Sampling Plan Design

Design sampling plans online and quantify acceptance and rejection risk.

Live calculation

DESIGN • COMPARE • DECIDE

设计一个满足风险目标的抽样方案

先定义良好与不良质量水平及双方风险目标，系统自动生成可行的 n–Ac/Re 方案，并解释风险与检验量取舍。

1定义质量与风险目标→2生成并比较可行方案→3查看推荐与工程解释
说明/示例
计数型 n-Ac/Re 抽样方案设计

定义类别并按 Ac/Re 判断本批样本结果。

默认参数仅用于演示；请按当前企业质量目标、批量和风险约束确认方案。

高级结果包含完整任务导航和高级结果查看。

1

接收抽样类型

先根据检验结果的数据形式选择计数型或计量型。

2

分析任务

再选择要设计方案、分析风险、判定批次还是比较方案。

Continuous measurement → n & k → Lot decision

Variables Acceptance Sampling

Design a plan for one measurable characteristic, then evaluate the lot using the sample mean, standard deviation and critical distance k.

Use only when measurements are approximately normal, the measurement system is suitable, and the characteristic has at least one specification limit. This risk-based calculator is not an embedded ISO 3951 table.

Characteristic & Specification

The variables plan evaluates one continuous characteristic at a time.

Use a historical σ only when the process is stable and sufficient historical evidence supports it.

Create Sampling Plan

Enter the good and poor quality levels and the maximum producer and consumer risks.

Enter the design targets and calculate the plan.
Sample size n——
Critical distance k——
Mean acceptance boundary——
Method——
—

Accept / Reject Lot

Paste the actual measurements. The decision compares Z.LSL and/or Z.USL with k.

Observed n——
Sample mean——
Standard deviation——
Z value(s)——

Lot result

—

—

Variables OC Curve

Exact for a single specification limit with known historical σ.

Engineering Interpretation

The interpretation separates plan performance from the actual lot decision.

iCalculate a plan or evaluate a lot to generate an interpretation.

Use Boundary

1Confirm that the measured characteristic is reasonably normal, or that the selected procedure is otherwise justified.
2The measurement system must be capable; precise formulas cannot correct unreliable measurements.
3Each plan evaluates one continuous characteristic. Multiple characteristics require separate plans.
4For formal ISO 3951 implementation, use the licensed standard and its switching, table and applicability rules.

Lot → Categories → Decisions

计数型 n-Ac/Re 抽样方案设计

输入质量参考点、风险上限和样本量约束，生成一个可执行的单次计数型抽样方案，并查看风险、检验量和后续行动建议。

默认参数仅用于演示；请按当前企业质量目标、批量和风险约束确认方案。

批次与检验设置

批量和检验设置由所有类别共享；每个类别仍可拥有独立的 AQL、n、Ac/Re 和风险参考点。

反向设计抽样方案

输入良好与不良质量水平以及双方风险上限，自动搜索同时满足目标的 n–Ac 方案。

选择类别或修改目标，然后点击“设计方案”。

结果只保证在当前模型、批量和风险参考点下满足目标；它是自定义 n–Ac/Re 方案，不是自动放行结论。

缺陷类别与判定条件

同一物理样本可以同时检查多个类别。每个类别分别累计并判定。

“实际发现”在不合格品百分率模式下指该类别的不合格产品数；同一产品可同时进入 Major 与 Minor 分类统计，但整批唯一不合格品数量不能简单相加。

抽样方案综合评价

按照当前定义的风险、输出质量和检验负担目标逐项判断，不使用所谓行业通用固定阈值。

α和β始终按当前方案中设置的目标评价。AOQL、ATI和抽样比例只有在勾选时才作为强制条件；未勾选时仍会作为工程权衡信息显示。

综合结论

—

—

整批综合状态

—

—

所选类别分析

—

p₁ 处接收概率
—
—
生产方风险 α
—
—
使用方风险 β
—
—
AOQL
—
—
预计质量水平下 ATI
—
—
抽样比例
—
—

OC 曲线：各类别对比

每条曲线对应一组独立 n–Ac 方案。

AOQ 曲线：各类别对比

仅在拒收批筛选纠正假设下显示。

类别汇总

风险目标与批次 Ac/Re 判定是两件事：前者评价方案，后者评价本次样本结果。

类别AQLnAc/Re实际发现αβAOQL本批判定方案风险

工程解释

先给出结论、关键证据和优先行动；详细风险、检验量及质量影响可展开查看。

—
高级结果与假设
查看详细工程解释