Multi-mode Composite Guidance Data Fusion Algorithm Based on Optimized Convex Combination Theory

To improve composite guidance accuracy in complex interference environments, this paper proposes a guidance information fusion algorithm PR-DWF (Polynomial Regression - Dynamic Weighted Fusion) based on an optimized convex combination for multimode guidance information fusion. After time registration and outlier rejection, the algorithm improves the polynomial regression analysis prediction method to alleviate overfitting under small samples and performs dynamic weighted fusion output based on the deviation of the estimated value from the observed value. In this study, a set of captive-carry flight experimental platform is constructed based on the XX multimode composite seeker. The experimental results show that the algorithm has certain robustness and significantly improves tracking accuracy, the integrated tracking bias index improved by 38.38% over the cross-covariance fusion method, 9.93% over the Bar-Shalom-Campo fusion method, and 27.26% over the MAP-based fusion method.