This page configuration references FastGPT v4.17.0.
What this category of data looks like
Non-ferrous metal smelting ledger data comes from on-site industrial control systems, daily patrol records, material in-and-out warehouse documents, and quality inspection center test reports. There are three update frequencies: real-time process parameters update every 10 seconds, shift ledgers are finalized at 24:00 daily, and monthly comprehensive ledgers are generated by the 3rd day of the following month. Document formats include single-device CSV/Excel ledgers, multi-sheet workshop summary tables, and entire plant structured reports. Fields include shift output, smelting temperature, auxiliary material unit consumption, power consumption, and more. Some patrol documents are stored as image files, containing equipment numbers and working condition descriptions.
What constraints these characteristics impose on workflow orchestration
Multi-source heterogeneous data sources require workflows to support three types of access nodes: industrial control API, Excel upload, and image upload, to avoid data silos. High-frequency updates of real-time parameters require workflows to allow configuring trigger intervals from 10 seconds to 1 hour, and support event triggering to respond to real-time data pushes from industrial control systems. The three-level summary logic of ledgers requires workflows to support nested aggregation nodes, to integrate data step-by-step by production line, workshop, and entire plant. Image-format patrol documents require transcoding processing, so workflows need built-in custom script nodes to complete base64 conversion. Cross-cycle data requirements for monthly reports require workflows to allow configuring natural time range filter parameters, to ensure accurate data intervals.
How to set configurations
| Configuration Item | Recommended Setting | Rationale |
|---|---|---|
| Trigger Mode | Dual mode of event trigger + scheduled trigger, set scheduled interval to 10 seconds–1 hour | Matches the update rhythm of real-time process data, supports on-demand triggering and automatic summary |
| Data Source Access | Support three types of access: industrial control system API, Excel upload, image upload | Covers three data source types: single-device ledgers, workshop summary tables, and image patrol documents |
| Data Cleaning Rules | Filter missing value fields, unify parameter units | Ensures consistent data caliber for ledgers across devices and workshops |
| Transcoding Processing Node | Configure custom scripts to handle base64 conversion | Adapts to analysis requirements for image-format patrol documents |
| Summary Logic Configuration | Three-level nesting by production line → workshop → entire plant | Matches the ledger hierarchy structure of non-ferrous metal smelting |
variables | Preset global variables include knowledge base ID and target storage path | Simplifies parameter passing during API calls, aligns with global variable configuration requirements |
nodeId | Fill in according to the actual generated node ID in the workflow canvas | Ensures accurate targeting of specified workflow nodes during API calls |
Three Common Misconfigurations
- Symptom: Ledger data fields generated after workflow runs are missing, or units are inconsistent. Cause: Required field validation rules for data cleaning nodes are not configured, causing invalid data to be included in the summary process.
- Symptom: After uploading an image via the API-triggered workflow, the base64 encoding result is empty. Cause: The file reading logic in the transcoding node is not configured correctly, and the uploaded image file is not converted into an encodable binary stream.
- Symptom: The workflow returns a parameter error prompt when calling the knowledge base. Cause: The knowledge base ID is not correctly written to the
variablesglobal variable, causing the workflow to fail to read the preset knowledge base configuration.
How to Confirm Configuration is Correct
- Manually trigger the workflow, upload a single-device Excel ledger, and check whether the summarized workshop data matches the original data.
- Upload a site patrol image, check whether the base64 encoding output by the workflow is a valid string, which can be verified via an online decoding tool.
- Call the API with the preset
variablesparameters, check whether the workflow can correctly read the global variables and associate with the knowledge base. - Configure a scheduled trigger task, wait for one trigger cycle, and check whether the automatically generated monthly ledger includes all data within the specified time range.
The values given are common starting points and should be measured against the reader's own samples.
Question material comes from public community discussions. Configuration values are common starting points and should be measured against your own samples. Verified on 2026-09-28.