A novel adaptive multi-sensor UKF anti-interference method of four-axis dual-band infrared attitude measurement for spinning projectile

Abstract

In infrared radiation attitude measurement of the spinning projectile, traditional two-axis infrared measurements are prone to failure when subjected to interference. And, due to the random occurrence of interference sources, traditional filters exhibit significant errors. To address these issues, a four-axis dual-band infrared radiation (FDBIR) attitude measurement device and an Adaptive Multi-sensor Unscented Kalman Filter (AM-UKF) algorithm are proposed. The device is capable of operating normally under interference within a 180-degree range. The sensor output curve distortion detection function and the dual-band interference intensity quantification function are employed to address the issue of random interference. What is more, Multi-channel covariance is utilized to automatically adjust or discard channels affected by interference. This allows the FDBIR measurement to resolve attitude angles under 180-degree interference coverage. Then the actual infrared radiation environment and simulated motion are used for semi-physical experiments. The results of experiments show that in the absence of interference, the AM-UKF achieves more than twice the accuracy compared to other recent algorithms. In the presence of interference, the AM-UKF can still maintain the roll angle error within $\pm$1.3 degrees , and the pitch angle error within $\pm$1.2 degrees, which is more than three times smaller than other algorithms. The proposed method enhances the accuracy and reliability of dual-band infrared rotational attitude measurement and provides technical references for rotational projectile attitude measurement.