A-MHA*: Anytime Multi-Heuristic A*

2025-08-29

AI Summary

Existing Multi-Heuristic A* (MHA*) suffers from non-progressive solution improvement, reliance on manually tuned inflation factors, and dependence on domain-specific, globally admissible heuristics—often difficult to design. To address these limitations, this paper proposes Anytime Multi-Heuristic A* (AMHA*), the first algorithm to rigorously integrate the anytime repairing A* paradigm into the MHA* framework while preserving bounded suboptimality and completeness. AMHA* introduces a dynamic weight inflation mechanism and an incremental repair strategy that jointly leverage multiple *partially effective but inadmissible* heuristics. Empirical evaluation on 3D pathfinding and sliding-tile puzzles demonstrates that AMHA* significantly outperforms MHA* and other anytime algorithms: it delivers high-quality initial solutions faster and continuously refines solution quality as computation proceeds.

Technology Category

• Search and Optimization: Heuristic Search
• Planning, Routing, and Scheduling: Replanning and Plan Repair
• Multiagent Systems: Multiagent Planning

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

Designing good heuristic functions for graph search requires adequate domain knowledge. It is often easy to design heuristics that perform well and correlate with the underlying true cost-to-go values in certain parts of the search space but these may not be admissible throughout the domain thereby affecting the optimality guarantees of the search. Bounded suboptimal search using several such partially good but inadmissible heuristics was developed in Multi-Heuristic A* (MHA*). Although MHA* leverages multiple inadmissible heuristics to potentially generate a faster suboptimal solution, the original version does not improve the solution over time. It is a one shot algorithm that requires careful setting of inflation factors to obtain a desired one time solution. In this work, we tackle this issue by extending MHA* to an anytime version that finds a feasible suboptimal solution quickly and continually improves it until time runs out. Our work is inspired from the Anytime Repairing A* (ARA*) algorithm. We prove that our precise adaptation of ARA* concepts in the MHA* framework preserves the original suboptimal and completeness guarantees and enhances MHA* to perform in an anytime fashion. Furthermore, we report the performance of A-MHA* in 3-D path planning domain and sliding tiles puzzle and compare against MHA* and other anytime algorithms.

Problem

• Extends MHA* to an anytime algorithm for continual solution improvement
• Addresses lack of solution refinement over time in original MHA*
• Enables quick feasible solutions with progressive optimality enhancement

Innovation

• Anytime extension of Multi-Heuristic A* algorithm
• Combines multiple inadmissible heuristics for faster search
• Continuous solution improvement until time expiration