苹果采摘路径规划最优化算法与仿真实现

Algorithm for optimization of apple harvesting path and simulation

采摘路径规划对苹果采摘机器人的工作效率有很大的影响,为了提高苹果采摘机器人采摘效率,研究了采摘路径规划最优化方法。将苹果采摘的路径规划问题转化为三维的旅行商问题进行求解,结合图像识别得到的苹果位置特征,提出了有限域信息素自适应更新的改进蚁群算法,避免了基本蚁群算法求解过程中的早熟和局部收敛的问题,研究了三维模型的建模和驱动方法。试验结果表明将蚁群算法用于解决苹果采摘路径规划问题,当苹果数量达到250个时,改进蚁群算法迭代次数是基本算法的25.3%,而搜索到的最优路径是其94.3%,可见改进算法在搜索次数和最优结果上都有明显的优势。本研究为苹果采摘机器人采摘路径规划的提供理论参考。

Path planning is pivotal to the working efficiency of apple harvesting robot, in order to improve the picking efficiency of harvesting robot, the algorithm for optimization of apple harvesting path was studied. In this study, path planning is translated into three dimensional Travel Salesman Problem. According to the apple position obtained through image processing, an improved ant colony algorithm was studied in the optimization of apple-picking path, which can overcome premature convergence and local convergence of the search. Three-dimensional model building and driving methods were studied. Experimental results show that when there are 250 apples, the improved ant colony algorithm can reduce alternate times to 25.3% of basic algorithm, while the total distance is 94.3%, therefore, the improved ant colony algorithm cannot only reduce algorithm times, but also can obtain better values. This study is of great importance to the optimization of apple-picking path.