摘要
掘进机定位定姿技术作为掘进机自动化和无人化的基础,亟需深入研究。当前方法以SINS定位传感器信号为基础,完成定位定姿。但是,矿井下的传感信号极易受到现场环境干扰,导致单一的定位方法无法精准获取掘进机传感信号,会直接影响后续的坐标精度,导致纯捷联惯导系统中存在累积误差。提出基于视觉坐标误差校准的煤矿井下掘进机自动化定位定姿方法。通过采集掘进机工作位姿图像,利用图像处理和计算机视觉技术,针对纯捷联惯导系统中的累积误差问题,建立了坐标系之间的转换关系,,并确定掘进机自身坐标系。根据初始位置坐标和相关坐标系定义,结合四元数方法构建一个自动化的定姿定位模型;通过建立的模型,获取掘进机工作状态的位姿角和位置更新方程,用于实时计算和更新掘进机位置和位姿,实现煤矿井下掘进机自动化定位定姿。实验结果表明:利用视觉定位开展煤矿井下掘进机自动化定位定姿时,捷联惯导系统的累积误差在俯仰角上得到有效抑制,x和y方向上的平均定位误差为12.08 cm和19.77 cm,比单纯SINS定位效果更好。
The positioning and pose determination technology of tunneling machines,as the foundation of automation and unmanned tunneling machines,urgently needs in-depth research.The current method is based on SINS positioning sensor signals to complete positioning and pose determination.However,the sensing signals under the mine are highly susceptible to on-site environmental interference,resulting in a single positioning method not being able to accurately obtain the sensing signals of the tunneling machine,which directly affects the subsequent coordinate accuracy and leads to accumulated errors in the pure strapdown inertial navigation system.Propose an automated positioning and pose determination method for coal mine underground tunneling machines based on visual coordinate error calibration.By collecting the working posture images of the tunneling machine,utilizing image processing and computer vision technology,a transformation relationship between coordinate systems was established to address the cumulative error problem in the pure strapdown inertial navigation system,and the tunneling machine's own coordinate system was determined.Based on the definition of initial position coordinates and related coordinate systems,an automated pose determination and positioning model is constructed using quaternion methods;By establishing a model,the pose angle and position update equation of the working state of the tunneling machine are obtained,which are used to calculate and update the position and pose of the tunneling machine in real time,achieving automatic positioning and pose determination of the underground tunneling machine in coal mines.The experimental results show that when using visual positioning for automated positioning and pose determination of coal mine underground tunneling machines,the accumulated error of the strapdown inertial navigation system is effectively suppressed in the pitch angle,and the average positioning errors in the x and y directions are 12.08cm and 19.77cm,which is better than the position
作者
李福来
LI Fulai(China Shenhua Energy Co.,Ltd.,Shendong coal branch,Shenmu,Shaanxi 719300,China)
出处
《自动化与仪器仪表》
2024年第10期347-351,共5页
Automation & Instrumentation
基金
中国神华能源股份有限公司神东煤炭分公司资助项目(SHGF-06-08)。
关键词
掘进机
视觉矫正
图像处理
坐标转换
定位定姿
SINS定位
tunneling machine
visual correction
image processing
coordinate conversion
positioning and pose determination
SINS positioning
