摘要
在AGV沿预设磁条运动的过程中,由于受到磁条周边磁性杂质产生的异常点的影响,会使传感器获取磁条位置信息时出现偏差,使得AGV运动的平稳性降低,严重时会出现脱轨运行。为了识别并消除异常点,设计以四阶相关峭度作为代价函数的反卷积逆滤波器,对磁导航混合信号进行盲提取,有效地识别出野点,恢复出源信号。实验结果说明,该算法能够有效地识别异常点,并且具有较快的收敛速度和较好地稳定性。
During the movement of the AGV along the preset magnetic stripe,due to the influence of abnormal points generated by the magnetic impurities around the magnetic stripe,deviations will occur when the sensor obtains the position information of the magnetic stripe,which reduces the stability of the AGV movement. There will be derailment operation. In order to identify and eliminate the abnormal points,a deconvolution inverse filter with the fourth-order correlation kurtosis as the cost function is designed to blindly extract the magnetic navigation mixed signal,identify and remove the abnormal points,and restore the source signal. The experimental results show that the algorithm can effectively identify abnormal points,and has a faster convergence speed and better stability.
作者
黄卓群
杨光永
程满
刘叶
HUANG Zhuoqun;YANG Guangyong;CHENG Man;LIU Ye(College of Electronics and Information,Yunnan University of Nationalities,Kunming 650500)
出处
《计算机与数字工程》
2022年第12期2803-2808,共6页
Computer & Digital Engineering
基金
国家自然科学基金项目(编号:61761049,61261022)
云南省专业学位研究生教学案例库及研究生优质课程项目资助。
关键词
野点识别
四阶相关峭度
逆滤波器
盲提取
outlier recognition
fourth-order correlation kurtosis
inverse filter
blind extraction
