SMT贴片工序外观质量统计数据集

原始数据来源于AOI检测设备每日导出的检验记录及人工补充的日报。首先对表格中的合并单元格进行拆分，并通过向下填充方式补全日期、线别、班次、客户、生产订单号、产品型号等维度字段，确保每条记录具有完整的标识信息。随后，将日期格式统一转换为“YYYY-MM-DD”；插入良率字段（（当日生产数－不良数）/当日生产数）计算，以及是否达标字段（采用IF函数进行判定，当良率≥99.7%时返回“达标”，否则返回“未达标”）；对于缺陷列中的空单元格，统一填充为0，表示该缺陷类型未发生；为保护商业信息，将客户名称进行匿名化处理。最终形成结构化、可直接用于统计分析的数据集。The raw data is sourced from daily inspection records exported by AOI testing equipment and manually supplemented daily reports. First, merged cells in the tables are split, and dimensional fields including date, production line, shift, customer, production order number, and product model are supplemented via the fill-down method, ensuring that each record has complete identifying information. Subsequently, the date format is uniformly converted to "YYYY-MM-DD"; a yield rate field is added, calculated as (daily production quantity - defective quantity)/daily production quantity, and a compliance status field is added, which is determined using the IF function: returns "Compliant" when the yield rate ≥ 99.7%, otherwise returns "Non-compliant". Empty cells in the defect columns are uniformly filled with 0, indicating that the corresponding defect type did not occur. To protect commercial information, customer names are anonymized. Finally, a structured dataset that can be directly used for statistical analysis is formed.

搜集汇总

背景与挑战

背景概述

该数据集基于AOI检测设备和人工日报的原始数据，经过系统预处理，包括字段补全、格式统一、良率计算和匿名化处理，形成了结构化的SMT贴片工序外观质量统计记录。它包含日期、线别、班次、客户、生产订单号、产品型号等维度信息，以及良率和是否达标等关键指标，可直接用于质量分析和统计建模。数据集旨在支持生产过程中的质量监控和改进决策。