结论：选对代表型号，论证清楚差异，就能覆盖整个系列

轮椅企业常用一个基础平台衍生出多个型号——不同座宽、不同轮径、不同配置、不同颜色。如果每个型号都做完整的型式检验，成本会很高。

可行的做法是覆盖论证：选择代表性型号做完整测试，对其余型号论证其被覆盖。

论证成立的条件是：代表型号在相关性能上不优于其他型号，且差异点的影响被说明或单独验证。

论证写得好，能省下大量重复测试；论证草率，会被要求补测，反而拖延进度。

论证的基本思路

第一步，明确系列的范围。 哪些型号属于这个系列。

第二步，列出型号间的差异。 逐项列清楚差异在哪里。

第三步，分析差异对各测试项目的影响。 某个差异会不会影响某个项目的结果。

第四步，确定代表性型号。 按各项目的最不利原则选择。

第五步，对无法被覆盖的差异做单独验证。

第六步，形成覆盖说明文件。

第三步是论证的核心。 差异与项目的对应关系要说清楚，不能笼统说「差异不影响性能」。

代表性型号的选择

选择的原则是「最不利」，但不同项目的最不利型号可能不同：

测试项目 最不利的型号特征
静态强度 额定载重最大、座宽最大
疲劳 载重最大、结构最简的
稳定性 重心位置偏上、轮距较窄、座位较高
制动 整机最重、速度最快
电气安全 功率最大、电池容量最大
电磁兼容 电子件最多、功率最大
阻燃 面料与填充材料的组合
折叠尺寸 各型号分别测量

注意各项目的最不利型号可能不是同一个。 所以可能需要选择多个代表型号，每个覆盖一组项目。

「稳定性」是容易判断错的项目。 最不利的不一定是最大最重的型号，而是重心相对偏高、支撑基面相对偏小的型号，这需要分析而不是凭直觉。

常见的差异类型与处理

座宽差异。 影响强度（座面跨度）、稳定性（宽度）、折叠尺寸。通常取最大和最小座宽做验证。

轮径差异。 影响稳定性、越障、行驶性能。可能需要分别验证。

电池容量差异。 影响重量、续航、电气安全。取最大容量验证安全，各容量分别测续航。

电机功率差异。 影响速度、爬坡、电气安全、EMC。取最大功率验证。

颜色差异。 通常不影响机械性能，但可能影响阻燃和生物相容性（不同色母）。

配置差异。 选装件的有无，需要单独考虑。

面料差异。 不同面料的阻燃和生物相容性可能不同，需要分别验证或论证等同。

「颜色差异」的处理常被简化。 机械性能确实不受颜色影响，但不同色母的化学成分不同，阻燃和生物相容性可能有差异。建议对颜色做表征对比，差异不大的可以合并，差异明显的分别处理。

