首页|期刊导航|智能化农业装备学报(中英文)|畜禽舍移动机器人自主导航关键技术试验研究

畜禽舍移动机器人自主导航关键技术试验研究OA

Experimental study on key technologies of autonomous navigation for mobile robots in livestock and poultry houses

中文摘要英文摘要

随着我国畜禽养殖业规模化、集约化程度持续提升,养殖密度不断增大,不仅加剧畜禽疫病传播风险,也对环境监测、粪污清理等作业提出更高要求.传统人工作业模式效率低下,而畜禽舍通道狭窄、障碍物分布不规律、地面湿滑且灰尘较多的复杂环境,导致移动机器人自主导航准确性和稳定性较差.为解决上述问题,研究提出融合改进跳跃点搜索算法(jump point search,JPS)与动态窗口法(dynamic window approach,DWA)的全局-局部联合路径规划方法.在全局规划层面,通过引入多维度综合代价函数、自适应跳点优化策略、贝塞尔曲线插值重构以及道格拉斯-普克冗余点剔除算法,全方位优化全局路径平滑度,大幅减少冗余路径节点与尖锐转角;局部规划层面,针对性优化动态窗口法目标函数权重参数与运动学约束,强化机器人实时动态避障与轨迹跟踪能力.仿真结果显示,在40×40和60×60标准栅格地图中,改进跳跃点搜索算法较传统算法,路径节点数分别减少77.55%、87.14%,累计转角分别降低70.91%、77.97%,搜索时间分别缩短 48.94%、75.15%,且始终保持最优路径长度.实物试验表明,融合 JPS-DWA算法能够成功避开未知障碍物,3个目标点重复导航的横向、纵向、航向平均偏差分别为0.48 cm、-0.66 cm和-0.17°,均优于文献方法,显示出较高的定位稳定性与导航重复精度.试验结果表明,研究所提出的方法能够在畜禽舍复杂非结构化环境中实现高精度高鲁棒性的自主导航,有效突破传统导航技术的应用局限,为畜禽舍移动机器人自动化巡检、消毒、清洁等作业提供核心技术支撑,助力畜禽养殖智能化与高效化转型.

Large-scale and intensive livestock production in China has greatly raised breeding density,elevating pathogen transmission risks and imposing stricter requirements on environmental monitoring and manure handling.Traditional manual methods suffer from low efficiency,and the unstructured barn environment,such as narrow walkways,randomly scattered obstacles,slippery and dusty ground,severely degrades navigation accuracy and robustness of mobile robots.To address these issues,this study proposes a global-local integrated path planning method combining an enhanced Jump Point Search(JPS)algorithm with the Dynamic Window Approach(DWA).For global planning,a multi-dimensional comprehensive cost function,an adaptive jump point optimization strategy,Bezier curve interpolation,and the Douglas-Peucker algorithm for redundant point elimination are introduced to comprehensively optimize global path smoothness,significantly reducing redundant path nodes and sharp turns.For local planning,the weighting parameters of the DWA objective function and kinematic constraints are specifically optimized to enhance the robot real-time dynamic obstacle avoidance and trajectory tracking capabilities.Simulation results on 40×40 and 60×60 standard grid maps show that,compared to the traditional algorithm,the improved JPS algorithm reduces the number of path nodes by 77.55%and 87.14%,cumulative turning angles by 70.91%and 77.97%,and search time by 48.94%and 75.15%,respectively,while consistently maintaining the shortest path length.Physical experiments demonstrate that the integrated JPS-DWA algorithm successfully avoids unknown obstacles.The average lateral,longitudinal,and heading deviations for repeated navigation to three target points are 0.48 cm,-0.66 cm,and-0.17°,respectively,outperforming existing methods and indicating high positioning stability and navigation repeatability.The proposed algorithm achieves accurate and robust autonomous navigation inside complex and unstructured livestock houses,effectively overcoming the limitations of traditional navigation technologies.It provides core technical support for automated mobile robot tasks in livestock houses,such as inspection,disinfection,and cleaning,thereby facilitating the intelligent and efficient transformation of livestock and poultry farming.

焦星惠;刘恩光;王锡泉;王帅智;史书阳;丁攀

河南农业大学机电工程学院,河南 郑州,450002||郑州市农业装备智能化技术与系统重点实验室,河南 郑州,450002河南农业大学机电工程学院,河南 郑州,450002||郑州市农业装备智能化技术与系统重点实验室,河南 郑州,450002河南农业大学机电工程学院,河南 郑州,450002||郑州市农业装备智能化技术与系统重点实验室,河南 郑州,450002河南农业大学机电工程学院,河南 郑州,450002||郑州市农业装备智能化技术与系统重点实验室,河南 郑州,450002河南农业大学机电工程学院,河南 郑州,450002||郑州市农业装备智能化技术与系统重点实验室,河南 郑州,450002河南农业大学机电工程学院,河南 郑州,450002||郑州市农业装备智能化技术与系统重点实验室,河南 郑州,450002

农业科技

移动机器人自主导航路径规划跳跃点搜索算法动态窗口法

mobile robotautonomous navigationpath planningJPS algorithmDWA algorithm

《智能化农业装备学报(中英文)》 2026 (3)

49-60,12

河南省重点研发专项(251111110400)河南省高等学校重点科研项目计划(25B460006) Key Research and Development Program of Henan Province(251111110400)Key Scientific Research Project Plan for Higher Education Institutions in Henan Province(25B460006)

10.12398/j.issn.2096-7217.2026.03.005

评论