摘要
由一名新闻记者-机器人与机器学习每日新闻的工作人员新闻编辑-调查人员发布了关于机器人的新报告。根据NewsRx Edi Tors在中华人民共和国武汉的新闻报道，研究表明：“机器人在现代英特尔智能制造和无人生产系统中扮演着举足轻重的角色，通常任务是准确地执行特定的路径。然而，机器人的输入是一条带有时间信息的路径。”本研究的资金支持包括湖北省重点研发项目。新闻记者引用华中大学科技研究的一句话:“由此产生的核心技术是轨迹规划方法,大致分为两大类:最大速度曲线(MVC)方法和多相直接配置(MPDC)方法。本文着重讨论了后一种方法所面临的挑战。在MPDC方法中,离散节点的数目对求解效率和精度有很大的影响，在处理复杂动力学系统时，如何在求解时间和路径离散误差之间取得平衡变得至关重要，本文在保证路径离散误差的前提下，采用网格细化(MR)算法来寻找合适的节点数目，因此，MR算法的传统应用要求在每次迭代中求解原问题,这些过程非常耗时,在处理复杂的动态系统时可能无法求解。本文提出了一种序贯最优轨迹规划方案，将原最优控制(OC)问题分为路径规划(PP)和路径规划(TP)两个阶段，在PP阶段，采用基于弧长的DC方法和MR算法识别指定路径上的关键节点，以最小化DU离散化引入的逼近误差。
Abstract
By a News Reporter-Staff News Editor at Robotics & Machine Learning Daily News – Investigators publish new report on robotics. Acc ording to news reporting out of Wuhan, People’s Republic of China, by NewsRx edi tors, research stated, “Robotic manipulators play a pivotal role in modern intel ligent manufacturing and unmanned production systems, often tasked with executin g specific paths accurately. However, the input of the robotic manipulators is t rajectory which is a path with time information.” Financial supporters for this research include The Key R&D Program of Hubei Province. The news reporters obtained a quote from the research from Huazhong University o f Science and Technology: “The resulting core technology is trajectory planning methods which are broadly classified into two categories: maximum velocity curve (MVC) methods and multiphase direct collocation (MPDC)methods. This paper conc entrates on addressing challenges associated with the latter methods. In MPDC me thods, the solving efficiency and accuracy are greatly influenced by the number of discretization nodes. When dealing with systems with complex dynamics, such a s robotic manipulators, striking a balance between solving time and path discret ization errors becomes crucial. We use a mesh refinement (MR) algorithm to find a suitable number of nodes under the premise of ensuring the path discretization error. So, the actual device can effectively implement the planned solutions. N onetheless, the conventional application of the MR algorithm requires solving th e original problem in each iteration; these processes are extremely time-consumi ng and may fail to solve when dealing with a complex dynamic system. As a result , we propose a sequential optimal trajectory planning scheme to solve the proble m efficiently by dividing the original optimal control (OC) problem into two sta ges: path planning (PP) and trajectory planning (TP). In the PP stage, we employ a DC method based on arc length and an MR algorithm to identify key nodes along the specified path. This aims to minimize the approximation error introduced du ring discretization.”
NETL
NSTL