Abstract :

The Increasing Dependence On Space-based Systems, Supported By Thousands Of Active Satellites And Significant Economic Value, Has Made Operational Reliability A Critical Concern In Aerospace Applications. A Large Proportion Of Mission Failures Can Be Attributed To Unnoticed Irregularities In Telemetry Streams, Creating A Strong Need For Intelligent Systems Capable Of Accurate Prediction And Early Anomaly Detection. Telemetry Data Provides Detailed Insights Into Spacecraft Subsystems And Operational Conditions; However, Its High Dimensionality, Presence Of Noise, Missing Values, And Complex Variability Make Analysis Challenging For Traditional Techniques. To Address These Issues, This Study Introduces A Scalable Framework For Telemetry Analysis Using A Locally Implemented Ensemble Learning Approach Based On OPSSATAD Data. The System Integrates Multiple Learning Models Within A Unified Structure Inspired By Classification And Regression Tree (CART) Principles, Including Random Forest (RF-CART), Support Vector Machine (SVMCART), And Gradient Boosting (GB-CART) For Both Classification And Prediction Tasks. Furthermore, A Hybrid Ensemble Model Combining Extra Trees (ET-CART) And CatBoost (CatBoost-CART) Is Proposed Using A Voting-based Aggregation Mechanism To Enhance Predictive Performance. The ET Component Captures Hierarchical Decision Patterns, While CatBoost Improves Learning Through Efficient Handling Of Feature Interactions And Categorical Data. A Structured Pipeline Involving Preprocessing, Feature Refinement, Training, And Evaluation Is Implemented Using Performance Metrics Such As Accuracy, Precision, Recall, F1- Score, MAE, MSE, RMSE, And R². The Results Indicate That The Ensemble Approach Consistently Achieves Higher Accuracy And Robustness Compared To Individual Models, Making It Effective For Reliable Mission Prediction And Analysis.

Published:

24-4-2026

Issue:

Vol. 26 No. 4 (2026)

Page Nos:

3001-3013

Section:

Articles

License:

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

How to Cite