What the data for this category looks like
Data sources for photovoltaic intelligent due diligence reports include structured operation logs from power station operation and maintenance platforms, compliance test reports from third-party testing institutions, grid connection settlement documents from power grid companies, and factory parameter documents from component manufacturers. There are two types of data update schedules: basic records and factory parameters are statically updated and delivered along with project progress; operation data and settlement data are synchronized and updated daily. The document structure is fixed into four sections: power station basic information page, equipment parameter page, operation log page, and compliance verification page. Core fields include power station record number, component model, installed capacity, daily irradiation, monthly power generation, inverter operation parameters, etc. Installed capacity is measured in kilowatt peak (kWp), irradiation is measured in kilowatt hours per square meter (kWh/㎡), and power generation is measured in megawatt hours (MWh).
What constraints do these characteristics impose on the "workflow orchestration" link
The multi-source and heterogeneous nature of photovoltaic due diligence data requires the workflow to adapt to multiple input formats, including structured Excel, encrypted PDF reports, and plain text logs. Multiple nodes must be configured to handle different types of data sources separately. The daily updated operation data requires the workflow to be bound to a scheduled trigger mechanism to ensure that the latest operation data can be obtained each time it is executed. The fixed fields and units require a format verification link to be added to the workflow to avoid abnormal data with inconsistent units from affecting subsequent analysis and report generation. The fixed structure of compliance reports requires the workflow to extract content in the order of preset fields to avoid missing core compliance items such as power station record numbers and grid connection permit numbers.
How to set the configurations
| Configuration Item | Recommended Approach | Basis for This Approach |
|---|---|---|
PARSE_FILE_TIMEOUT_SECONDS | 600 seconds | Photovoltaic due diligence reports often contain multi-page equipment parameter logs, which take a long time to parse. 600 seconds covers the complete parsing process |
workflow_trigger_cron | 0 0 2 * * * | Photovoltaic operation data is synchronized every early morning. Triggering at this time allows access to the latest daily operation data |
max_extract_length | 8000–12000 characters | The text volume of the equipment parameter page of a single photovoltaic power station due diligence report is large. This range can fully extract core fields |
http_request_timeout | 300 seconds | The batch write interface of Feishu multidimensional tables takes a long time to respond. This duration avoids request interruptions mid-execution |
variable_validation_rule | {"unit":["kWp","kWh/㎡","MWh"]} | Photovoltaic due diligence data has fixed unit requirements, which can filter abnormal inputs with inconsistent units |
The parameter values provided on this page are all conventional recommendations used to determine the starting point for configuration. Actual values are affected by material form, data volume, and business rules. Specific issues require specific analysis. It is recommended to test on your own samples before finalizing the configuration.
Three common mistakes
- Symptom: The HTTP node returns a 400 status code when calling Feishu multidimensional tables, and the written fields are empty. Cause: The Authorization parameter in the request header is not configured correctly, and the Feishu API key stored in the environment variable is not correctly substituted into the request header.
- Symptom: Variables output by workflow nodes cannot recognize space-separated content in subsequent prompts. Cause: No format processing is added to variable references in the prompt, causing spaces to be automatically compressed and unable to match preset field splitting rules.
- Symptom: The workflow stalls after reaching the file parsing node with no error message. Cause: A reasonable value for
PARSE_FILE_TIMEOUT_SECONDSis not configured. After the photovoltaic due diligence report parsing times out, no retry mechanism is triggered, and the workflow enters a suspended state.
How to confirm the configuration is complete
- Manually trigger the workflow once, view the output logs of each node, and check whether the extracted fields include the core parameters required for photovoltaic due diligence.
- Check the Cron expression of the scheduled trigger node to confirm that it matches the update schedule of the data source.
- Call the test interface of the Feishu API to verify that the configured key and request parameters can normally write data to the multidimensional table.
- Simulate abnormal data input with inconsistent units, and check whether the variable verification node can filter this type of content.
Question material comes from public community discussions. Configuration values are common starting points and should be measured against your own samples. Verified on 2026-09-14.