从农田垄线等间距地形特征着手,基于粒子滤波提出了一种基于低成本定位器的高精度跟踪算法.对农机工作路线进行优化设计的基础上,确定了适合农田作业的2种运动模型,并采用交互式多模型(IMM)粒子滤波算法对田间运动目标进行了定位滤波处理;然后以粒子滤波算法为基础,将农田地形特征作为一个限制因素融入到算法中,在滤波过程中对穿越垄线的粒子进行剔除,从而提高估计精度;最后基于改进的IMM粒子滤波算法进行了定位跟踪仿真和分析.结果表明:该算法借助普通低成本GPS应用于垄线规则的农田导航,可有效地提高定位跟踪精度.
According to the regular farmland topography,a new tracking algorithm with low cost and high precision was proposed.Based on the optimized driving routes of agricultural vehicle,two maneuver mo-dels were determined for farmland work,and the farmland work target was tracked by interacting multiple model(IMM) particle filter.Considered the farmland topography as a factor in particle filter algorithm,the particles which cross ridges were ticked out to improve the estimated precision.The improved IMM particle filter algorithm was used to simulate and analyze the moving target.The results show that the proposed algorithm can improve the positioning tracking precision effectively only by general low cost GPS for the navigation of regular farmland topography.