摘要
[目的]农田环境的复杂性对履带式收割机的路径跟踪精度提出了严峻挑战。为了提升收割机在农田作业中的路径跟踪性能,减少跟踪偏差,本研究提出了一种基于改进鱼群算法的履带式收割机全田块路径跟踪方法。[方法]根据履带式收割机的结构特性,将其作业过程简化为二维平面上的运动形式。通过结合全局与局部坐标系的转换,构建了收割机运动轨迹的数学模型,并进一步建立了相邻时刻履带式收割机全田块作业的运动模型。随后,根据不同作业状态,以收割机的前视距离作为关键参数,确定增益系数,从而获取其实时控制变量。为优化路径跟踪效果,引入了粒子滤波算法对鱼群算法进行改进,并以此构建了目标函数。在目标求解过程中,通过算法的不断迭代和优化,实现了收割机路径的精准跟踪。[结果]经过多次试验验证,在设定不同起始偏差点的情况下,本文提出的方法表现出了良好的跟踪性能。应用本文方法后,作业路径跟踪平均响应时间为0.52 s,最小转弯半径为5.0 m,平均偏差0.8 m,最小偏差0.5 m,与设定路线基本一致。这一结果充分证明了本文设计方法的有效性。[结论]综上所述,本文提出的基于改进鱼群算法的履带式收割机全田块路径跟踪方法,能够准确实现收割机在复杂农田环境中的精准跟踪,跟踪效果好,具有广泛的应用价值。
[Objective]The complexity of farmland environment poses severe challenges to the path tracking accuracy of crawler harvesters.To improve the path tracking performance of harvesters in field operations and reduce tracking deviation,this study proposed a full-field path tracking method for crawler harvesters based on an improved fish swarm algorithm.[Methods]According to the structural characteristics of the crawler harvester,its operation process was simplified into a two-dimensional plane motion form.By combining global and local coordinate transformations,a mathematical model of the harvester's motion trajectory was constructed,and a motion model of the crawler harvester's full-field operation at adjacent times was further established.Then,based on different operating states,the harvester's forward-looking distance was taken as a key parameter to determine the gain coefficient,thereby obtaining its real-time control variables.To optimize the path tracking effect,a particle filter algorithm was introduced to improve the fish swarm algorithm,and a target function was constructed accordingly.During the target solving process,continuous iteration and optimization of the algorithm achieved precise path tracking of the harvester.[Results]After multiple experimental verifications,the proposed method demonstrates good tracking performance under different initial deviation points.Using this method,the average response time for path tracking was 0.52 seconds,the minimum turning radius was 5.0 meters,the average deviation is 0.8 meters,and the minimum deviation was 0.5 meters,which was basically consistent with the set route.This result fully demonstrated the effectiveness of the design method.[Conclusion]In summary,the proposed full-field path tracking for crawler harvester based on an improved fish swarm algorithm can accurately track the harvester in complex farmland environment with good tracking effects and wide application value.
作者
王芳
张伟
申珂
刘中峰
Wang Fang;Zhang Wei;Shen Ke;Liu Zhongfeng(Shanxi Agricultural Machinery Development Center,Taiyuan 030027,China;Shennong Agricultural Machinery Development(Shanxi)Group Co.,Ltd,Taiyuan 030031,China;Shennong Technology Group Co.,Ltd.,Taiyuan 030045,China)
出处
《山西农业大学学报（自然科学版）》
CAS
北大核心
2024年第4期109-117,共9页
Journal of Shanxi Agricultural University(Natural Science Edition)
基金
山西省科技重大专项计划“揭榜挂帅”项目(202201140601023)。
关键词
粒子滤波
鱼群算法
履带式收割机
路径跟踪
农田作业
Particle filtering
Fish school algorithm
Crawler harvester
Path tracking
Farmland operations
