Air-cooled inverter failure prediction decision assistance system
By constructing a multi-dimensional sensing module and a hybrid intelligent prediction module, the problem of multi-factor coupling characteristics of electrical-heating-environmental factors in air-cooled frequency converters was solved, achieving high-precision fault prediction and intelligent decision-making, and improving operation and maintenance response efficiency.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- DATANG HEBEI NEW ENERGY ZHANGBEI
- Filing Date
- 2026-02-06
- Publication Date
- 2026-06-09
AI Technical Summary
Existing frequency converter fault monitoring systems fail to fully consider the multi-factor coupling characteristics of air-cooled frequency converters, including electrical, heat dissipation, and environmental factors, resulting in low prediction accuracy, a lack of intelligent decision-making capabilities, and low operation and maintenance response efficiency.
A multi-dimensional sensing module is constructed to collect data, which is then fused using weighted principal component analysis. This is combined with a hybrid intelligent prediction module for hierarchical prediction, generating fault classification results and providing customized operation and maintenance solutions.
It improves the accuracy of predicting specific faults of air-cooled frequency converters and the efficiency of operation and maintenance response, thereby reducing operation and maintenance costs.
Smart Images

Figure CN122174146A_ABST