A Gaussian-Shaped Fuzzy Inference System for Multi-Source Fuzzy Data

Fuzzy control theory has been extensively used in the construction of complex fuzzy inference systems. However, we argue that existing technologies focus mainly on single-source information system, disregarding complementary nature multi-source data. In this paper, develop a novel Gaussian-shaped Inference System (GFIS) driven by To end, first propose an interval-value normalization method to address heterogeneity The contribution our involves mapping heterogeneous data unified distribution space adjusting mean and variance from each source. As result combining normalized descriptions various sources for object, can obtain fused representation object. We then derive adaptive membership function based addition law Gaussian distribution. GFIS uses it dynamically granulate fusion inputs design rules. This proposed advantage being able adapt changing sources. Finally, integrate Takagi–Sugeno (T–S) model present modified framework. Applying methodology four datasets, confirm do lend support implying improved performance effectiveness.