基于图像识别的深海RMR液位测控系统设计研究

Design and research of deep-sea RMR liquid level identification and control system based on image recognition

摘要:

无隔水管泥浆循环钻井（RMR）是一种新兴的绿色环保钻井技术，旨在解决没有泥浆循环通道的开放水域钻井中泥浆排放带来的材料损耗和环境污染等问题。创新的泥浆循环液位测控系统解决了开放水域钻井浑浊视界下泥浆-海水交融区域界面识别难题。泥浆循环液位测控系统由泥浆液位测算系统与泥浆液位控制系统组成，首先将收集的井口泥水图像传输至泥浆液位测算系统得到测算泥浆液位，然后基于测算泥浆液位与预设泥浆液位的偏差，利用泥浆液位控制系统自动调节泵量，使井口液位维持在预设水平。通过搭建陆地测试试验台进行试验，验证了RMR液位测控系统能够较为准确地识别井口液位并且自动控制泵的流量，能够自动应对井涌与井漏等状况，整体性能可靠，实现了深海钻探技术装备的原始创新。

Abstract:

Riserless mud recovery drilling （RMR） is an emerging green and environmentally friendly drilling technology aimed at solving the problems of material loss and environmental pollution caused by mud discharge in open water drilling without mud recovery channels. The innovative mud recovery liquid level identification and control system solves the problem of identifying the interface between mud and seawater in the turbid field of open water drilling. The system consists of the mud level identification system and mud level control system. Firstly， the collected images of mud and seawater are transmitted to the mud level identification system to obtain the measured mud level. Then， based on the deviation between the measured mud level and the preset mud level， the mud level control system is used to automatically adjust the pump volume to maintain the wellhead mud level at the designed level. By building the land testing platform for experimentation， it has been verified that the RMR liquid level identification and control system can accurately identify the mud level and automatically control the pump flow rate， and can automatically respond to well kick and leakage. The overall performance of the system is reliable， achieving the original innovation of deep-sea drilling technology equipment.