Adaptive autonomous robot path planning using hybrid probabilistic roadmaps and bidirectional a star search

In modern robotic applications, including search and rescue, manufacturing, transportation, disaster response, and defense, sophisticated autonomous path-planning and navigation systems have become increasingly important. Notwithstanding significant progress, contemporary path-planning methodologies sometimes encounter challenges related to computational complexity, scalability, and real-time adaptability, particularly in environments rife with barriers and dynamic alterations. To address these limitations, this research proposes a dependable bidirectional hybrid path-planning framework that integrates the Probabilistic Roadmap (PRM) methodology with an Enhanced $$A^{*}$$ algorithm for efficient autonomous robot path-planning. The proposed architecture leverages the global connection and search-space reduction capabilities of PRM as well as the heuristic search efficiency of bidirectional $$A^{*}$$ . The enhanced bidirectional $$A^{*}$$ algorithm includes several improvements, including efficient roadmap-guided graph exploration, priority-based node expansion, early search termination through frontier intersection, closed-set pruning, dynamic cost updating, and simultaneous forward and backward searches. By restricting the search to collision-free roadmap nodes and performing simultaneous exploration from both the start and destination locations, the proposed method significantly reduces search depth, computational complexity, and path-planning time. Adaptive replanning and collision-avoidance algorithms are also used to extend the system to multi-robot scenarios and enable path-planning in static, dynamic, and hybrid static-dynamic settings. The effectiveness of the proposed system is evaluated using comprehensive performance measures, including path length, computing time, path smoothness, success rate, and adaption factor. Extensive MATLAB/Simulink simulations are performed in a variety of path-planning scenarios and contrasted with conventional path-planning methods, including Dijkstra-based and conventional $$A^{*}$$ -PRM approaches. The gathered results demonstrate that the proposed hybrid PRM-bidirectional $$A^{*}$$ framework consistently achieves shorter path lengths, faster convergence, increased computation efficiency, and higher adaptation to environmental changes. These findings validate its suitability as a scalable and dependable method for autonomous robot path-planning in complex and dynamic scenarios.