A platform for research: civil engineering, architecture and urbanism
A platform for research: civil engineering, architecture and urbanism
Full-rotation pile driving barge control method and system based on deep reinforcement learning
The invention discloses a full-rotation pile driving barge control method and system based on deep reinforcement learning, and the method comprises the steps: constructing a pile driving strategy model, and carrying out the pile driving simulation training of the pile driving strategy model; a sensor is installed on the pile driving barge, a pile driving instruction is responded, the pile driving instruction comprises a target pile body and the target state of the target pile body, and the pile driving barge is controlled to go to the target pile body; and the sensor is used for obtaining state parameters of the pile driving barge, the state parameters are input into the pile driving strategy model, the pile driving strategy model outputs a pile driving action instruction, the pile driving barge operates the target pile body according to the pile driving action instruction till the pile driving instruction is completed, and pile driving is ended. According to the method, the strategy model is constructed by using the DQN, the strategy model is trained by using the experience playback and exploration-learning mechanism, and the real-time state parameters are input into the strategy model to obtain the piling action instruction, so that the piling machine operation is automatically performed in real time according to the sea condition, the ship condition and the like, the human intervention is reduced, and the piling precision and efficiency are improved.
本发明公开了一种基于深度强化学习的全回转打桩船控制方法及系统，包括：构建打桩策略模型，对打桩策略模型进行打桩仿真训练；在打桩船上安装传感器，响应于打桩指令，所述打桩指令包括目标桩体和目标桩体的目标状态，控制打桩船前往目标桩体；利用传感器获取打桩船状态参数，向打桩策略模型输入状态参数，打桩策略模型输出打桩动作指令，打桩船根据打桩动作指令对目标桩体进行操作，直至完成打桩指令，结束打桩。本发明利用DQN构建策略模型，利用经验回放和探索‑学习机制对策略模型进行训练，将实时状态参数输入策略模型以获取打桩动作指令，从而实时根据海况、船只状况等自动进行桩机操作，减少人为干预，提升打桩精度及效率。
Full-rotation pile driving barge control method and system based on deep reinforcement learning
The invention discloses a full-rotation pile driving barge control method and system based on deep reinforcement learning, and the method comprises the steps: constructing a pile driving strategy model, and carrying out the pile driving simulation training of the pile driving strategy model; a sensor is installed on the pile driving barge, a pile driving instruction is responded, the pile driving instruction comprises a target pile body and the target state of the target pile body, and the pile driving barge is controlled to go to the target pile body; and the sensor is used for obtaining state parameters of the pile driving barge, the state parameters are input into the pile driving strategy model, the pile driving strategy model outputs a pile driving action instruction, the pile driving barge operates the target pile body according to the pile driving action instruction till the pile driving instruction is completed, and pile driving is ended. According to the method, the strategy model is constructed by using the DQN, the strategy model is trained by using the experience playback and exploration-learning mechanism, and the real-time state parameters are input into the strategy model to obtain the piling action instruction, so that the piling machine operation is automatically performed in real time according to the sea condition, the ship condition and the like, the human intervention is reduced, and the piling precision and efficiency are improved.
本发明公开了一种基于深度强化学习的全回转打桩船控制方法及系统，包括：构建打桩策略模型，对打桩策略模型进行打桩仿真训练；在打桩船上安装传感器，响应于打桩指令，所述打桩指令包括目标桩体和目标桩体的目标状态，控制打桩船前往目标桩体；利用传感器获取打桩船状态参数，向打桩策略模型输入状态参数，打桩策略模型输出打桩动作指令，打桩船根据打桩动作指令对目标桩体进行操作，直至完成打桩指令，结束打桩。本发明利用DQN构建策略模型，利用经验回放和探索‑学习机制对策略模型进行训练，将实时状态参数输入策略模型以获取打桩动作指令，从而实时根据海况、船只状况等自动进行桩机操作，减少人为干预，提升打桩精度及效率。
Full-rotation pile driving barge control method and system based on deep reinforcement learning
一种基于深度强化学习的全回转打桩船控制方法及系统
WANG XUEGANG (author) / YING ZONGQUAN (author) / SHEN WENGENG (author) / LIN MEIHONG (author) / DONG HONGJING (author)
2025-01-10
Patent
Electronic Resource
Chinese
IPC:
G06F
ELECTRIC DIGITAL DATA PROCESSING
,
Elektrische digitale Datenverarbeitung
/
E02D
FOUNDATIONS
,
Gründungen
/
G01D
MEASURING NOT SPECIALLY ADAPTED FOR A SPECIFIC VARIABLE
,
Anzeigen oder Aufzeichnen in Verbindung mit Messen allgemein
/
G06N
COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS
,
Rechnersysteme, basierend auf spezifischen Rechenmodellen
| European Patent Office | 2023