A SELF-ADAPTIVE GENERIC IMM DATA FUSION ALGORITHM

• Published in: Natsional'nyi Hirnychyi Universytet. Naukovyi Visnyk no. 1; p. 122
• Main Authors: Yi, Yingmin, Chen, Weiduo
• Format: Journal Article
• Language: English
• Published: Dnipropetrosk State Higher Educational Institution "National Mining University" 01.01.2016
• Subjects: Algorithms; Computer simulation; Data fusion; Data integration; Equivalence; Mathematical models; Optimization; Similarity
• ISSN: 2071-2227; 2223-2362

For the problem of hybrid estimation, this paper proposes the self-adaptive generic interacting multiple-model (IMM) data fusion algorithm for solving the model selection problem of IMM. To find the optimal solution of the hybrid estimation problem, the history information of all the models was considered. According to the prior knowledge, the parameter space describing the model is mapped to the model set. According to the similarity of the parameter variations, the parameter space is divided into several sub-spaces. The center model of every sub-space was calculated out self adaptively. The center models were organized as the model set of the IMM algorithm. The final output of the algorithm is the data fusion of the model set estimations using IMM algorithm. At last, the simulation experiments showed that the proposed algorithm is superior to the traditional IMM algorithms under the condition of equivalent computation quantity.