A Hierarchical DWD-PLRM Framework for Real-Time Autonomous Navigation in Dynamic Environments

Planning of autonomous mobile robots in unpredictable and dynamic environments in real time is a significant challenge. Conventional techniques, including A, DLite, RRT and Velocity Obstacle (VO) techniques, either assume the world is statically known or are reactive in nature. Global planners can be effective in constant conditions, but not able to preserve their approach in the shifting ones, whereas reactive local strategies such as Dynamic Window Approach (DWA) and Artificial Potential Fields (APF) are quicker but tend to create unstable or oscillating paths. Such computationally intense control methods as Model Predictive Control (MPC) also increase the load on onboard processing, restricting speed and safety. In order to overcome these drawbacks, a hierarchical hybrid path planning algorithm based on Dynamic Wavefront Decomposition (DWD), and Predictive Local Replanning Module (PLRM) is presented. DWD actively separates the environment, and calculates parallel costmaps to ensure continuity of world paths, whereas PLRM anticipates motion of obstacles, to produce smoother and adaptive local paths. Experimental findings indicate that DWD-PLRM framework decreases the time of replanning by 65.9 percent over <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$D^{*}$</tex>-Lite and 66 percent over A+ and still ensures latencies of less than <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\mathbf{1 0 0 ~ m s}$</tex>. It has a 92 % obstacle avoidance success rate which is more than 10 times better than the existing methods. The suggested solution provides the computational efficiency, predictive, and adaptability to realtime autonomous navigation and manages the global optimality of structured planners and the flexibility of reactive systems to guarantee reliable multi-robot operations.

Ҳали таржима қилинмаган