What the data for this category looks like
Energy storage-related data primarily comes from product development documents, grid connection compliance documents, marketing promotional materials, and industry technical white papers. Update cycles vary irregularly with new product launches and policy adjustments. Single-update document volumes range from single-page parameter sheets to dozens of pages of compliance manuals. Most documents exist in hierarchical markdown or PDF formats, and include fields such as product model, rated charge-discharge capacity, conversion efficiency, installation dimensions, and certification number. These fields come with standard units such as kWh, kW, %, and mm.
Constraints imposed by these characteristics on knowledge base retrieval and recall
Energy storage product categories have many parameter fields with units. Retrieval requires precise unit matching to avoid ambiguity. Hierarchical document structures require retention of subordinate relationships, otherwise recall by product category is not possible. Update frequencies are irregular, so regular synchronization of the latest data is required. Single-document volumes vary widely, so adaptation to different lengths of text parsing is needed. These characteristics mean general-purpose configurations cannot meet requirements, and parameters for parsing, recall, synchronization and other links require targeted adjustments.
How to set the configurations
| Configuration Item | Recommended Value Range | Rationale |
|---|---|---|
parse_chunk_size | 800–1200 characters | Energy storage documents contain long strings of parameters and hierarchical content. This segment length balances parameter completeness and retrieval accuracy |
recall_top_k | Top 8–12 results | Energy storage product parameter entries are numerous. Excessive recall leads to redundant results. This range covers core retrieval needs |
similarity_threshold | 0.72–0.85 | Energy storage parameter matching requires high precision. This threshold filters low-relevance non-target parameter content |
parse_keep_markdown_structure | Enabled | Most energy storage documents include hierarchical headings and classifications. Retaining structure prevents loss of subordinate logic |
refresh_cron | 0 0 */3 * * ? | Updates to new energy storage products and policies occur frequently. Refreshing every three days ensures data timeliness |
UPLOAD_FILE_MAX_SIZE | 1000 MB | Some energy storage compliance manuals and technical white papers have large volumes. This upper limit covers standard upload requirements |
The parameter values provided on this page are common starting points for configuration. Actual values are affected by material form, data volume and business rules. Specific issues require individual analysis. It is recommended to test on your own samples before finalizing settings.
Three common configuration mistakes
- Phenomenon: After importing energy storage markdown documents, heading levels are lost, and sub-model content cannot be retrieved by product category. Cause: The parsing configuration to retain markdown structure is not enabled. Subordinate relationships between headings and content are discarded during segmentation.
- Phenomenon: Extra spaces appear between numbers and units of energy storage parameters in retrieval results, while no spaces are present during knowledge base entry. Cause: The automatic formatting switch is enabled by default during parsing, which adds spaces to numbers and adjacent characters.
- Phenomenon: After uploading an Excel-format energy storage product parameter sheet, extracted fields are empty or have chaotic formatting. Cause: The knowledge base's Excel parsing configuration is not enabled, or field extraction mapping rules are not configured.
How to confirm configurations are correct
- Upload an energy storage markdown document with hierarchical headings. Check if parsed segments retain the subordinate relationship between headings and corresponding content.
- Enter energy storage parameter keywords. Check if the format of numbers and units in retrieval results matches the format used during knowledge base entry.
- Upload an Excel-format energy storage parameter sheet. Check if preset fields such as product model and rated capacity are correctly extracted.
- Manually trigger a knowledge base refresh. Check if new product parameters after the update can be retrieved normally.
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.