Aero-engine fault detection method based on support vector data description and transfer learning
An aero-engine, data description technology, applied in engine testing, machine learning, machine/structural component testing, etc., can solve problems such as insufficient fault data and inability to train performance, so as to improve effectiveness and solve insufficient fault data. , the effect of good fault detection effect
Pending Publication Date: 2022-08-02
NANJING UNIV OF AERONAUTICS & ASTRONAUTICS
View PDF0 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0008] Purpose of the invention: In order to solve the problem that a fault detection model with good performance cannot be trained due to insufficient fault data when the aero-engine degrades to a certain stage, this invention adopts the idea of transfer learning and uses the data collected when it is not degraded as the source domain Data, using the data collected in the current degradation state as the target domain data, a SVDD-based aero-engine fault detection model is designed to improve the effectiveness of engine fault detection
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0031] An aero-engine fault detection method based on support vector data description and transfer learning of the present invention, such as figure 1 shown, including the following steps:
[0032] Step 1: Collect historical flight data as source domain data under the engine's nominal state;
[0033] The flight data includes multiple groups, and each group of flight data comes from 10 sensors, including: high-pressure rotor speed N H , the low-pressure rotor speed N L , the fan outlet temperature T 22 , the compressor outlet pressure T 3 , the compressor outlet pressure P 3 , the low pressure turbine inlet temperature T 45 , the low pressure turbine outlet temperature T 46 , the low pressure turbine outlet pressure P 46 , the mixing chamber inlet temperature T 65 , and the main fuel flow WFB.
[0034] The historical operation data of the nominal state collected includes fault data and non-fault data, wherein the fault data types include fan fault data, compressor faul...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention discloses an aero-engine fault detection method based on support vector data description and transfer learning, which is based on support vector data description and introduces the thought of transfer learning. According to the method, the difference of engine operation data under different degradation degrees in the degradation process of an aero-engine is considered, historical engine operation data serves as source domain data, operation data collected in the current state of the engine serves as target domain data, the sphere center of an SVDD model serves as a knowledge structure, and the SVDD model is migrated from a source domain to a target domain; and the target domain is assisted to establish a fault detection model with relatively high accuracy. According to the method, the problem of lack of fault data in the field of aero-engine fault detection is solved through transfer learning, and the fault detection effect of the fault detection model under the condition of less fault data is improved in combination with the single classification algorithm SVDD.
Description
technical field [0001] The invention belongs to the field of aero-engine fault detection, in particular to an aero-engine fault detection method based on support vector data description and migration learning. Background technique [0002] Aero-engine fault detection refers to judging the state of the engine, fault state or normal state according to the real-time operating data of the engine. Fault detection is an important part of aero-engine health management, which is of great significance to ensure the healthy operation of aircraft and engines. Since 90% of the failures of aero-engines are the faults of the air circuit components, it is very necessary to detect the faults of the air circuit components. Changes in flow and efficiency can directly reflect the health of engine components, but they cannot be directly observed. However, some measurable parameters on the engine, such as temperature, pressure, speed, etc., cannot directly reflect the health status of aero-eng...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More IPC IPC(8): G06K9/62G06N20/00G01M15/00
CPCG06N20/00G01M15/00G06F18/214G06F18/2411Y02T90/00
Inventor 赵永平陈耀斌
Owner NANJING UNIV OF AERONAUTICS & ASTRONAUTICS