Skip to content

ADAPTIVE_LN_SIBLING_DISCOVERY_FIX.md

Latest commit

 

History

History
27 lines (19 loc) · 1.62 KB

ADAPTIVE_LN_SIBLING_DISCOVERY_FIX.md

File metadata and controls

27 lines (19 loc) · 1.62 KB
Raw

Adaptive LN Sibling Discovery Fix

This patch fixes cases where a live IED exposes only the first logical-node instance through the current ARIEC61850 model builder, while commercial browsers such as IEDScout can browse additional instances such as MMXU3, MMXU4, MMXU5, and so on.

What changed

  1. Added adaptive logical-node sibling proof-probing.

    • The engine no longer hardcodes MMXU1..MMXU24.
    • It observes real discovered logical-node patterns such as MMXU1.
    • It then probes adjacent IEC 61850 logical-node instances using real MMS reads.
    • A sibling instance is added only when the IED proves it exists by returning at least one readable value.
    • The probe stops after consecutive misses, so the runtime does not blindly generate endless fake signals.

  2. Added VAA sibling expansion in the clean-room MMS browser.

    • If GetNameList returns a shallow root like MMXU1, the discovery engine can probe MMXU2, MMXU3, etc. using GetVariableAccessAttributes and expand the actual type tree when the IED accepts the object.

  3. Improved MMS item parser.

    • Discovery artifacts can now be parsed whether they use $ or . separators.
    • Examples normalized correctly:
      • MMXU3$MX$PhV$phsA$cVal$mag$f
      • MMXU3.MX.PhV.phsA.cVal.mag.f
      • GD7700LD01/MMXU3$MX$PhV$phsA$cVal$mag$f

Why this is not hardcoding

The engine does not assume that a Moxa IED has 24 MMXU nodes. It only starts adaptive sibling discovery from logical-node classes and instance patterns that were actually discovered from the IED, then confirms every new instance with a real MMS read or GetVariableAccessAttributes response.