TS_Predictor

A user-friendly GNSS Time Series Prediction toolbox combining Deep Learning and Signal Decomposition

Overview

TS_Predictor is an open-source MATLAB toolbox for predicting Global Navigation Satellite System (GNSS) coordinate time series. It integrates signal decomposition methods with machine learning and deep learning models, providing a complete workflow from raw data preprocessing to prediction accuracy assessment.

Key Features

• Data Import: Supports multiple formats from NGL, PBO, and CEA data centers
• Preprocessing: Missing data interpolation, outlier detection, offset correction, CME removal
• Signal Decomposition: EMD, EEMD, CEEMD, CEEMDAN, ICEEMDAN, VMD
• Prediction Models: SVM, RBF, ELM, BP, LSTM, GRU
• Accuracy Evaluation: RMSE, MAE, sMAPE, R², and unified WQE index
• Visualization: Interactive plots and exportable results

Installation

Requirements

• MATLAB 2024b or later
• Deep Learning Toolbox
• Signal Processing Toolbox
• Statistics and Machine Learning Toolbox

Steps

1. Download or clone this repository:

   git clone https://github.com/SpaceGeodesyLab/TS_Predictor.git

2. Open MATLAB and navigate to the TS_Predictor directory

3. Right-click on TS_predictor.mlappinstall and select Install

4. After installation, TS_Predictor will appear in the Apps toolbar

5. Click the TS_Predictor icon to launch the program

Quick Start

1. Load Data

TS_Predictor accepts GNSS time series from:

• Nevada Geodetic Laboratory (NGL): .txt format
• Plate Boundary Observatory (PBO): .csv and .pos formats
• China Earthquake Administration (CEA): .txt format

Use the Data Conversion module to convert your data to the unified CSV format.

2. Preprocess

The preprocessing module (based on gCMEbox) provides:

• Interpolation: Kriging-Kalman filtering for missing data
• Outlier detection: IQR and MAD methods
• Offset correction: Weighted least-squares fitting
• CME removal: PCA, ICA, FastICA, vbICA

3. Decompose

Select a decomposition method to separate your time series into components:

Method	Description	Best for
EMD	Empirical Mode Decomposition	Basic separation
EEMD	Ensemble EMD	Reducing mode mixing
CEEMDAN	Complete Ensemble EMD with Adaptive Noise	Cleaner separation
ICEEMDAN	Improved CEEMDAN	Best overall performance
VMD	Variational Mode Decomposition	Known number of modes

4. Predict

Choose a prediction model:

Model	Type	Characteristics
SVM	Machine Learning	Good for small datasets
RBF	Machine Learning	Fast training
ELM	Machine Learning	Very fast, less accurate
BP	Neural Network	Classic approach
LSTM	Deep Learning	Best for long sequences
GRU	Deep Learning	Good balance of speed/accuracy

Recommended: ICEEMDAN + GRU for most GNSS applications.

5. Evaluate

Compare models using multiple metrics:

• RMSE: Root Mean Square Error
• MAE: Mean Absolute Error
• sMAPE: Symmetric Mean Absolute Percentage Error
• R²: Coefficient of determination
• WQE: Weighted Quality Evaluation (unified metric)

Results are exported to Excel for further analysis.

Example

Test data from 10 GNSS stations in Yunnan, China (2010-2025) is provided in the Test data folder.

% Example workflow
% 1. Launch TS_Predictor from Apps toolbar
% 2. Load test data from 'Test data/YNTC.csv'
% 3. Apply preprocessing (interpolation, outlier removal)
% 4. Decompose using ICEEMDAN (default parameters)
% 5. Predict using GRU model
% 6. Evaluate and export results

Project Structure

TS_Predictor/
├── main/                    # Core application files
├── gCMEbox_v5/             # Preprocessing toolbox
├── Test data/              # Example GNSS time series
├── TS_predictor.mlappinstall   # Installation package
├── LICENSE                 # MIT License
└── README.md              # This file

Citation

If you use TS_Predictor in your research, please cite:

@article{he2025ts_predictor,
  title={TS_Predictor: An Open-Source Toolbox for GNSS Time Series Prediction Combining Deep Learning and Signal Decomposition},
  author={He, Xiaoxing and Zhou, Yu and Li, Jun and Kermarrec, Ga{\"e}l and Fernandes, Rui and Montillet, Jean-Philippe},
  journal={SoftwareX},
  year={2025},
  note={Submitted}
}

Related Software

• gCMEbox - GNSS Common Mode Error extraction toolbox
• Hector - GNSS time series noise analysis
• TSAnalyzer - GNSS time series analysis

Authors

• Xiaoxing He - Jiangxi University of Science and Technology - xxh@jxust.edu.cn
• Yu Zhou - Jiangxi University of Science and Technology
• Jun Li - Chang'an University
• Gaël Kermarrec - Leibniz University Hannover
• Rui Fernandes - University of Beira Interior
• Jean-Philippe Montillet - University of Beira Interior

Acknowledgments

This work was supported by:

• National Natural Science Foundation China (42364002, 42474028)
• Natural Science Foundation of Jiangxi (20252BAC220015, 20252BAC200264)

License

This project is licensed under the MIT License - see the LICENSE file for details.

Contributing

Contributions are welcome! Please feel free to submit a Pull Request. For major changes, please open an issue first to discuss what you would like to change.

Support

• Issues: GitHub Issues
• Email: xxh@jxust.edu.cn

---

SpaceGeodesyLab | Jiangxi University of Science and Technology