复杂工况下油菜机械化生产机器田间作业路径规划及多目标优化

For agricultural operations in fields of known boundary information, it is very important in practice to obtain optimal operation direction, trajectory and scheduling strategies by systematic planning and optimization, which helps improving the agricultural operations’ efficiency, quality and utilization rate of machinery, reducing operating consumption and labor intensity, increasing land productivity and effective crop yields. Based on the existing work, taking key links of Canola mechanized production like seeding, spraying and harvesting as the main applications, aiming to optimize the operation efficiency, quality and overall consumption, using technical means including theoretical analysis and calculation, quality assessment of software simulation and field trials, this proposal will investigate the key issues of "people - machinery - crops - soil" interaction system of the agricultural field operations in several different complex conditions such as a variety of equipment characteristics, complex terrain in non-plain areas, multi-machine collaboration and so on. It will focus on: ①multi-objective path planning and optimization considering different machines feature for operation efficiency, quality and overall consumption; ②coverage path planning considering differences in the field terrain especially three dimensions; ③planning algorithm for multiple cooperative machinery in field of large area; ④scheduling optimization strategies for machinery in several adjacent fields. Research results will directly provide theoretical and technical support for precision agricultural operations based on GNSS navigation and autopilot and intelligent scheduling and management of field machinery.

对已知边界信息田地进行农机作业路径的系统规划优化,获得合理的作业方向、轨迹和调度策略，对于提高作业效率和质量，降低作业消耗，增加土地产出和作物有效产量，有着重要实践意义。在已有工作基础上，以作业效率、质量和总消耗等为优化目标，以直播、植保和收获等油菜机械化生产关键环节为主要应用对象，考量多种机具特性、非平原地区复杂地形、多机并行/协同及跨田块等多种不同复杂工况，采用理论分析计算、软件仿真评估和实地田间试验等技术手段，探讨农机田间作业时 “人-机具-作物-土壤”互作系统中的基础路径规划优化方法，重点研究:①不同机具特性条件下规则边界田块单机多目标全覆盖路径规划；②差异地形及边界条件下全覆盖路径规划；③大面积地块中多机并行/协同条件下作业路径规划；④相邻多田块条件下农机作业调度及优化策略。研究将为主要农作物机械化生产的GNSS导航自动驾驶农机的田间精准作业及农机智能调度管理提供理论及技术支撑。

事先设定的目标作业路径是农机“铁牛”田间作业行进时的“牵牛绳”，使能、使优智能农机作业过程，一定程度上决定着农业机械田间作业过程的质量、效率和燃油、种子、化肥等农资消耗量。以地面机器联合直播、飞播/防、联合收获等油菜全程机械化典型作业环节的作业过程为具体研究对象，本项目研究利用计算机技术自动化计算已知边界信息田块中农机作业路径过程中的基础问题。以作业效率、质量和总消耗等为优化目标，采用理论分析计算、软件仿真评估和实地田间试验等技术手段，分析地面机具传统U型、Ω型和T型转弯策略的基础上，设计了M型和T’型2种新型转弯策略及其对应的转弯路径生产算法，减少转弯路径长度和地头空间；在开展多边形快速填充、旅行商问题等基础算法的研究基础上，针对常见矩形和凸边形边界田块中的履带式联合收获作业和轮式拖拉机牵引的耕播联合作业过程，设计了全覆盖作业路径规划算法；针对农用多旋翼无人机，设计了凸多边形、凹多边形、含障碍物孔洞多边形及多个多边形等任意复杂边界田块和单机单航次、单机多航次及多机多航次等复杂工况条件下的全覆盖飞播/飞防作业路径规划算法。在《农业机械学报》和《农业工程学报》等期刊上发表学术论文8篇，获得软件著作权1项，申请国家发明专利6项，其中授权5项、审查中1项、转化3项。研究成果支撑稻油轮作无人农场的建设，为主要农作物机械化生产的GNSS导航自动驾驶农机的田间精准作业及农机智能调度管理提供理论及技术支撑。