What the Data for This Category Looks Like
Data for this category comes from design BOMs, silicon steel sheet and insulating material procurement data, vacuum drying equipment logs, oil test reports, high-voltage test records from test stations, and product nameplate parameter tables. Daily documents include winding and core assembly records, vacuum drying records, insulating oil test reports, partial discharge and withstand voltage test reports, and product nameplate parameter tables. Data is updated per single product or each batch of raw materials. Core fields include insulation distance parameters, water content test values, partial discharge thresholds, and tap position identifiers. Common units are millimeters, percentages, picocoulombs per second, and kilovolts.
Constraints Imposed on Workflow Orchestration
The update rhythm for single products or each batch of raw materials requires workflows to bind a unique product identifier to trigger execution. This prevents cross-batch data confusion. Data from multiple sources is scattered across device logs, assay reports, and test records. Workflows must support initiating queries across multiple specified databases and integrating all scattered full documents. Core fields such as insulation distance and partial discharge thresholds have strict precision requirements. It is necessary to limit the scope of fields for retrieval and matching to avoid irrelevant data interfering with review results. Failure consequences include insulation breakdown or grid site faults. Workflows must be configured with exception branch handling to trigger alerts for matching exceptions. Products in this category require single-unit traceability. Workflows must accurately associate all documents for a single product, and cannot use batch fuzzy matching logic.
Configuration Settings
This configuration is based on FastGPT v4.17.0.
| Configuration Item | Recommended Setting | Rationale |
|---|---|---|
| variables | {"product_id": "${trigger.product_id}", "insulation_threshold": 10 picocoulombs per second, "required_fields": ["insulation distance", "partial discharge value"]} | Binds a single product identifier, limits core retrieval fields, and meets the single-unit traceability requirement for this category |
| Single Search Limit | Top 6 entries | The number of core test reports for this category is limited; limiting result entries avoids redundant data interference |
| Timeout Threshold | 120 seconds | Covers query time across device logs and test reports, prevents execution interruptions |
| Re-ranking | Sort by matching field weight | Prioritizes matching core fields such as insulation distance and partial discharge value to improve review accuracy |
| nodeId | Bind cross-database query node | Integrates data sources from test stations, oil assays, and device logs, covering all document types |
Three Common Mistakes
- Phenomenon: When using a channel trigger to start a workflow, the system returns "undefined" is not valid JSON. Cause: The product_id parameter passed by the channel was not configured in variables, leading to variable parsing failure.
- Phenomenon: The number of results returned by workflow searches exceeds expectations. Cause: The retrieval field scope was not limited, and non-core document data was mixed in, interfering with review logic.
- Phenomenon: Workflow execution times out and interrupts. Cause: A reasonable timeout threshold was not set, and cross-multiple data source query time exceeded the default limit.
How to Confirm Successful Configuration
- Manually pass a preset product ID to trigger the workflow. Check if insulating oil reports, test reports, and other documents for the corresponding product can be retrieved.
- Review the workflow's variable configuration. Confirm that parameters such as product_id have been correctly bound to the trigger source or input items.
- Simulate a scenario where the partial discharge value exceeds the threshold. Check if the preset exception alert branch is triggered.
- Verify the cross-data source query node configuration. Confirm that all required document source types have been covered.
The values provided are common starting points and should be measured against the reader's own samples.