一种实时在线超声速进气道不起动状态预警的方法
By automatically selecting sensors and generating classifiers using the CD-WPT-CNN algorithm, the problem of manual analysis in the early warning of the non-starting state of supersonic air intakes is solved, realizing real-time online early warning, adapting to different operating conditions and engines, and reducing workload and professional knowledge requirements.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- NANJING UNIV OF AERONAUTICS & ASTRONAUTICS
- Filing Date
- 2022-08-26
- Publication Date
- 2026-07-17
AI Technical Summary
Existing technologies rely on manual analysis and parameter setting for early warning of non-starting status of supersonic inlets, resulting in a large workload, high difficulty, and inability to adapt to different operating conditions. Furthermore, it is difficult to accurately identify the sensor position and signal fluctuation time, which affects the early warning effect.
The CD-WPT-CNN combined algorithm, which combines Cumulative Sum Change Detection (CUSUM), Wavelet Packet Transform (WPT), and Deep Learning Convolutional Neural Network (CNN), automatically filters sensors and generates classifiers to achieve real-time online early warning.
It reduces the difficulty and workload of early warning tasks, improves adaptability and accuracy, is applicable to different operating conditions and engines, reduces reliance on professional knowledge, and has real-time performance and feasibility.
Smart Images
Figure CN115408940B_ABST