Seamless Micro-Electro-Mechanical System-Inertial Navigation System/Polarization Compass Navigation Method with Data...

Area of Science:

• Robotics and Autonomous Systems
• Sensor Fusion
• Navigation Technology

Background:

• Micro-electro-mechanical system-inertial navigation systems (MEMS-INS) integrated with polarization compasses (PCs) are crucial for vehicle navigation in Internet of Things (IoT) applications.
• These integrated systems face challenges like cumulative errors and time-varying noise covariance in complex, dynamic, and occluded environments.
• Existing methods struggle to maintain accuracy and reliability under such adverse conditions.

Purpose of the Study:

• To propose a novel dual data- and model-driven seamless navigation method for MEMS-INS/PC integrated systems.
• To address the limitations of cumulative errors and time-varying measurement noise covariance.
• To enhance the overall performance, robustness, and accuracy of navigation in challenging environments.

Main Methods:

• A nonlinear autoregressive neural network (NARX) with Gauss-Newton Bayesian regularization trains to model MEMS-INS outputs and PC heading increments for data-driven operation.
• A nonlinear MEMS-INS/PC loosely coupled navigation model is established for the model-driven component.
• Variational Bayesian methods estimate time-varying measurement noise covariance, and the cubature Kalman filter solves the nonlinear navigation problem.

Main Results:

• The proposed dual data- and model-driven method demonstrates robustness and effectiveness in experimental verification.
• The system successfully models the dynamic characteristics of the integrated navigation system.
• High-precision heading information is stably provided even in complex, occluded, and dynamic environments.

Conclusions:

• The developed dual data- and model-driven approach significantly improves the performance of MEMS-INS/PC integrated navigation systems.
• The method effectively mitigates cumulative errors and handles time-varying noise covariance.
• This advancement offers a reliable solution for high-precision vehicle navigation in challenging real-world IoT applications.