Enhancement of the HILOMOT Algorithm with Modified EM and Modified PSO Algorithms for Nonlinear Systems Identification

Developing a mathematical model has become an inevitable need in studies of all disciplines. With advancements technology, there is emerging to develop complex models. System identification popular way constructing models highly processes when analytical not feasible. One the many architectures system utilize Local Model Network (LMN). Hierarchical Tree (HILOMOT) iterative LMN training algorithm that uses axis-oblique split method divide input space hierarchically. The positions local directly influence accuracy entire model. However, finding best presents nonlinear optimization problem. This paper optimized HILOMOT with enhanced Expectation–Maximization (EM) and Particle Swarm Optimization (PSO) algorithms which includes normalization parameter utilizes reduced-parameter vector. Finally, performance improved compared existing by modeling NOx emission gas turbine multiple test functions different orders structures.