An multi-platform systematic error registration method based on Multi-criteria Adaptive Cubature Kalman Filter with Singer model

Systematic error registration is a critical issue in multi-platform, multi-sensor target tracking. For noncooperative targets moving under the Singer model, this paper proposes a systematic error registration algorithm based on a multi-criteria adaptive cubature Kalman filter. By comparing the innovation obtained during the measurement update process with different decision thresholds, the algorithm selects appropriate compensation corrections to improve estimation accuracy and convergence speed. Numerical experiments reveal that the proposed methodology achieves notable improvements in parameter estimation fidelity and accelerated convergence rates compared to conventional CKF-based error registration approaches, demonstrating effective real-time compensation capabilities for systematic biases in multi-platform, multi-sensor architectures. The enhanced performance metrics confirm the algorithm’s practical applicability in operational scenarios involving distributed sensor networks.