| 摘要: |
| 针对现有激光增材再制造性评价方法主观性强、效率低等不足，提出一种失效数据驱动的磨损失效退役零件增材再制造性神经网络量化评价方法。根据磨损失效退役零件的激光增材再制造修复难易程度与失效模式相关的特点，以修复路径规划可行性、运动轨迹规划可行性、激光熔覆材料选择、再制造时长、再制造经济性为评价指标，构建磨损失效退役零件激光增材再制造性层次评价模型；通过退役零件失效区域的修复路径规划、修复设备运动轨迹模拟和碰撞检测进行修复路径可行性和运动轨迹可行性指标量化，并定义再制造时长、经济性指标的量化公式。以多个同类零件为对象，通过上述量化评价方法构建样本空间，基于神经网络训练获得再制造性神经网络量化评价模型，实现同类零件的快速激光增材再制造性评估。最后，以碎煤机端盘为例对所提方法进行验证，结果表明了该方法的可行性与有效性。根据失效数据与退役零件再制造性之间的映射关系，可实现磨损失效零件增材再制造性的快速量化评价，为其工程应用提供技术支撑。 |
| 关键词: 激光增材再制造 量化评价 轨迹规划 路径规划 神经网络 |
| DOI：10.11933/j.issn.1007?9289.20220301001 |
| 分类号:TH17;TP24 |
| 基金项目:国家自然科学基金(51965049)、内蒙古自治区关键技术攻关计划(2021GG0261)和内蒙古自治区高等学校创新团队发展计划支持(NMGIRT2213)资助项目 |
|
| Quantitative Evaluation of Wear Failure Part’ s Remanufacturability for Laser Additive Remanufacturing Based on Neural Network |
|
ZHANG Qi1,2, ZHANG Xiufen1,2, YU Gang3
|
|
1.College of Mechanical Engineering, Inner Mongolia University of Technology, Hohhot 010051 , China;2.Inner Mongolia Manufacturing Informatization Productivity Promotion Center, Hohhot 010051 , China;3.Inner Mongolia Vocational College of Mechanical and Electrical Technology, Hohhot 010070 , China
|
| Abstract: |
| Aiming at the shortcomings of high subjectivity and low efficiency of the existing laser additive remanufacturing evaluation methods, a neural network quantitative evaluation method for additive remanufacturability of wear failure retired parts driven by failure data is proposed. According to the relationship between the difficulty of laser additive remanufacturing repair and failure modes, the feasibility of repair path planning, motion trajectory planning, laser cladding parameter selection, remanufacturing time and remanufacturing economy are taken as the evaluation indexes, and the remanufacturability hierarchical evaluation model of wear failure retired parts for laser additive remanufacturing is constructed. Moreover, the repair path planning of the failure area of retired parts, repair equipment trajectory simulation and collision detection are used to quantify the feasibility index of repair path and motion trajectory, and the quantitative formulas of remanufacturing time and economic index are defined. Furthermore, in order to realize the rapid evaluation of laser additive remanufacturability for the same type parts, taking several similar parts as the objects, the sample space is constructed through the above quantitative evaluation method, and the remanufacturability neural network quantitative evaluation model is obtained based on neural network training. Finally, the proposed method is verified by taking a coal crusher end disk as an example, and the results show that the feasibility and effectiveness of this method. According to the mapping relationship between failure data and remanufacturing of retired parts, the rapid quantitative evaluation of additive remanufacturing of worn-out parts is realized, which provides technical support for its engineering application. |
| Key words: laser additive remanufacturing quantification trajectory planning route planning neural network |
