Fire-fighting equipment whole life cycle predictive operation and maintenance method and system

By constructing a context importance assessment mechanism and key vector similarity neighborhood clustering, and combining it with a summary generation network to optimize KVCache management, the problems of cache expansion and prediction stability in multimodal long sequence data processing are solved, thereby improving the accuracy and efficiency of fire protection facility operation and maintenance.

CN122390725APending Publication Date: 2026-07-14COHEN THINK TANK FIREZHEJIANG CO
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
COHEN THINK TANK FIREZHEJIANG CO
Filing Date
2026-06-11
Publication Date
2026-07-14

AI Technical Summary

Technical Problem

Existing methods for the operation and maintenance of fire protection facilities are difficult to effectively process multimodal long-sequence data, which limits the accuracy of predictions of equipment failure probability, remaining lifespan and maintenance level. Furthermore, KVCache bloat can easily lead to the loss of key degradation clues, affecting prediction stability.

Method used

By constructing a contextual importance evaluation mechanism that integrates modality weights, knowledge graph entity centrality, and cumulative attention values, and combining it with a clustering method based on key vector similar neighborhoods, a summary generation network is used to perform semantic compression of redundant information clusters and evict isolated information points, thereby optimizing KVCache management.

Benefits of technology

It improves the stability and accuracy of predicting the failure probability, remaining lifespan, and maintenance level of fire protection facilities, reduces cache usage and inference latency, and enhances the operational efficiency of long-cycle operation and maintenance prediction tasks.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122390725A_ABST
    Figure CN122390725A_ABST
Patent Text Reader

Abstract

The application provides a fire-fighting facility whole life cycle predictive operation and maintenance method and system, the method comprises collecting multi-modal operation and maintenance data such as sensor time series data, maintenance text log and structured event alarm, inputting the unified coding into the autoregressive prediction model, and storing the context key-value pair at each time step in KVCache during the inference process. In view of the problem that the cache capacity continues to grow with long sequence inference, the importance score of the key-value pair is calculated by fusing the modal weight, the entity centrality of the operation and maintenance knowledge graph and the cumulative attention value at each inference step; when the cache capacity reaches the threshold, based on the similar neighborhood clustering of the key vector, the redundant information cluster and the isolated information point with low importance are identified, the redundant cluster is compressed and the isolated point is expelled by using the abstract generation network, the cache optimization is completed, and then the failure probability, the remaining life and / or the maintenance level of the fire-fighting facility in the future period are predicted.
Need to check novelty before this filing date? Find Prior Art