论证不成立的情形

以下情况下覆盖论证难以成立：

结构有实质差异。 不只是尺寸缩放，而是结构形式不同。

材料不同。 主要受力件用了不同材料。

工艺不同。 不同型号用不同的生产工艺或场地。

功能有增减。 某些型号有额外的功能。

适用人群不同。 比如成人型号与儿童型号。

这些情况下应当分别测试，或者至少对差异部分做单独验证。

资料的组织

覆盖说明文件建议包含：

系列的定义。 包含哪些型号，型号的命名规则。

差异对照表。 逐项列出各型号的差异。

影响分析。 每个差异对各测试项目的影响判断。

代表型号的选择及理由。 为什么选这些型号。

覆盖关系表。 哪个型号的哪个项目由哪份报告覆盖。

单独验证的项目。 无法覆盖的部分及其验证数据。

结论。 整个系列的符合性论证。

「覆盖关系表」是文件的核心。 有了这张表，任何一个型号的任何一个项目，都能追溯到支撑它的数据。

后续维护

系列会扩展，覆盖论证要相应维护：

新增型号时。 判断是否落在现有覆盖范围内，超出的部分补验证。

型号变更时。 变更是否影响覆盖关系。

代表型号变更时。 如果代表型号本身变了，覆盖关系要重新评估。

标准更新时。 新标准的要求是否仍被覆盖。

建议把覆盖文件作为受控文件维护，每次系列变动时更新，而不是每次重新做一份。

论证的写法要点

覆盖论证容易写空，几个要点：

具体到项目。 说明某个差异对某个具体测试项目的影响，而不是笼统说不影响性能。

说明机理。 为什么不影响，依据是什么。

量化差异。 差异有多大，比如座宽相差多少。

引用数据。 如果有支持数据（比如表征对比），一并提供。

承认局限。 无法覆盖的部分明确说明并单独验证。

「承认局限」反而增强论证的可信度，全部声称被覆盖而不做任何单独验证的论证，通常会被质疑。

与变更管理的衔接

覆盖论证与变更管理是相关的：新型号的加入本质上是一次变更；现有型号的变更可能影响覆盖关系；代表型号的变更影响整个系列。

建议把覆盖文件的更新纳入变更评审的固定环节，避免系列扩展后覆盖关系失效而无人察觉。

我们的做法

承接系列产品委托时，我们会先看型号间的差异清单，据此与委托方讨论代表型号的选择。 有时委托方选的代表型号在某些项目上不是最不利的，这时论证会有缺口。

对于差异较多的系列，我们可能建议选择多个代表型号，每个覆盖一组项目 ——这比选一个「综合最不利」的型号更准确，因为综合最不利的型号未必存在。

测试安排上，我们会在报告中明确样品的型号和规格，便于后续的覆盖论证引用。

如果你有系列产品需要做覆盖论证，可以把型号清单和差异说明发过来一起讨论，或者直接联系：132 4819 8029。检测能力见服务介绍，产品分类见手动轮椅检测与电动轮椅检测，案例见案例。

English version

Conclusion. Wheelchair manufacturers commonly derive several models from one base platform, differing in seat width, wheel size, configuration and colour. Full type testing of every model would be expensive. A worked alternative is coverage justification: testing representative models completely and arguing that the remainder are covered. The argument holds where the representative model is no more favourable than the others in the relevant performance aspects and where the effects of the differences are either explained or separately verified. A well-written justification saves a great deal of duplicated testing; a careless one draws requests for supplementary testing and delays the schedule instead.

The basic approach. Define the scope of the family, stating which models it contains. List the differences between models item by item. Analyse the effect of each difference on each test item, establishing whether it could change the result. Determine representative models on a least-favourable basis. Verify separately any difference that cannot be covered. And produce a coverage statement document. The third step is the core: the correspondence between differences and test items must be explicit, not a general assertion that differences do not affect performance.

Choosing representative models. The principle is least favourable, though the least favourable model differs between items. For static strength it is the highest rated load and widest seat. For fatigue it is the highest load and simplest structure. For stability it is the highest centre of gravity, narrowest track and highest seat. For braking it is the heaviest and fastest. For electrical safety it is the highest power and largest battery. For electromagnetic compatibility it is the most electronics and highest power. For flammability it is the combination of upholstery and filling materials. And folded dimensions must be measured on each model. Note that the least favourable model differs by item, so several representatives may be needed, each covering a group of items. Stability is the item most often misjudged: the least favourable is not necessarily the largest and heaviest but the one with the highest centre of gravity relative to the smallest support base, which requires analysis rather than intuition.

Common differences and their handling. Seat width affects strength through seat span, stability through width, and folded dimensions; usually the widest and narrowest are verified. Wheel size affects stability, obstacle negotiation and ride, and may need separate verification. Battery capacity affects weight, range and electrical safety; verify safety at the largest capacity and measure range for each. Motor power affects speed, gradient capability, electrical safety and electromagnetic compatibility; verify at the highest power. Colour generally does not affect mechanical performance but may affect flammability and biocompatibility through different masterbatches. Configuration differences from optional items need separate consideration. And upholstery differences may affect flammability and biocompatibility, requiring separate verification or an equivalence argument. Colour is commonly oversimplified: mechanical performance is indeed unaffected, while masterbatches differ chemically and flammability and biocompatibility may differ. Compare colours by characterisation, combining those that match closely and handling divergent ones separately.

Where the argument fails. Coverage justification is hard to sustain where structures differ substantively rather than merely in scale; where materials differ in principal load-bearing parts; where processes differ, with different models made by different processes or at different sites; where functions are added or removed; or where the intended population differs, as between adult and paediatric models. In those cases test separately, or at least verify the differing parts individually.

Organising the documentation. A coverage statement should include the family definition, stating which models it contains and the naming convention; a difference table listing each model's differences item by item; an impact analysis judging each difference against each test item; the representative model selection and its reasoning; a coverage table showing which report covers which item for which model; separately verified items with their data; and a conclusion on the family's conformity. The coverage table is the heart of the document: with it, any item for any model traces to the data supporting it.

Ongoing maintenance. Families expand, and the justification must keep up. When models are added, judge whether they fall within the existing coverage and verify what does not. When models change, judge whether the coverage relationships still hold. When a representative model itself changes, reassess the coverage. And when standards are updated, check whether the new requirements remain covered. Maintain the coverage document under document control, updating it whenever the family changes rather than producing a new one each time.

How we handle it. For family submissions we review the difference list first and discuss representative model selection with the client. Sometimes a client's chosen representative is not the least favourable for certain items, which leaves a gap in the argument. For families with many differences we may suggest several representatives, each covering a group of items, which is more accurate than seeking one overall least-favourable model, since such a model may not exist. In reporting we state the sample's model and specification explicitly so that the coverage justification can cite it.

Send us the model list and difference description and we will work through the coverage argument. Phone or WeChat: +86 132 4819 8029.