What Data Looks Like for This Category
Core data for petroleum refining scenarios comes from DCS/APC control systems, LIMS laboratory management systems, tank farm metering systems, crude oil evaluation reports, equipment management systems, and energy management systems. Daily documents cover operating procedures, production plans, material balance sheets, distillate export quality reports, interlock ledgers, shift handover records, and change management documents.
Data update frequencies include real-time, shift-based, and daily updates. Fields include crude oil distillation range, unit load, product distillate indicators, tank farm liquid levels, and some fields have dedicated units. For example, distillation range is measured in degrees Celsius, and material balance is measured in tons.
Constraints Imposed on Workflow Orchestration
The differing update rhythms of multi-source heterogeneous data require workflows to support mixed trigger rules, balancing real-time plant data and daily updated crude oil evaluation data. Multiple continuous production units are coupled through logistics and energy networks. Workflows must connect multiple data pull nodes to ensure consistent parameter linkage across all stages.
There are diverse document types and many specialized fields. Workflows must adapt to the parsing logic of different documents to accurately extract core indicators required for tendering. Data anomalies may trigger severe production risks. Workflows must be configured with clear failure handling and alert links to ensure the stability of the preparation process.
Configuration Settings
This configuration is based on FastGPT v4.17.0.
| Configuration Item | Recommended Setting | Rationale |
|---|---|---|
| Trigger Mode | Real-time data update trigger + daily scheduled trigger | Matches the real-time update rhythm of plant operating data and the daily update rhythm of crude oil evaluation reports |
variables | Variable group containing crude oil properties, unit load, and product quality indicators | Provides core business parameters for dynamic insertion into tender documents and quotations |
| Node Timeout Setting | 600 seconds | Covers the time required for multi-unit data pulling and complex material balance calculations |
| Knowledge Base Retrieval Tag Filtering | Enabled, matched by document type | Accurately retrieves specialized documents such as operating procedures and material balance sheets, avoiding interference from irrelevant content |
| Failure Retry Count | 2–3 times | Addresses temporary network fluctuations or short system exceptions during real-time data pulling |
nodeId | Bound to core data nodes and quotation generation nodes | Ensures the accuracy of parameter transmission and the integrity of the link in the workflow |
Three Common Mistakes
- Symptom: Two thinking process outputs appear during workflow debugging. Cause: The automatic follow-up configuration of the tool call node is not disabled, causing an additional thinking link to be triggered after tool calls.
- Symptom: API calls cannot obtain MCP tool call records during workflow execution. Cause: The tool call log collection switch is not enabled in the workflow configuration, resulting in no logs being generated.
- Symptom: Code running component reports an error, prompting abnormal input parameter format. Cause: Specialized fields in the refining scenario (such as distillation range data) are not converted to a unified format, causing the component to fail to recognize them.
How to Verify Successful Configuration
- Trigger the preset trigger rules to verify whether the workflow starts as expected.
- Check the documents and quotation content output by the workflow to confirm whether core parameters for the current business scenario are included.
- Simulate a data anomaly scenario to verify whether failure retry and alert triggering work normally.
- Call the specified API interface to verify whether tool call logs during workflow execution can be obtained.
The values provided here are common starting points and should be measured against the reader's own samples.