Ground Moving Target Tracking Filter Considering Terrain and Kinematics.

Area of Science:

• Radar Systems
• Signal Processing
• Geospatial Analysis

Background:

• Ground target tracking (GTT) for airborne radar commonly utilizes Digital Terrain Elevation Data (DTED) as prior information.
• Existing methods often apply DTED as a hard constraint on target position only, neglecting velocity constraints.
• DTED limitations include lack of continuous and accurate terrain elevation measurements.

Purpose of the Study:

• To develop an improved GTT method that overcomes limitations of existing DTED fusion techniques.
• To introduce a novel soft terrain constraint that considers both target position and velocity.
• To enhance tracking accuracy in airborne radar systems.

Main Methods:

• Reconstruction of ground-truth terrain elevation using Gaussian processes, treating DTED as noisy observations.
• Formulation of terrain constraints as joint soft constraints on target position and velocity.
• Development of a Soft Terrain Constrained Particle Filter (STC-PF) incorporating these soft constraints.

Main Results:

• The proposed STC-PF effectively propagates particles while adhering to terrain constraints during the prediction step.
• Numerical simulations demonstrate superior tracking performance of STC-PF compared to the Smoothly Constrained Kalman Filter (SCKF).
• The improved performance is attributed to the STC-PF's ability to incorporate soft position and velocity constraints, unlike SCKF's hard constraints.

Conclusions:

• The novel soft terrain constraint and STC-PF offer a significant advancement in airborne radar GTT.
• The method effectively addresses the limitations of traditional DTED-based tracking filters.
• STC-PF provides more accurate and robust ground target tracking by leveraging kinematic terrain constraints.