交互多模型与简单凸组合下的多传感器融合算法OA

Aiming at the problems of low tracking accuracy and poor robustness in single-sensor systems,a multi-sensor information fusion algorithms is investigated to improve the performance of maneuvering target state estimation.Segmented motion models,namely,the constant velocity(CV),constant acceleration(C A),and constant turning(CT)models,are adopted to simulate target maneuvers.By integrating the interacting multiple model(IMM)algorithm with Kalman filtering(KF),nonlinear estimation under a single sensor is achieved.Fur-thermore,the simple convex combination(SCC)algorithm is employed to fuse the local estimation results from multiple sensors,with a focus on addressing challenges such as adaptive model matching and consistent fusion of multi-source information.Simulation results demonstrate that the proposed algorithm significantly enhances estimation accuracy,tracking stability,and system robustness.By combining segmented multi-model modeling with IMM adaptive filtering,the characterization of complex maneuvers is improved,and the SCC fusion strategy is adopted to reduce computational complexity while maintaining accuracy,thereby providing an effective solu-tion for engineering applications.