A Reliable Deep Learning Approach for Time-Varying Faults Identification: Spacecraft Reaction Wheel Case Study
نویسندگان
چکیده
Reaction wheels are key components for the spacecraft attitude control subsystems. Faults in reaction may lead to high energy consumption, lack of control, and case failure, loss spacecraft. The accurate identification anomalies is a challenging task due internal nonlinearities wheels. This study proposes fast end-to-end architecture detecting identifying occurring using One-Dimensional Convolutional Neural Network (1D-CNN) with Long Short-Term Memory (LSTM) network architecture. 1D-CNN used capture useful features from raw residual signals. layer its effectiveness handling time series data capabilities learning long-term dependencies. proposed directly trained torque signals captured 3-axis subsystem simulation model. In this way, scheme eliminates need specific feature extraction method. Results showed that algorithm represents reliable robust anomaly detection mechanism compact system Furthermore, obtained results revealed superiority generalizability model diagnosing time-varying wheel faults over other recent approaches. Ultimately, approach considered be generic fault diagnosis safety-critical systems. dataset available download at: https://dx.doi.org/10.21227/jr1c-bm66.
منابع مشابه
A Finite Frequency Approach to Reliable H1 Filtering for Linear Continuous-Time Systems with Sensor Faults
The paper deals with the reliable H∞ filtering problem for linear uncertain continuous-time systems with bounded disturbances. Different from the existing approaches, the filter is designed in certain finite frequency ranges, which is important in practice, since the full frequency approaches are conservative to some extent for the case when the frequency ranges of disturbances are known before...
متن کاملA Deep Learning Approach for Norm Conflict Identification
Regulations are often applied to social members in a society in order to minimize conflicting behaviors [9]. Such regulations, also known as social norms, define expected behaviors for society members [7] and help ensure that individuals act in socially acceptable behavior. Besides regulating entire societies, social norms are also used to regulate interactions in smaller groups, and are often ...
متن کاملIdentification of Time-varying Modal Parameters Using a Time-varying Autoregressive Approach
A time-varying autoregressive model with time-varying coefficients is introduced in this paper for parameter extraction from non-stationary vibration signals. With this model, the relationship between linear time-varying modal parameters, i.e., instantaneous frequencies and damping factors, and time-varying autoregressive model coefficients is established. The time-varying autoregressive modeli...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: IEEE Access
سال: 2022
ISSN: ['2169-3536']
DOI: https://doi.org/10.1109/access.2022.3191331