy enhances the efficiency and effectiveness of HTN planning,outperforming traditional methods in terms of solution quality and computational performance.
基金Supported by the National Natural Science Foundation of China(61103136,61370156,61503074)Open Research Foundation of Science and Technology on Aerospace Flight Dynamics Laboratory(2014afdl002)
文摘Action model learning has become a hot topic in knowledge engineering for automated planning.A key problem for learning action models is to analyze state changes before and after action executions from observed"plan traces".To support such an analysis,a new approach is proposed to partition propositions of plan traces into states.First,vector representations of propositions and actions are obtained by training a neural network called Skip-Gram borrowed from the area of natural language processing(NLP).Then,a type of semantic distance among propositions and actions is defined based on their similarity measures in the vector space.Finally,k-means and k-nearest neighbor(kNN)algorithms are exploited to map propositions to states.This approach is called state partition by word vector(SPWV),which is implemented on top of a recent action model learning framework by Rao et al.Experimental results on the benchmark domains show that SPWV leads to a lower error rate of the learnt action model,compared to the probability based approach for state partition that was developed by Rao et al.
