名优茶采摘机械手与采摘策略研究

发布时间：2018-10-21 18:19

【摘要】：名优茶具有较高的实用性与经济价值,其产业价值占茶产业的比例逐渐增高。现在市场上的采茶机采用一刀切方式,无选择性,难以保证名优茶的品质,而名优茶采摘要求匀嫩齐,仍为人工采摘,虽能满足严苛的采摘要求,但由于经济转型和劳动力的缺乏导致名优茶采摘不及时,造成巨大的浪费,采摘难成为了名优茶产业发展的瓶颈。实现名优茶机械化采摘的难点主要有2个,一个是鲜叶识别,另一个是高效采摘。本文提出的名优茶采摘机,使用图像技术识别鲜叶,使用机械手进行单芽采摘。机械手的采摘系统是本文研究重点,为提高采摘速率,系统使用双采摘机械手,并采用蚁群算法规划机械手采茶路径。主要研究内容如下:1)阐述了名优茶采摘机的工作原理与名优茶采摘机系统组成。名优茶采摘机主要包括行走装置、采摘机械手、图像识别系统和和控制系统等。使用高地隙动力底盘作为采茶机的前进动力与支撑平台,使用图像识别技术获取鲜叶坐标,使用嵌入式系统开发控制平台。2)提高茶叶采摘效率的措施之一是使用双采摘机械手。2个三自由度机械手可独立工作且有重叠的工作区域。机械手用双圆轴导向,由步进电机驱动。机械手的末端执行器采用曲柄滑块机构,鲜叶剪切部件结构为圆筒式,由气缸驱动刀片开合。末端执行器中内筒的两端连接塑料软管,以吹气加吸气的方式收集已采摘鲜叶。3)提高茶叶采摘效率的另外一个措施是使采摘机械手的运动距离最短。蚁群算法在解决旅行商(TSP)问题上有独特的优势,鲜叶的采摘路径类似于旅行商问题。本文将蚁群算法与鲜叶采摘技术相结合来规划采摘路径。论文分别进行了2区、4区和6区的茶叶分区对路径规划影响的研究,分别使用基本蚁群算法和优化蚁群算法进行了仿真分析。仿真结果表明分4区的采摘效果较好,优化算法后规划路径更佳。论文对鲜叶分布均匀、极不均匀、较多和较少的情况进行了研究,结果表明本文的采摘策略适应性较好。
[Abstract]:Famous and excellent tea has high practicability and economic value, and its industrial value has gradually increased in the proportion of tea industry. At present, tea pickers in the market adopt a one-size-fits-all approach without selectivity, so it is difficult to guarantee the quality of famous and excellent tea. However, the requirements for picking famous and excellent tea are even and tender and still artificial, although they can meet the stringent requirements of picking. However, due to the economic transformation and the lack of labor force, famous and excellent tea picking is not timely, resulting in a huge waste, picking difficulty has become the bottleneck of the development of famous and excellent tea industry. There are two difficulties in realizing mechanized picking of famous and excellent tea, one is the recognition of fresh leaves, the other is efficient picking. In this paper, the famous and excellent tea picking machine uses image technology to identify fresh leaves and manipulator to pick single bud. The picking system of manipulator is the focus of this paper. In order to improve the picking rate, the dual picking manipulator is used in the system, and ant colony algorithm is used to plan the tea picking path of manipulator. The main contents are as follows: 1) the working principle of famous and excellent tea pickers and the system composition of famous and excellent tea pickers are expounded. The famous and excellent tea picking machine mainly includes walking device, picking manipulator, image recognition system and control system. The dynamic chassis of highland gap is used as the forward power and support platform of tea picker, and the image recognition technology is used to obtain the coordinates of fresh leaves. One of the measures to improve the efficiency of tea picking is to use double picking manipulator. Two three degree of freedom manipulators can work independently and have overlapping working areas. The manipulator is guided by a double circular shaft and driven by a step motor. The end actuator of the manipulator adopts the crank slider mechanism, the fresh blade shearing part is cylindrical, and the blade is opened and closed driven by the cylinder. The plastic hose is connected to the two ends of the inner cylinder of the end actuator to collect the fresh leaves collected by blowing and inhaling. 3) another measure to improve the efficiency of tea picking is to make the moving distance of the picking manipulator shortest. Ant colony algorithm has a unique advantage in solving the (TSP) problem of traveling salesman, and the picking path of fresh leaves is similar to that of traveling salesman problem. In this paper, ant colony algorithm and fresh leaf picking technology are combined to plan picking path. In this paper, the influence of tea partition in region 2, 4 and 6 on path planning is studied, and the basic ant colony algorithm and the optimized ant colony algorithm are used to simulate the path planning. The simulation results show that the picking effect of the four areas is better, and the planning path after the optimization algorithm is better. In this paper, the distribution of fresh leaves is uniform, very uneven, more and less. The results show that the picking strategy of this paper is more adaptable.
【学位授予单位】：华中农业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：S225.99

【参考文献】