红花采摘机器人集条预定位机构设计与试验

针对红花选择性采收中枝条摆动干扰识别和采摘的问题,该研究结合红花独特的生长特性,提出一种在识别前调整红花植株形态的集条预定位原理,设计了一种红花采摘机器人集条预定位机构,确定了红花植株集条机构和主干位置识别机构的主要结构参数,并通过分析获得集条作业与主干位置识别作业中凸轮各工作段的对应关系。田间试验表明:在集条夹持板间距和上沿圆弧半径分别为50和292 mm的情况下,防摆动枝条露出长度合格率为80.53%,夹果率为2.04%,植株损伤率为0.47%,果球损伤率为1.04%,幅宽比为16.64%,果球分散度为83.76%,果球遮挡率为6.51%,集条预定位作业可有效降低枝条的摆动,使整株红花果球呈条状有序分布,最终降低整株红花的识别和采摘难度。该研究可为红花自动化采摘提供参考。

Design and experiment of the strip-collected pre-positioning mechanism for safflower picking robots

Mechanical harvesting safflower has increased rapidly, with the continuous expansion of planting areas in recent years. However, the branch swing of safflower generally interferes with accurate visual recognition positioning and harvesting quality in selective harvesting of safflower. In this study, a strip-collected pre-positioning mechanism was proposed for a picking robot to adjust the morphology of safflower plants before visual recognition. The key factors were clarified on the swing of branches during harvesting, according to the growth characteristics of safflower, including the single safflower, the high growth center of the top of the fruit ball, the different opening time, and symmetrical distribution of spatial morphology of safflower plants. A safflower plant strip-collected mechanism also ran with the trunk position recognition to explore the plant shape for the pre-positioned state of strip-collected. The key working parts were designed to determine the structural parameters of safflower plant strip-collected and trunk position recognition mechanism. Specific works included: 1) A kind of safflower plant club mechanism was designed to meet the requirements of strip-collected operation, according to the four-bar kinematics. The key parameters were also optimized for the strip-collected trajectory drive mechanism and clamping plate. 2) The cam motion was utilized to determine the key parameters of trunk position recognition. The position relationship between the strip-collected operation and each working section of the cam was further analyzed to match the working stroke of strip-collected operation and the trunk position recognition. 3) A field test showed that the qualified rate of exposed length of anti swing branch was 80.53%, the fruit ratio of safflower fruit ball was 2.04%, the damage rate of plant was 0.47%, the damage rate of fruit ball was 1.04%, the width-to-width ratio was 16.64%, the dispersion of fruit ball was 83.76%, and the shading rate of fruit ball was 6.51%. Consequently, the strip-collected pre-positioning operation can effectively reduce the swing of the branches, where the whole safflower fruit balls are orderly distributed, and finally reduce the difficulty of identification and picking of the whole safflower. This finding can provide a sound theoretical basis for automatic safflower picking in modern mechanical harvesting.