Odometer Velocity and Acceleration Estimation Based on Tracking Differentiator Filter for 3D-Reduced Inertial Sensor...

Area of Science:

• Robotics
• Sensor Fusion
• Navigation Systems

Background:

• Reduced Inertial Sensor Systems (RISS) rely heavily on odometer velocity, which is susceptible to noise.
• Conventional differentiation methods amplify noise, compromising navigation accuracy and reliability in 3D RISS.
• Accurate velocity and acceleration estimation are crucial for robust RISS performance.

Purpose of the Study:

• To propose and evaluate a tracking differentiator (TD) filter for improved velocity and acceleration tracking in 3D RISS.
• To enhance navigation accuracy and reliability by mitigating noise in odometer velocity data.
• To demonstrate the TD filter's effectiveness compared to traditional numerical differentiation.

Main Methods:

• Implementation of a tracking differentiator (TD) filter to estimate velocity and acceleration from odometer signals.
• Utilizing the TD filter's prediction process to reduce phase lag.
• Conducting numerical simulations and analyzing offline data from actual vehicle experiments.

Main Results:

• The TD filter accurately estimates velocity and acceleration, filtering out random noise and outliers.
• Numerical simulations show superior performance of the TD filter over traditional differentiation methods.
• Vehicle experiment data confirm the TD filter's effectiveness in a real-world 3D RISS context.

Conclusions:

• The TD filter effectively filters odometer velocity and estimates acceleration for 3D RISS.
• The proposed method significantly improves the accuracy and reliability of 3D RISS navigation.
• The TD filter offers a computationally efficient solution without requiring object models.