|
所在平台: Udemy |
课程主页: https://www.udemy.com/course/tabtrainer-minitab-spc-charts-for-attribute-quality-data/
课程评论：没有评论
Coursera 课程总结：Tabtrainer Minitab：属性质量数据的 SPC 图 本课程是 Tabtrainer® 系列的进阶培训，专注于使用 Minitab 进行属性数据的统计过程控制 (SPC)。课程由 Tabtrainer® 创始人 Prof. Dr. Murat Mola 教授讲授，理论与实际应用相结合，旨在提供清晰、精确且可操作的教学。 **课程内容概览：** 课程深入浅出地介绍了属性控制图在 Minitab 中的应用，以滑板公司最终组装过程中的两个真实场景为例。重点讲解了用于属性数据的 P 图、NP 图、Laney P′ 图、U 图、C 图和 Laney U′ 图的理解、选择、应用、解释和区分。 **模块一：监控缺陷单位（二项分布）** * **P 图：** 监控不同子组大小下缺陷产品比例。 * **NP 图：** 评估恒定子组大小下的缺陷数量。 * **Laney P′ 图：** 修正 P 图，处理过度分散或分散不足的数据，提供更可靠的控制限。 * **关键学习点：** * 通过概率图诊断二项分布的适用性。 * 何时使用 Laney P′ 图避免误报或漏报过程偏移。 * 使用 Western Electric 控制规则检测特殊原因变异。 * 结合实际生产班次和检验质量解释控制图结果。 **模块二：监控缺陷计数（泊松分布）** * **U 图：** 跟踪每单位缺陷数，尤其适用于子组大小变化的情况。 * **C 图：** 分析恒定子组大小下的总缺陷计数。 * **Laney U′ 图：** 泊松分布不完全满足时使用的分散调整 U 图。 * **高级技术：** * 运行 U 图诊断，检查泊松分布拟合度。 * 根据 AIAG 公式手动计算控制限。 * 使用 Minitab 的“阶段”功能分割图表，对比改进前后的过程。 * 在不同条件下，对相同数据进行 U 图与 C 图的视觉比较。 * **数据分割的重要性：** 课程强调将不同过程阶段数据混合在同一图表中会导致控制限失真，正确分割可实现有意义的解释和真实的过程洞察。 **课程目标：** 学完本课程后，您将能够： * 根据缺陷类型、数据结构和分布选择合适的属性控制图。 * 理解二项分布和泊松分布质量数据的区别。 * 对 P、NP、U 和 C 图进行诊断并验证数据适用性。 * 解释概率图中的符合率和置信限。 * 使用 Laney 图校正过度分散或分散不足的数据，避免误读。 * 应用控制规则检测可分配原因和过程不稳定性。 * 使用阶段控制将分析分割为改进前后的过程阶段。 * 手动计算控制限，验证软件生成的结果。 * 在 Minitab 项目中以结构化的方式呈现和记录发现，用于质量报告。 本课程结合了理论、诊断和应用分析，是掌握 Minitab 属性 SPC 方法的完整学习之旅，对工业实践和学术发展都极具价值。
Welcome to this advanced training from the Tabtrainer® Series - a recognized learning platform for high-impact statistical training in industry and academia.This course is developed and taught by Prof. Dr. Murat Mola, founder of Tabtrainer®, certified by TÜV and awarded "Professor of the Year 2023" in Germany. Tabtrainer® courses are known for bridging the gap between theory and industrial application - with clarity, precision, and actionable outcomes.What This Course CoversThis comprehensive training course provides a deep, practice-driven introduction to Statistical Process Control (SPC) using attribute control charts in Minitab. It is based on two detailed real-world scenarios from the final assembly process of skateboards at Smartboard Company. The training focuses on understanding, selecting, applying, interpreting, and differentiating the most relevant SPC tools for attribute data: P charts, NP charts, Laney P′ charts, U charts, C charts, and Laney U′ charts.Participants learn not only the technical application of each control chart but also the underlying statistical distributions (binomial and Poisson), diagnostics, interpretation of process instabilities, and the impact of subgroup structure and process changes on control chart accuracy.Module 1: Monitoring Defective Units (Binomial Distribution)In the first part of the course, you will work with a dataset that reflects the number of defective skateboards identified during final surface inspection. The analysis focuses on:P Chart - to monitor the proportion of defective products across subgroups of varying size.NP Chart - to evaluate the number of defectives in subgroups of constant size.Laney P′ Chart - a modified version of the P chart that adjusts for overdispersion or underdispersion, providing more reliable control limits.Key learning points include:How to diagnose binomial suitability using probability plots.When to apply the Laney P′ chart to avoid false alarms or missed process shifts.How to detect special cause variation using built-in Western Electric control tests.How to interpret control chart results in the context of real production shifts and inspection quality.Module 2: Monitoring Defect Counts (Poisson Distribution)In the second part, you transition from the classification of defective units to analyzing the number of defects per product-such as surface scratches detected per skateboard. This requires a different statistical approach based on Poisson distribution and the use of:U Chart - for tracking the defects per unit, especially when subgroup sizes vary.C Chart - for analyzing total defect counts in subgroups of constant size.Laney U′ Chart - a dispersion-adjusted U chart used when Poisson assumptions are not fully met.This module also introduces advanced techniques such as:Running a U chart diagnostic to check Poisson distributional fit.Manual calculation of control limits based on AIAG formulas.Understanding and applying the "Stages" function in Minitab to split charts before and after process improvements.Visual comparison of U chart vs. C chart when working with the same data under different conditions.Learners explore how mixing data from two different process phases in a single chart leads to distorted control limits, and how correct segmentation enables meaningful interpretation and true process insight.By the End of the Course, You Will Be Able To:Select the appropriate attribute control chart based on defect type, data structure, and distribution.Understand the difference between binomially and Poisson-distributed quality data.Perform diagnostics and validate data suitability for P, NP, U, or C charts.Interpret agreement rates and confidence limits in probability plots.Use Laney charts to correct overdispersed or underdispersed data and avoid misinterpretation.Apply control tests to detect assignable causes and process instability.Split your analysis into pre- and post-improvement process phases using stage control.Manually calculate control limits to validate software-generated results.Present and document your findings in a structured Minitab project for quality reporting.This course combines theory, diagnostics, and applied analytics into a complete learning journey for mastering attribute SPC methods in Minitab-ideal for both industrial practice and academic advancement.
