This page is compatible with FastGPT v4.17.0.
What the Data for This Category Looks Like
Data for petroleum refining scenarios comes from DCS/APC control systems, LIMS laboratory information management systems, tank farm metering systems, crude oil evaluation databases, equipment management systems, and energy management systems. Daily documents include operating procedures, production plans, material balance sheets, distillate export quality reports, interlock ledgers, shift handover records, and change management documents. Data update rhythms cover real-time, per-shift, and daily frequencies. Fields include equipment tag numbers, unit numbers, inspection cycles, maintenance work order numbers, crude oil property parameters, and process fluctuation thresholds. Units include megapascals (MPa) for pressure, degrees Celsius (°C) for temperature, cubic meters per hour (m³/h) for flow, and other standard units.
Constraints Imposed on HTTP Interfaces and External Systems by These Characteristics
Multiple continuous units couple via material and hydrogen energy networks. Interfaces must support batch retrieval of real-time process data across units to avoid business disruptions from single-point request delays. Data update frequencies span a wide range. Interfaces must support data retrieval at three granularities: real-time, per shift, and daily to meet query needs for different business scenarios. Document types are diverse and fields are complex. Interfaces must support different request formats for structured ledgers and unstructured procedure texts. Data related to equipment inspection and maintenance links multiple systems. Interfaces must support cross-system data aggregation requests. They must also align request frequencies with the concurrency limits of each system to avoid triggering rate limiting rules.
How to Set Configurations
| Configuration Item | Recommended Setting | Rationale |
|---|---|---|
Authorization | Bearer {enterprise-specific API key} | Complies with HTTP interface authentication standards, and adapts to permission control requirements for multi-system docking in petroleum refining enterprises |
variables | {"deviceNo": "equipment tag number", "shift": "current shift"} | Matches business parameters for equipment inspection and maintenance Q&A in petroleum refining scenarios, and accurately associates historical data for the corresponding equipment |
responseData | Only return content, deviceId, updateTime fields | Filters redundant data, reduces interface transmission load, and adapts to efficiency requirements for high-frequency data requests in petroleum refining scenarios |
| API request timeout | 600 seconds | Cross-system data aggregation requests in petroleum refining scenarios may take a long time, preventing task failure due to mid-request disconnections |
| Batch request limit | Top 20 entries | Matches the conventional batch retrieval scale of shift inspection records in petroleum refining enterprises, avoiding exceeding system concurrency limits |
| Q&A context window | 8000–12000 token | Adapts to Q&A requirements for splicing multiple documents in petroleum refining scenarios, covering multiple text types such as operating procedures and maintenance records |
Three Common Configuration Mistakes
- Phenomenon: An "incorrect parameter" error returns when calling the
cloud.fastgpt.cnAPI to upload documents. Cause: Metadata fields for petroleum refining scenario documents are not configured correctly, such as failing to add business parameters like deviceNo and shift, resulting in interface verification failure. - Phenomenon: The Q&A result returned by the interface callback does not include the thinking process content of deepseek. Cause: The return field containing the thinking process is not specified in the
responseDataconfiguration, or the thinking process output setting for the corresponding model is not enabled. - Phenomenon: When proxyAI and one-api coexist, calling the equipment inspection interface returns
404 body not found. Cause: The dedicated petroleum refining equipment inspection interface path is not separately configured in proxy forwarding rules, resulting in requests being incorrectly forwarded to a mismatched interface pool.
How to Confirm Proper Configuration
- Initiate a request to the authentication interface using the configured
Authorizationparameter, and confirm that a 200 status code is returned along with a valid session identifier. - Submit an equipment inspection Q&A request with the
variablesparameter, and confirm that the returned result associates historical maintenance records for the corresponding equipment tag number. - Review the
responseDatafield returned by the interface, and confirm that only preset valid business fields are included with no redundant content. - Simulate a cross-system data retrieval request, and confirm that the interface returns results matching the data update frequency of petroleum refining scenarios within the configured timeout period.
The values provided here are common starting points, and should be evaluated against the reader's own samples.