考虑倒伏与粮仓容积的再生稻头季机收路径规划

为提升机具作业效率和质量、减少土壤碾压等，该研究提出一种同时考虑作物倒伏状态、碾压面积和收获机粮仓容积的路径规划算法（ harvester grain bin capacitated arc routing problem，HGBCARP）。该算法由作业信息处理模块和作业路径规划模块组成，作业信息处理模块将农田边界、卸粮点位置、作物倒伏方向及面积、位置等信息转化成可处理的数据形式并传输给路径规划模块，然后由路径规划模块进行作业行方向划分、作业行遍历顺序寻优、转弯方式生成、碾压面积计算等，最终得到最优路径规划结果。采用改进遗传算法，分别以3种再生稻收获机、2种传统水稻收获机和3种不同田块为对象，以行驶路径长度、碾压面积、收获粮食量为评价指标，采用回转式收获路径和HGBCARP式收获路径规划开展对比试验。结果表明，HGBCARP式比回转式收获路径碾压面积减少11.79%～27.20%，可使倒伏的机收头季稻增收1.64%～1.95%；同时在3种不同田块条件下进行仿真试验，HGBCARP式比回转式收获路径可使碾压面积减少7.25%～20.09%。使用电动无人履带式底盘对不同收获路径进行田间模拟收获试验，HGBCARP式收获路径与传统牛耕往复式及回转式收获路径相比，碾压面积减少约11.21%～28.03%，在路径长度减少约6.81%～23.46%，验证了HGBCARP式路径规划方法的有效性，研究结果可为智能化作业路径规划研究提供参考。

Path planning of mechanical harvesting considering the lodging and grain bin capacity for the ratoon rice in main season

Ratoon rice is a typical cultivation to allow for the growth of another crop of rice using dormant seedlings from the previous harvest. Seedling, raising, and transplanting of double-season rice can be removed to reduce the harm from pesticide residues. Cultivated land quality can also be protected at a low cost. Additionally, the quality and fullness of the second crop of ratoon rice are superior to that of double-season rice. The planting area of machine-harvested ratoon rice has increased rapidly in recent years. Therefore, most research has been focused on the harvest technology and equipment of ratoon rice. Particularly, ratoon rice harvesting is required for the low rolling and high stubble retention for optimal agronomic conditions. The rolling of rice stubble during harvest in the main season can significantly reduce the yield of ratoon rice in the ratoon season. Once the lodging occurs at the mature stage of crop growth, it is detrimental to the harvesting yield and quality. Therefore, the optimal path of harvesting is crucial to improve the efficiency of grain production for better harvesting with minimal loss. Concurrently, the volume limit of the grain bin in the harvesters and the location of the unloading points also exert a significant impact on the loss of harvesting and low rolling of the ratoon rice. Therefore, it is necessary to design the harvest path of the ratoon rice harvesters for the maximum yield in the main season. The optimal path for high-quality agricultural machinery is crucial to saving time and energy, particularly for better operation efficiency and quality with less soil rolling. In this study, a path planning was proposed, called the harvester grain bin capacitated arc routing (HGBCARP). Two modules consisted of information processing and path planning. The information processing module was used to convert the operation information (such as farmland boundaries, unloading point positions, crop lodging directions, and area positions) into a processable data form, and then transmit it to the path planning module. The optimal path planning was obtained after the tasks, such as the direction division of the operation line, the optimization of the operation line traversal sequence, the generation of turning modes, and the calculation of the rolling area. An improved genetic algorithm was utilized to evaluate the performance of three new types of ratoon rice harvesters, as well as two traditional rice harvesters. Three parameters of the field test were taken as evaluation indicators, including the length of the driving path, the rolling area, and the amount of harvested grain. A comparative simulation was then conducted under three field conditions using rotary and HGBCARP harvesting path planning. It was found that the rolling area of the HGBCARP harvesting path was 11.79% to 27.20% less than that of the rotary type. The rolling area was reduced to increase the yield of the main season rice by 1.64% to 1.95%. Additionally, the HGBCARP harvesting path was found to reduce the rolling area by 7.25% to 20.09%, compared with the rotary type. Field harvesting experiments were carried out using an electric unmanned crawler chassis on various harvesting paths. The HGBCARP harvesting path was found to reduce the rolling area by 11.21% to 28.03% and the path length by 6.81% to 23.46%, compared with the traditional cattle tillage reciprocating and rotary harvesting paths. The high effectiveness was achieved in the HGBCARP path planning. This finding can also provide a valuable reference for intelligent operation path planning.