作者: Gheorghe Puscasu , Bogdan Codres
DOI: 10.1142/S0129065711002869
关键词:
摘要: A new approach for nonlinear system identification and control based on modular neural networks (MNN) is proposed in this paper. The computational complexity of can be greatly reduced if the whole decomposed into several subsystems. This obtained using a partitioning algorithm. Each local model associated with controller. These are also implemented by networks. switching between controllers done dynamical switcher, networks, that tracks different operating points. multiple modelling strategy has been successfully tested simulated laboratory scale liquid-level system.