A combined logic- and model-based approach to fault detection and identification （FDI） in a suction foot control system of a wall-climbing robot is presented in this paper. For the control system, some fault models ...A combined logic- and model-based approach to fault detection and identification （FDI） in a suction foot control system of a wall-climbing robot is presented in this paper. For the control system, some fault models are derived by kinematics analysis. Moreover, the logic relations of the system states are known in advance. First, a fault tree is used to analyze the system by evaluating the basic events （elementary causes）, which can lead to a root event （a particular fault）. Then, a multiple-model adaptive estimation algorithm is used to detect and identify the model-known faults. Finally, based on the system states of the robot and the results of the estimation, the model-unknown faults are also identified using logical reasoning. Experiments show that the proposed approach based on the combination of logical reasoning and model estimating is efficient in the FDI of the robot.展开更多
文摘为了解决果园因农药过量使用导致的环境污染与农药浪费问题,提出了一种基于改进YOLACT的果树叶墙区域(Leaf wall area,LWA)实时检测方法,用于计算深度彩色双目相机采集视频中的叶墙区域距离及密度,为果园农药智慧喷施作业中农药喷洒剂量与喷洒距离的实时调整提供依据。首先,使用ConvNeXt主干网络改进了YOLACT模型,并引入NAM通道注意力机制对模型进行了优化;其次,提出了基于深度学习的果树叶墙密度检测方法;最后,通过阈值法排除深度图像中的干扰信息,简化了果树叶墙平均距离计算方法的处理流程。实验结果表明,改进YOLACT模型分割的APall为91.6%,相较于原始模型上升3.0个百分点,与YOLACT++、Mask R CNN和QueryInst模型相比分别高2.9、1.2、4.1个百分点;叶墙密度估计算法在叶墙顶部、中部和底部的均方根误差(Root mean square error,RMSE)分别为1.49%、0.82%、2.20%;叶墙区域实时检测方法的处理速度可达29.96 f/s。
基金supported by the Hi-tech Research and Development Program of China (No.2006AA420203)
文摘A combined logic- and model-based approach to fault detection and identification （FDI） in a suction foot control system of a wall-climbing robot is presented in this paper. For the control system, some fault models are derived by kinematics analysis. Moreover, the logic relations of the system states are known in advance. First, a fault tree is used to analyze the system by evaluating the basic events （elementary causes）, which can lead to a root event （a particular fault）. Then, a multiple-model adaptive estimation algorithm is used to detect and identify the model-known faults. Finally, based on the system states of the robot and the results of the estimation, the model-unknown faults are also identified using logical reasoning. Experiments show that the proposed approach based on the combination of logical reasoning and model estimating is efficient in the FDI of the robot.
基金The National Engineering Laboratory for Wheat&Corn Further Processing(NL2016012)The Innovation Scientists and Technicians Troop Construction Projects of Henan Province(114100510015)The Nature Science Foundation of Education Department of Henan Province(16A413003)
文摘采用CFD(Computational Fluid Dynamic)技术进行湍流效应数值模拟时,经常需要计算流场网格点到最近壁面的距离。当网格规模很大,特别是针对一些动网格问题,壁面距离的计算量很大且费时,影响整个流场求解效率。本文对壁面距离计算的直接算法和基于二叉树技术的ADT(Alternating Digital Tree)快速算法进行了对比分析和研究,提出了一种高效的、基于方盒切割技术的快速计算方法,并采用不同外形的CFD计算网格对该方法进行了验证。计算结果表明,新方法的壁面距离计算效率明显高于直接算法和ADT算法,并具有较好的鲁棒性和通用性,可以便捷地移植到现有的CFD计算程序中,从而提高整个流场的数值计算效率。
