A telemetry data based diagnostic health monitoring strategy for in-orbit spacecrafts with component degradation

Abstract

Diagnostic health monitoring without prior knowledge is still a hard problem in the prognostic and health management field. A multivariate diagnostic health monitoring strategy is proposed based on telemetry data for in-orbit spacecrafts with component degradation. Compared with the existing univariate or direct diagnostic health monitoring methods, multivariate diagnostic health monitoring methods can avoid constructing one-dimensional synthesized health index and setting empirical thresholds for different health states. In our developed strategy, a deep forest algorithm combined with an effective feature extraction approach and fuzzy C-means clustering algorithm is proposed to achieve more accurate assessment of the current health state. First, a partitioning window is utilized to deal with the raw telemetry data and then features which have high monotonicity and trends are extracted for diagnostic health monitoring. Then, a fuzzy Cmeans algorithm is used to handle unlabeled telemetry data and determine states of degrading component. Finally, a deep forest classifier is adopted to obtain the prognostic model for online probabilistic diagnostic health monitoring. Verification results on a simulated spacecraft attitude control system can demonstrate the effectiveness and feasibility of the proposed multivariate diagnostic health monitoring strategy.

Similarity-based probabilistic remaining useful life estimation for an aeroengine under variable operational conditions

“…Based on the centers of six operational conditions, each testing sample data utilizes equation ( 14) to match its corresponding operational condition and is normalized by equation (15). Then, the normalized testing data X′ (33991 × 5) uses the parameters (α, β) of the SHI library to construct SHIs Y ′ (I ′ × 1 × K ′ ) according to equation (16).…”

“…Their degradation states can be briefly described as a single-dimensional quantity utilizing one extracted effective feature; subsequently, their corresponding RULs are estimated accurately. In contrast, an engineering system contains multiple interdependent components, and no individual obvious degraded component can be found due to the closed-loop compensation [9,15]. Furthermore, complex and nonlinear relationships are presented between the degradation states of an engineering system and the end of life (EOL) [16][17][18][19].…”

Systemic symptom detection in telemetry of ISS with explainability using FRAM and SpecTRM

“…Especially machine-learning based anomaly detection methods are widely used. As an example of application in aerospace, Wang et al (2019) proposed diagnostic health monitoring for in-orbit spacecrafts. Whereas such methods provide high accuracy for anomaly detection, explainability for operators is lacking.…”

Research on Performance Evaluation Method of Communication System Based on FAHP-LSTM

“…System performance evaluation collects various types of information status, conducts learning analysis and obtains operating rules to ensure the reliable operation of the system [2] . In reference [3], an improved deep forest algorithm is proposed to evaluate the current health status of orbital spacecraft. In reference [4], data clustering analysis is used to classify and identify abnormal data of power equipment and identify the risk of power grid equipment.…”