A TSK-type recurrent fuzzy network for dynamic systems processing via supervised and reinforcement learning

In this paper, a TSK (Takagi-Sugeno-Kang) recurrent fuzzy network (TRFN) structure is proposed. The proposal calls for a design of TRFN under either supervised or reinforcement learning. Set forth first is a recurrent fuzzy network which is developed from a series of recurrent fuzzy IF-THEN rules with TSK-type consequent parts. TRFN design under the two learning environments (supervised and reinforcement) is next advanced. For a TRFN with supervised learning (TRFN-S), an online learning algorithm with a concurrent structure and parameter learning is proposed. For reinforcement learning, a TRFN with genetic learning (TRFN-G) is put forward. To demonstrate the superior properties of TRFNs, the TRFN-S is applied to dynamic system identification and the TRFN-G is applied to dynamic system control, and the efficiency of TRFNs is verified