What this category of data looks like
Data for this category originates from battery cell grading systems, workshop MES systems, welding and torque collection equipment, BMS software version repositories, and EOL test benches. Daily documents include cell grouping rules, BOM lists, welding torque records, BMS flashing logs, insulation and voltage withstand reports, EOL test reports, and system configuration lists. Data is updated each time a single PACK or energy storage cabinet is produced, and synced each time a BMS software version changes. Data fields include cell grading parameters, welding torque values, insulation test voltage, BMS version number, and system configuration parameters. Data is stored structurally based on actual collected content.
What constraints do these characteristics impose on the citation and traceability link
Data characteristics for this category impose multiple constraints on the citation and traceability link. First, data is updated per single PACK or energy storage cabinet production or single software change. It must be bound to a unique serial number or batch number for citation to avoid cross-batch confusion. Second, data sources cover multiple systems including industrial equipment, MES systems, and test benches. Cross-data-source associated traceability must be supported, such as binding welding torque records to fastener information from the corresponding BOM. Third, daily documents include multiple types of test reports. Citations must match test items to corresponding production batches to ensure accurate traceability information. Fourth, data updates from BMS version changes must be associated with version identifiers during citation to avoid using expired configuration data.
How to configure this
This page’s configuration is based on FastGPT v4.17.0. Specific configuration items are listed below:
| Configuration Item | Recommended Setting | Rationale |
|---|---|---|
| Similarity Threshold | 0.75–0.85 | This category of data includes highly relevant fields such as precision torque and insulation parameters. A threshold that is too low will introduce irrelevant data, while a threshold that is too high will fail to retrieve valid records |
| Single Search Limit | Top 8 entries | The volume of associated data for a single PACK is limited. Too many retrieved entries will exceed the context window and reduce retrieval efficiency |
quoteMaxToken | 1200–1500 token | This category of documents includes multiple sets of test data and configuration parameters. Sufficient token count is required to store complete traceability information |
responseData | Only return associated data bound to the batch number | Ensure that cited data is strictly bound to the current audit batch to avoid incorrect cross-batch citations |
detail | Return complete source path and timestamp | Process quality audits require tracing data collection time and source equipment to ensure compliance |
| Citation Template | Concatenate in the format "batch number + source equipment + collection time" | This category requires batch-based traceability, and this format allows quick location of corresponding production records |
Three common configuration mistakes
- Phenomenon: Citation results include production records from non-current batches. Cause: No batch number binding retrieval rule is configured, leading to retrieval of irrelevant cross-batch data.
- Phenomenon: Form node reference variables cannot obtain corresponding batch data. Cause: Batch number is not used as a pre-filter condition for retrieval, and the retrieval scope bound to the variable does not cover the current audit order.
- Phenomenon: Cited BMS configuration parameters do not match the actual batch. Cause: No retrieval is performed using the BMS software version field, and expired version data is used, leading to incorrect audit results.
How to verify correct configuration
- Enter the batch number of a single PACK, retrieve associated welding torque records, and check that the returned results only include data from that batch.
- Review the source path and timestamp of cited content to confirm that collection information and test time from the corresponding equipment are included.
- Trigger a simulated audit for an EOL test, and check that the returned cited data matches the currently used BMS software version.
- Adjust the test value of the similarity threshold, confirm that the relevance of retrieval results meets business requirements, and that there is no excessive irrelevant data.
The values provided are common starting points and should be measured against the reader's own samples.