煤炭技术

针对地下矿山深部开采环境复杂、安全风险高等问题，构建了集环境感知、路径规划、远程通信、智能决策于一体的无人驾驶采矿装备远程智能控制系统。系统采用“云—边—端”3层架构，通过5G专网与Wi-Fi 6融合实现井下全覆盖通信；集成激光雷达、双目视觉、毫米波雷达构建360°环境感知体系；基于UWB与IMU组合定位实现±15 cm动态精度。采用改进A*算法和动态窗口法实现路径规划与避障，通过D-S证据理论融合多源信息，设计基于有限状态机的自主决策逻辑，远程控制响应时间45 ms。

To address the challenges of complex environments and high safety risks in deep underground mining, an intelligent remote control system for unmanned mining equipment has been developed, integrating environmental perception, path planning, remote communication, and smart decision-making. The system adopts a "cloud-edge-device" three-tier architecture, achieving full underground coverage through 5G dedicated networks and Wi-Fi 6 integration. It establishes a 360°environmental perception system by combining LiDAR, binocular vision, and millimeter-wave radar,with dynamic positioning accuracy of ±15 cm achieved via UWB and IMU integration. Path planning and obstacle avoidance utilize an improved A* algorithm and dynamic window method, while D-S evidence theory merges multi-source information to design finite state machine-based autonomous decision logic, achieving 45 ms remote control response time.