Thermal Modeling and Calibration Method in Complex Temperature Field for Single-Axis Rotational Inertial Navigation...

Area of Science:

• Navigation Systems Engineering
• Inertial Navigation Technology
• Sensor Calibration Techniques

Background:

• Single-axis rotational inertial navigation systems (RINS) rely on stable inertial measurement unit (IMU) biases for high accuracy.
• Temperature fluctuations, from environmental conditions and internal heating, destabilize IMU biases in RINS.
• Existing thermal calibration models inadequately represent the complex temperature fields within RINS.

Purpose of the Study:

• To develop advanced thermal calibration methods for single-axis RINS.
• To improve the precision and reliability of RINS navigation by accurately calibrating IMU biases.
• To address the limitations of traditional calibration models in capturing complex temperature effects.

Main Methods:

• Proposed a multiple regression method incorporating temperature gradients to model RINS temperature fields.
• Introduced a BP neural network approach considering coupled temperature variables for comprehensive thermal field description.
• Conducted experimental validation under laboratory and moving vehicle conditions.

Main Results:

• The proposed multiple regression and BP neural network methods significantly outperformed traditional thermal calibration.
• Navigation accuracy improvements of up to 47.41% were achieved in lab conditions.
• Navigation accuracy improvements of up to 65.11% were demonstrated in moving vehicle experiments.

Conclusions:

• Advanced thermal calibration methods, including multiple regression with temperature gradients and BP neural networks, effectively improve single-axis RINS precision.
• Accurate calibration of temperature-induced IMU bias variations is critical for enhancing RINS navigation performance.
• The developed methods offer a substantial advancement over traditional approaches for RINS thermal calibration.