Abstract:Effective real-time autonomous navigation for mobile robots in static and dynamic environments has become a challenging and active research topic. Although the simultaneous localization and mapping (SLAM) algorithm offers a solution, it often heavily relies on complex global and local maps, resulting in significant computational demands, slower convergence rates, and prolonged training times. In response to these challenges, this paper presents a novel algorithm called PER-n2D3QN, which integrates prioritized experience replay, a noisy network with factorized Gaussian noise, n-step learning, and a dueling structure into a double deep Q-network. This combination enhances the efficiency of experience replay, facilitates exploration, and provides more accurate Q-value estimates, thereby significantly improving the performance of autonomous navigation for mobile robots. To further bolster the stability and robustness, meaningful improvements, such as target "soft" updates and the gradient clipping mechanism, are employed. Additionally, a novel and powerful target-oriented reshaping reward function is designed to expedite learning. The proposed model is validated through extensive experiments using the robot operating system (ROS) and Gazebo simulation environment. Furthermore, to more specifically reflect the complexity of the simulation environment, this paper presents a quantitative analysis of the simulation environment. The experimental results demonstrate that PER-n2D3QN exhibits heightened accuracy, accelerated convergence rates, and enhanced robustness in both static and dynamic scenarios.

Skill Q-Network: Learning Adaptive Skill Ensemble for Mapless Navigation in Unknown Environments

Unsupervised Reinforcement Learning of Transferable Meta-Skills for Embodied Navigation

Learning Reward Function with Matching Network for Mapless Navigation

Learning Autonomous Navigation in Unmapped and Unknown Environments

Spiking Reinforcement Learning with Memory Ability for Mapless Navigation

Mapless mobile robot navigation at the edge using self-supervised cognitive map learners

Interactive Semantic Map Representation for Skill-based Visual Object Navigation

Mapless Navigation With Safety-Enhanced Imitation Learning

End-to-End Navigation in Unknown Environments using Neural Networks

SocNavGym: A Reinforcement Learning Gym for Social Navigation

DYNAMIC ROUTING FOR NAVIGATION IN CHANGING UNKNOWN MAPS USING DEEP REINFORCEMENT LEARNING

QuasiNav: Asymmetric Cost-Aware Navigation Planning with Constrained Quasimetric Reinforcement Learning

Adaptive and Explainable Deployment of Navigation Skills via Hierarchical Deep Reinforcement Learning

Learning to Navigate Through Complex Dynamic Environment with Modular Deep Reinforcement Learning

Sim-to-Real: Mapless Navigation for USVs Using Deep Reinforcement Learning

Using Deep Reinforcement Learning with Automatic Curriculum Learning for Mapless Navigation in Intralogistics

Mobile Robot Navigation Based on Noisy N-Step Dueling Double Deep Q-Network and Prioritized Experience Replay

Robot path planning using deep reinforcement learning

Learning Exploration Policies for Navigation

Self-Learning Navigation Strategy of Mobile Robot