论文部分内容阅读
通过激光标志物的卷积神经网络(CNN)检测,与标志物中心点的奇异值分解(SVD)重构,实现了掘进机在巷道坐标系下的坐标估计。通过基于支撑向量数据描述(SVDD)的陀螺仪静止状态抖动抑制,与参考系变换,实现了机身与掘进臂的姿态检测。通过基于OpenGL的图形学引擎,实现了工作面场景的实时虚拟渲染。测试结果表明:系统能够准确可靠地完成工作面场景下掘进机监测任务。
Conjugation neural network (CNN) detection of laser markers and reconstruction of singular value decomposition (SVD) with the center of the marker were used to realize the coordinate estimation of roadheader in roadway coordinate system. Based on SVDD, the jitter suppression of the gyroscope and the transformation of the reference frame are used to detect the attitude of fuselage and driving arm. Through the OpenGL-based graphics engine, real-time virtual rendering of the work surface scene is realized. The test results show that the system can accurately and reliably complete the boring machine monitoring task under the working face.