Abstract:This paper combines a deep Q network, sample priority, long short‐term memory, and pre‐processing mechanisms to achieve the design of a complete coverage path planning algorithm for deep‐sea mining vehicle clusters. A map‐ fusion mechanism and a state matrix preprocessing method are proposed. Constraints based on distance variance are also designed to avoid the problem of hose entanglement. This paper proposes a deep reinforcement learning algorithm to achieve complete coverage path planning for deep‐sea mining vehicle clusters. First, the mining vehicles and the deep‐sea mining environment are modeled. Then, this paper implements a series of algorithm designs and optimizations based on Deep Q Networks (DQN). The map fusion mechanism can integrate the grid matrix data from multiple mining vehicles to get the state matrix of the complete environment. In this paper, a preprocessing method for the state matrix is also designed to provide suitable training data for the neural network. The reward function and action selection mechanism of the algorithm are also optimized according to the requirements of cluster cooperative operation. Furthermore, the algorithm uses distance constraints to prevent the entanglement of underwater hoses. To improve the training efficiency of the neural network, the algorithm filters and extracts training samples for training through the sample quality score. Considering the requirement of cluster complete coverage mission, this paper introduces Long Short‐Term Memory (LSTM) based on the neural network to achieve a better training effect. After completing the above optimization and design, the algorithm proposed in this paper is verified through simulation experiments.

A Multi-Objective Deep Reinforcement Learning Method for Path Planning in Shovel Loading Scenario

Spacecraft Attitude Maneuver Planning Based on Deep Reinforcement Learning under Complex Constraints

Path Planning Technology of Unmanned Vehicle Based on Improved Deep Reinforcement Learning

Deep Reinforcement Learning-based Multi-objective Path Planning on the Off-road Terrain Environment for Ground Vehicles

The Wide-Area Coverage Path Planning Strategy for Deep-Sea Mining Vehicle Cluster Based on Deep Reinforcement Learning

Deep Reinforcement Learning-Based 2.5D Multi-Objective Path Planning for Ground Vehicles: Considering Distance and Energy Consumption

An Algorithm of Complete Coverage Path Planning for Deep‐Sea Mining Vehicle Clusters Based on Reinforcement Learning

Deep Reinforcement Learning Approach with Multiple Experience Pools for UAV's Autonomous Motion Planning in Complex Unknown Environments

Path Planning for Autonomous Vehicles in Unknown Dynamic Environment Based on Deep Reinforcement Learning

Multi-objective Path Planning Based on Deep Reinforcement Learning

Multi-AGV Path Planning Method via Reinforcement Learning and Particle Filters

Unmanned Aerial Vehicle Path Planning Algorithm Based on Deep Reinforcement Learning in Large-Scale and Dynamic Environments

Path Planning of Autonomous Mobile Robot in Comprehensive Unknown Environment Using Deep Reinforcement Learning

3D Environment-based Multiobjective Path Planning for Unmanned Ground Vehicles.

Path Planning Algorithm for Unmanned Surface Vessel Based on Multiobjective Reinforcement Learning

Deep Reinforcement Learning Based Path Planning for Mobile Robots Using Time-Sensitive Reward

Variable-Step-Length Hybrid A* Based on Dichotomy Optimization for Path Planning of Autonomous Mining Trucks*

Two-Stage Dichotomy Hybrid A* for Path Planning of Unmanned Mining Trucks in Large-Scale Open-Pit Mine

Improved Robot Path Planning Method Based on Deep Reinforcement Learning

Deep Reinforcement Learning-based Multi-AMR Path Planning Algorithm

Pursuit Path Planning for Multiple Unmanned Ground Vehicles Based on Deep Reinforcement Learning