International Journal of Computer Applications
Foundation of Computer Science (FCS), NY, USA
|
Volume 187 - Issue 18 |
Published: July 2025 |
Authors: Thai-Hoang Huynh |
![]() |
Thai-Hoang Huynh . A Novel GA-based Fuzzy Extended State Observer for Fault Detection of Nonlinear Systems. International Journal of Computer Applications. 187, 18 (July 2025), 26-32. DOI=10.5120/ijca2025925243
@article{ 10.5120/ijca2025925243, author = { Thai-Hoang Huynh }, title = { A Novel GA-based Fuzzy Extended State Observer for Fault Detection of Nonlinear Systems }, journal = { International Journal of Computer Applications }, year = { 2025 }, volume = { 187 }, number = { 18 }, pages = { 26-32 }, doi = { 10.5120/ijca2025925243 }, publisher = { Foundation of Computer Science (FCS), NY, USA } }
%0 Journal Article %D 2025 %A Thai-Hoang Huynh %T A Novel GA-based Fuzzy Extended State Observer for Fault Detection of Nonlinear Systems%T %J International Journal of Computer Applications %V 187 %N 18 %P 26-32 %R 10.5120/ijca2025925243 %I Foundation of Computer Science (FCS), NY, USA
This paper presents a novel Genetic Algorithm based Fuzzy Extended State Observer (GA-FESO) to improve the estimation performance and reduce the peaking phenomenon of the classical Linear Extended State Observer (LESO). The proposed GA-FESO consists of a LESO integrated with a fuzzy supervisor designed to automatically adjust the observer bandwidth based on real-time estimation errors. The parameters of the fuzzy supervisor, including membership functions and scaling factors, are optimized by a real-coded genetic algorithm (GA). The integration of fuzzy logic and genetic algorithms into the classical LESO allows the observer to exhibit good transient response and accurate state estimation. As an illustrative application, the GA-FESO is applied to fault detection of the Van Der Pol process, a well-known nonlinear system exhibiting oscillatory behavior. Simulation results demonstrate that the GA-FESO significantly improves state estimation accuracy and fault detection effectiveness, and reduces the peaking phenomenon compared to traditional observer designs.