Modeling and control of carbon monoxide concentration using a neuro-fuzzy technique

摘要: Modeling and control of carbon monoxide (CO) concentration using a neuro-fuzzy technique are discussed. A self-organizing fuzzy identification algorithm (SOFIA) for identifying complex systems such as CO is proposed. The main purpose SOFIA to reduce the computational requirement model. In particular, authors simplify procedure finding optimal structure partition. /spl delta/ rule, which basic learning method in neural networks, used parameter consists four stages effectively realize identification. concretely demonstrated by simple example has been some modeling exercises. result shows effectiveness SOFIA. Next, apply prediction problem air at busiest traffic intersection large city Japan. Prediction results show that model much better than linear Furthermore, simulate system keeping constant level identified self-learning adaptively modifying controller parameters rule introduced because dynamics real changes gradually over long period time. Two controllers designed this simulation. One PI controller. other investigate robustness adaptability disturbance perturbation Simulation more robust adaptive. >